Long Noncoding RNA: Mechanistic Insights and Roles in Inflammation (Advances in Experimental Medicine and Biology, 1363) 303092033X, 9783030920333

Citation preview

Advances in Experimental Medicine and Biology 1363

Susan Carpenter   Editor

Long Noncoding RNA Mechanistic Insights and Roles in Inflammation

Advances in Experimental Medicine and Biology Series Editors Wim E. Crusio, Institut de Neurosciences Cognitives et Intégratives d’Aquitaine, CNRS and University of Bordeaux, Pessac Cedex, France Haidong Dong, Departments of Urology and Immunology Mayo Clinic, Rochester, MN, USA Heinfried H. Radeke, Institute of Pharmacology & Toxicology Clinic of the Goethe University Frankfurt Main, Frankfurt am Main, Hessen, Germany Nima Rezaei, Research Center for Immunodeficiencies, Children’s Medical Center, Tehran University of Medical Sciences, Tehran, Iran Ortrud Steinlein, Institute of Human Genetics, LMU University Hospital Munich, Germany Junjie Xiao, Cardiac Regeneration and Ageing Lab, Institute of Cardiovascular Science, School of Life Science, Shanghai University,  Shanghai, China

Advances in Experimental Medicine and Biology provides a platform for scientific contributions in the main disciplines of the biomedicine and the life sciences. This series publishes thematic volumes on contemporary research in the areas of microbiology, immunology, neurosciences, biochemistry, biomedical engineering, genetics, physiology, and cancer research. Covering emerging topics and techniques in basic and clinical science, it brings together clinicians and researchers from various fields. Advances in Experimental Medicine and Biology has been publishing exceptional works in the field for over 40 years, and is indexed in SCOPUS, Medline (PubMed), Journal Citation Reports/Science Edition, Science Citation Index Expanded (SciSearch, Web of Science), EMBASE, BIOSIS, Reaxys, EMBiology, the Chemical Abstracts Service (CAS), and Pathway Studio. 2020 Impact Factor: 2.622 More information about this series at http://link.springer.com/series/5584

Susan Carpenter Editor

Long Noncoding RNA Mechanistic Insights and Roles in Inflammation

Editor Susan Carpenter Molecular Cell and Developmental Biology University of California, Santa Cruz Santa Cruz, CA, USA

ISSN 0065-2598     ISSN 2214-8019 (electronic) Advances in Experimental Medicine and Biology ISBN 978-3-030-92033-3    ISBN 978-3-030-92034-0 (eBook) https://doi.org/10.1007/978-3-030-92034-0 © Springer Nature Switzerland AG 2022 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

Only since the advent of next generation sequencing have we come to appreciate the vast quantities of noncoding RNAs that are actively transcribed from the genome. Long noncoding RNAs (lncRNAs) represent the largest group of RNAs produced, and we know the function of approximately 3% to date. This book covers the emerging roles for these exciting genes in the context of inflammation. We know that transient inflammation is critical to keeping us healthy; however, any disruption to this process can have devasting affects leading to inflammatory conditions such as rheumatoid arthritis, diabetes, or cardiovascular disease. Here, we will cover the most up-to-date data available from the basics of utilizing sequencing to understand these genes to the role that structure plays in their mechanism of action in addition to the roles that lncRNAs play in contributing to control of the immune system and pathogenesis of inflammatory diseases. We will discuss the potential for use of lncRNAs as biomarkers for disease as well as their promise as targets for future therapeutic intervention of inflammatory diseases. Santa Cruz, CA, USA

Susan Carpenter

v

Contents

Part I Overview 1 Introduction and Overview ������������������������������������������������������������   3 Apple Vollmers and Susan Carpenter Part II Basics of Long Noncoding RNA Classifications and Functions 2 The Complexity of the Mammalian Transcriptome ��������������������  11 Sofie R. Salama 3 Towards Molecular Mechanism in Long Non-coding RNAs: Linking Structure and Function����������������������������������������������������  23 Karissa Sanbonmatsu Part III Long Noncoding RNA and Inflammation 4 Long Non-coding RNAs in Rheumatology������������������������������������  35 Susanne N. Wijesinghe, Mark A. Lindsay, and Simon W. Jones 5 LncRNAs and Cardiovascular Disease������������������������������������������  71 Elizabeth J. Hennessy 6 A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes ����������  97 Thomas C. Brodnicki Part IV Long Noncoding RNAs as Therapeutic Targets 7 LncRNA Biomarkers of Inflammation and Cancer �������������������� 121 Roman E. Reggiardo, Sreelakshmi Velandi Maroli, and Daniel H. Kim 8 Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs�������������������������������������������������������� 147 Ainara Castellanos-Rubio and Sankar Ghosh

vii

viii

9 Long Noncoding RNAs as Therapeutic Targets���������������������������� 161 Jacob B. Pierce, Haoyang Zhou, Viorel Simion, and Mark W. Feinberg Part V Challenges and Future Directions 10 Challenges and Future Directions for LncRNAs and Inflammation���������������������������������������������������������������������������� 179 Haley Halasz and Susan Carpenter Index���������������������������������������������������������������������������������������������������������� 185

Contents

About the Editor

Susan  Carpenter  is Associate Professor of Molecular Cell and Developmental Biology at UC Santa Cruz.Susan’s lab focuses on understanding the complex molecular mechanisms that control the body’s essential protection against infection. Any perturbation to these mechanisms can have devastating consequences and result in inflammatory conditions such as rheumatoid arthritis. Recent evidence shows that a group of RNA molecules known as long noncoding RNA or lncRNA plays important roles in diverse biological functions, including the inflammatory response. Her lab is studying lncRNAs to gain an understanding of their function within the immune system, which could lend insight into human disease and perhaps reveal novel targets for therapeutic intervention for inflammatory conditions.Susan earned her PhD in biochemistry from Trinity College in Dublin, Ireland, studying novel proteins important within innate immune signaling pathways, working under the supervision of Prof. Luke O’Neill. Her postdoctoral research with Prof. Kate Fitzgerald at UMASS medical school continued in this area, and there she focused her attention on the role of lncRNA in inflammatory signaling pathways. Since her research bridges the fields of immunology and RNA biology, she also took the opportunity to work at UCSF in the laboratory of Dr. Michael McManus, who is a pioneer in high throughput technologies. There she worked on developing high throughput genomic approaches to study lncRNAs important for human innate immune signaling in host defense mechanisms.

Editor Biography Susan  Carpenter  is Associate Professor of Molecular Cell and Developmental Biology at the University of California in Santa Cruz. She is an immunologist who studies the role that long noncoding RNAs play in contributing to the innate immune response. For more information on the Carpenter lab, please visit their website https://sites.google.com/a/ucsc.edu/ carpenter-­lab/

ix

Part I Overview

1

Introduction and Overview Apple Vollmers and Susan Carpenter

Abstract

As sequencing technologies improved, new classes of genes were uncovered. Initially, many of these were considered non-functional given their low protein-coding potential but have now emerged as important regulators of biological processes. One of the new classes of genes are called long noncoding RNAs (lncRNAs). LncRNAs are the largest group of transcribed RNA.  As their name suggests, they are non-protein coding genes. To differentiate them from other smaller, noncoding RNAs, lncRNAs are transcripts whose length are greater than 200 nucleotides. According to GENCODE Release 38, there are approximately 18,000 lncRNAs, of which only 4% have a known function. Of the lncRNAs characterized, many of them play regulatory roles in many biological processes, including regulation of gene expression, alternative splicing, chromatin modification, protein activity, and This opening chapter will review the basics of long noncoding RNA and the field of innate immunity and provide an overview of the book and the contents of the individual chapters. A. Vollmers · S. Carpenter (*) Department of Molecular, Cell and Developmental Biology, University of California Santa Cruz, Santa Cruz, CA, USA e-mail: [email protected]

posttranscriptional mechanisms. Compared to protein coding genes, lncRNAs show high cell type specificity. Many lncRNAs have been shown to be expressed in distinct immune cell populations and play RNA-mediated immune-­ regulatory roles. Many aspects of the immune response, including the duration, magnitude, and subsequent return to homeostasis are carefully controlled. Dysregulation of lncRNAs can result in an uncontrolled immune response, which can lead to a variety of immune-related diseases. This introduction aims to summarize the chapters highlighting the discovery of lncRNAs, their role in the immune response, and their functional characterization, either through interaction with DNA, RNA, and/or proteins in distinct immune cell populations or cells implicated in immune-related diseases. Additionally, the immune regulatory role of lncRNAs will be covered, and how lncRNA localization, sequence and secondary structure can inform function. Delving into this largely unexplored field can identify lncRNAs as potential therapeutic targets. Keywords

Long noncoding RNA · Sequencing · Inflammation · Dysregulation

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_1

3

A. Vollmers and S. Carpenter

4

1.1

Introduction

1.1.1 RNA Classification and Function For decades, the majority of biological research has focused on protein-coding genes. RNA was simply thought to be a bridge, a messenger conveying information encoded on DNA to translate into proteins [1]. For many years, proteins were considered the functional workhorse of cells, minimizing the functional potential of RNA. However, with advanced sequencing technologies, findings on an entire new group of noncoding RNA of unclear relevance and unknown function were discovered [2–4]. Follow up studies showed that the RNA itself can be functional, and that many biological processes are controlled by these noncoding RNAs that have little to no protein-coding potential [5, 6]. There is recent evidence to suggest that previously annotated lncRNAs may actually contain short open reading frames encoding small peptides [7–9], however, in general, most lncRNAs appear to be untranslated. The advent of deep sequencing technologies and how they have provided an unprecedented view of the transcriptome through use of RNA sequencing and more recently, single molecule nanopore sequencing or single cell RNA sequencing, is further discussed in Chap. 2. The importance and function of the noncoding transcriptome, which includes a newly identified class of regulatory RNA such as long noncoding RNAs (lncRNAs), has remained largely unexplored. There are approximately 18,000 lncRNAs in the human genome, and yet less than 4% have an ascribed function. However, a large number of the lncRNA that have been identified show important roles in the regulation of gene expression, cell fate, alternative splicing, posttranscriptional mechanisms, and much more [10–13]. Long noncoding RNAs represent the largest class of noncoding RNAs in the human genome. Noncoding RNAs include ribosomal RNA (rRNA), transfer RNA (tRNA) and regulatory RNAs which are arbitrarily classified based on transcript length. LncRNAs are classified as noncoding RNAs that are greater than 200 nucleo-

tides in length to distinguish them from smaller regulatory RNAs such as miRNAs and piRNAs [13–16]. Many of the features of lncRNAs are similar to protein-coding mRNAs in that their transcripts are capped, spliced, and polyadenylated, and typically transcribed by RNA Pol II [6]. But unlike mRNAs, which are evolutionary constrained to maintain an open reading frame to be translated into proteins, lncRNAs are typically not well conserved in terms of primary sequence [17]. As explored in Chap. 2, lncRNA conservation can be measured using comparative genomics. The origin of lncRNAs and their emergence is hypothesized to be due to transformation of a protein-coding gene or potential insertion or duplication event from transposable elements. As such, conservation of lncRNAs can be measured based on the presence of ultraconserved sequences within the gene. One example highlighted in Chap. 2 is Xist, which is one of the most well-characterized lncRNAs that appears to have evolutionarily evolved through a combination of protein-coding gene decay and insertion of transposable elements. Another example of lncRNA conservation demonstrated within the chapter based on identifiable ultraconserved sequences are Alu insertions in lncRNAs, called 1/2sbs RNAs. These repetitive Alu-repeat sequences provide binding sites for Staufen 1-mediated degradation of mRNAs. Additional metrics to assess lncRNA conservation is through genomic synteny, where the position of the neighboring protein coding genes surrounding a lncRNA are conserved, or through conservation of expression or regulatory function across species. One example of this is GAPLINC, a positionally conserved lncRNA found between the same two protein-coding genes, whose levels are downregulated upon human and mouse macrophage stimulation with various Toll-like receptor (TLR) ligands, including lipopolysaccharide (LPS) [18]. GAPLINC depletion shows enhanced baseline expression of NF-κB-dependent immune response genes in both human and mouse macrophages. This enhanced expression of immune response genes provides a protective effect in

1  Introduction and Overview

GAPLINC-KO mice after LPS-induced endotoxic shock. LncRNA conservation can also be measured at the structural level, where secondary structure and other regulatory elements such as enhancers within noncoding regions are conserved [6, 19]. In Chap. 3, the structure-function relationship of lncRNAs is further explored, providing examples of the current techniques used to study RNA to better understand mechanism, including chemical probing, NMR, small angle X-ray scattering, X-ray crystallography and cryoEM. Much of the current work on studying RNA structure has focused on interrogating large protein complexes like the ribosome or spliceosome. There are very few structural studies done to study lncRNAs, and most have been limited to biochemical or low-resolution assays. This chapter will explore examples of lncRNA structure that have recently been solved using high-resolution structural techniques and the impact recently discovered structural motifs can have on lncRNA function. Lastly, this chapter highlights the tools needed to identify the wide range of structures lncRNAs can adopt and how these structures can inform function in terms of specific protein or RNA or DNA binding partners or conformational changes.

1.1.2 LncRNAs and Inflammation Much of the current work on lncRNAs has aimed to better understand their expression, mechanism of action, and function. Emerging studies have shown lncRNAs to be critical regulators of the immune response, and perturbed expression of lncRNAs can impact many immune cell processes, including immune cell differentiation or cell function during inflammation [20, 21]. As the bulk of research has focused on studying protein-­coding genes and inflammation, the role of lncRNAs in immunity has yet to be fully studied. Current research on lncRNAs show high tissue-­specificity in that lncRNA expression is restricted to a particular cell type or condition [22, 23]. The first example of a lncRNA involved in inflammation was lincRNA-Cox2, which was found to be highly induced in innate immune

5

macrophage cells after LPS stimulation and shown to play both a positive and negative regulatory role in modulating proinflammatory cytokines and interferon response genes [24]. Since then, many lncRNAs in various cell types have been implicated in immune-cell differentiation (lnc-DC), epigenetic regulation (XIST), autoimmune disease (THRIL, MALAT1, NEAT1), as well as contribute to a host of genetic factors due to lncRNA dysfunction, increasing the risk of disease. Since lncRNAs can play roles in controlling homeostasis, it is now becoming clear that they can also be dysregulated and associated with autoimmune conditions. Several chapters will cover research on the various roles for lncRNA in autoimmune diseases with an emphasis on Rheumatoid Arthritis (RA), Osteoarthritis (OA), systemic lupus erythematosus (SLE) (Chap. 4) and Diabetes (Chap. 6). Chapter 4 provides an overview of lncRNAs involved in RA, OA, and SLE. All three diseases have shared inflammatory signaling pathways and understanding of lncRNA-mediated regulation of these pathways is still incomplete. Chapter 4 summarizes the immune-regulatory role of lncRNAs as either signals, decoys, or guides through interaction with DNA, RNA, and/or proteins. Many of the most well-characterized lncRNAs such as MALAT1, NEAT1, THRIL, and XIST are highlighted to show how lncRNA expression and function can contribute to the pathogenesis of disease by modulating proinflammatory cytokine expression, either acting as microRNA sponges to control inflammation, or binding to proteins to modulate the cytokine and antiviral response. In addition, other determinants like sex bias driven by lncRNA such as XIST is highlighted, showing how dysregulation of lncRNAs can drive risk of specific diseases. Understanding the signaling pathways and regulatory role of lncRNA in these diseases will be helpful in developing better therapeutics for patients with RA, OA and SLE. Genetic factors in lncRNA dysfunction such as single nucleotide polymorphisms (SNPs) increase risk of disease. Over 90% of disease-­ associated single nucleotide polymorphisms (SNPs) are located in noncoding regions and

6

account for disease phenotypes [25–27]. SNPs can affect binding of a specific TF or protein, create disruptions in splicing or cause changes to polyadenylation [28–30]. In Chap. 6, the role of lncRNAs in Type 1 Diabetes (T1D) in specific beta cell or infiltrating immune cell populations is described. Specifically, this chapter provides an overview on how a majority of T1D-associated SNPs map to noncoding regions of the genome that often encode for lncRNAs. Many of these SNPs show potential to disrupt the secondary structure of lncRNAs, impacting their function to regulate key signaling pathways important for insulin production and secretion, cell development, and apoptosis. Additionally, how lncRNAs regulate inflammation especially during significant crosstalk between beta cells and the infiltrating immune cells is explored. Studies on lncRNA expression and function in T1D elucidates how lncRNAs contribute to a complex signaling network that is important for regulating inflammation-­ induced responses in both beta cells and infiltrating immune cells. Additionally, recent work has highlighted the role of lncRNAs in cardiovascular disease. In Chap. 5, we review the impact lncRNAs can play in regulating cardiac development, lipid metabolism, vasculature development, cholesterol homeostasis  — all processes in which lncRNA dysregulation can promote cardiovascular disease. In this chapter, several key examples of lncRNAs are provided, including Braveheart, a lncRNA whose secondary structure is critical for controlling cardiovascular lineage commitment; defects in the development of cardiac tissue can lead to greater susceptibility for cardiac disease. Another lncRNA implicated in cardiovascular disease and important for maintaining lipid metabolism is highlighted within this chapter. CHROME is one example of a lncRNA that regulates signaling pathways that can affect onset of cholesterol-related pathologies. An accumulation of lipids within blood vessels can lead to atherosclerotic plaques which involves a complex interplay between cells that line the blood vessels, platelets, and infiltrating immune cells. Several lncRNAs including lincRNA-p21 and SMILR have been shown to affect vascular cell prolifera-

A. Vollmers and S. Carpenter

tion and migration. LncRNAs have also been implicated in cholesterol homeostasis, a tightly regulated process, including LeXis, which has been shown to control cholesterol levels, or Lnc-HC, which has been implicated in cholesterol metabolism. Not only do lifestyle factors or levels of cholesterol impact susceptibility to cardiovascular disease, but it has also shown that genetic variants, like SNPs, can also play major roles, and will be further discussed in a later chapter (Chap. 8). As such, the chapters highlighted here show how the study of noncoding RNA remains an active area for future research, especially in the development of future diagnostics and therapies.

1.1.3 LncRNAs and Therapeutics LncRNA expression levels often correlate with the progression or severity of a disease. As such, using lncRNAs as potential biomarkers is attractive for use in developing future therapies. As discussed in Chap. 7, the design and execution of biomarker discovery represents a gap in knowledge in the lncRNA field. In this chapter, the authors review many of the current approaches to using lncRNAs as biomarkers and their recommendations to improve technologies. Key features that make lncRNAs well suited as molecular markers are highlighted, which includes lncRNAs high tissue-specificity, high expression levels under inflammatory or oncogenic conditions, ability to package themselves into extracellular vesicles, and their easy detection in the blood using a variety of sequencing approaches. Given inflammation can be associated with increased risk for many immune-related diseases, including cancer, the authors discuss PCA3 as a promising example of a lncRNA for use as a biomarker. PCA3 is one of the first lncRNAs to be FDA-­ approved as a biomarker for prostate cancer. This chapter will explore the field of biomarkers and discuss the promise that lncRNAs bring to this field. As previously mentioned, the vast majority of SNPs lie inside noncoding regions and can affect binding of proteins through modifications in sec-

1  Introduction and Overview

ondary structure, or create disruptions in RNA processing by causing changes in expression levels or changes in the transcription of different isoforms. In Chap. 8, a more in-depth analysis of what is known about SNPs and lncRNAs and their disease relevance as it pertains to inflammatory diseases is explored. Previous discussion included SNPs associated in T1D (Chap. 6), however, this chapter extends the identification of disease-related lncRNAs to Celiac’s disease, HIV-1, RA, inflammatory bowel disease (IBD), and many more, highlighting the involvement of lncRNAs in the pathogenesis of immune-related disorders. Finally, in Chap. 9, an overview on the development of RNA-based therapies is provided, from the history to current therapies and advances in lncRNA therapeutics. The chapter reviews the range of potential delivery platforms, including adenoviruses, lentiviruses, nanoparticles, and antisense oligonucleotides (ASOs) with design considerations based on toxicity, off-target effects, immunogenicity and efficacy. Additionally, Chap. 9 provides examples of how lncRNAs have been successfully targeted in clinical therapies, highlighting companies like IONIS Pharmaceuticals that have effectively used ASOs to target lncRNAs. In one specific example, the company developed ASOs against a lncRNA that represses the production of a critical motor neuron protein, SMN, which is deficient in spinal muscle atrophy (SMA). Inhibition of this lncRNA resulted in increased protein production of SMN protein contributing to improved clinical outcome. As such, this chapter will explore a broad overview of using RNA-mediated approaches in developing therapeutics and the possible pros and cons to targeting lncRNAs in the clinic. In conclusions this book provides a broad overview of the lncRNA field as it relates to inflammation and inflammatory diseases. From signaling to therapeutics, it covers the gambit in what we know to date about these exciting genes and their potential in helping us better understand inflammatory diseases as well as the role they will play in designing drugs for therapeutic intervention for these devastating conditions.

7

References 1. Crick F (1970) Central dogma of molecular biology. Nature 227:561–563 2. Bertone P, Stolc V, Royce TE et  al (2004) Global identification of human transcribed sequences with genome tiling arrays. Science 306:2242–2246 3. Carninci P, Kasukawa T, Katayama S et al (2005) The transcriptional landscape of the mammalian genome. Science 309:1559–1563 4. Derrien T, Johnson R, Bussotti G et  al (2012) The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Res 22:1775–1789 5. Rinn JL, Chang HY (2012) Genome regulation by long noncoding RNAs. Annu Rev Biochem 81:145–166 6. Statello L, Guo C-J, Chen L-L, Huarte M (2021) Gene regulation by long non-coding RNAs and its biological functions. Nat Rev Mol Cell Biol 22:96–118 7. Anderson DM, Anderson KM, Chang C-L et  al (2015) A micropeptide encoded by a putative long noncoding RNA regulates muscle performance. Cell 160:595–606 8. Matsumoto A, Pasut A, Matsumoto M, Yamashita R, Fung J, Monteleone E, Saghatelian A, Nakayama KI, Clohessy JG, Pandolfi PP (2017) mTORC1 and muscle regeneration are regulated by the LINC00961-­ encoded SPAR polypeptide. Nature 541:228–232 9. Jackson R, Kroehling L, Khitun A et  al (2018) The translation of non-canonical open reading frames controls mucosal immunity. Nature 564:434–438 10. Khalil AM, Guttman M, Huarte M et  al (2009) Many human large intergenic noncoding RNAs associate with chromatin-modifying complexes and affect gene expression. Proc Natl Acad Sci U S A 106:11667–11672 11. Kotzin JJ, Spencer SP, McCright SJ et al (2016) The long non-coding RNA Morrbid regulates Bim and short-lived myeloid cell lifespan. Nature 537:239–243 12. Khan MR, Wellinger RJ, Laurent B (2021) Exploring the alternative splicing of long noncoding RNAs. Trends Genet. https://doi.org/10.1016/j. tig.2021.03.010 13. Mercer TR, Dinger ME, Mattick JS (2009) Long non-­ coding RNAs: insights into functions. Nat Rev Genet 10:155–159 14. Faghihi MA, Wahlestedt C (2009) Regulatory roles of natural antisense transcripts. Nat Rev Mol Cell Biol 10:637–643 15. Whitehead J, Pandey GK, Kanduri C (2009) Regulation of the mammalian epigenome by long noncoding RNAs. Biochim Biophys Acta 1790:936–947 16. Wilusz JE, Sunwoo H, Spector DL (2009) Long noncoding RNAs: functional surprises from the RNA world. Genes Dev 23:1494–1504 17. Johnsson P, Lipovich L, Grandér D, Morris KV (2014) Evolutionary conservation of long non-coding RNAs; sequence, structure, function. Biochim Biophys Acta 1840:1063–1071

8 18. Vollmers AC, Covarrubias S, Kuang D et al (2021) A conserved long noncoding RNA, GAPLINC, modulates the immune response during endotoxic shock. Proc Natl Acad Sci U S A. https://doi.org/10.1073/ pnas.2016648118 19. Novikova IV, Hennelly SP, Tung C-S, Sanbonmatsu KY (2013) Rise of the RNA machines: exploring the structure of long non-coding RNAs. J Mol Biol 425:3731–3746 20. Atianand MK, Fitzgerald KA (2014) Long non-­ coding RNAs and control of gene expression in the immune system. Trends Mol Med 20:623–631 21. Satpathy AT, Chang HY (2015) Long noncoding RNA in hematopoiesis and immunity. Immunity 42:792–804 22. Liu SJ, Horlbeck MA, Cho SW et al (2017) CRISPRi-­ based genome-scale identification of functional long noncoding RNA loci in human cells. Science. https:// doi.org/10.1126/science.aah7111 23. Djebali S, Davis CA, Merkel A et al (2012) Landscape of transcription in human cells. Nature 489:101–108 24. Carpenter S, Aiello D, Atianand MK et  al (2013) A long noncoding RNA mediates both activation

A. Vollmers and S. Carpenter and repression of immune response genes. Science 341:789–792 25. Farh KK-H, Marson A, Zhu J et al (2015) Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature 518:337–343 26. Tak YG, Farnham PJ (2015) Making sense of GWAS: using epigenomics and genome engineering to understand the functional relevance of SNPs in non-coding regions of the human genome. Epigenetics Chromatin 8:57 27. Castellanos-Rubio A, Ghosh S (2019) Disease-­ associated SNPs in inflammation-related lncRNAs. Front Immunol 10:420 28. Buckland PR (2004) Allele-specific gene expression differences in humans. Hum Mol Genet 13(Spec No 2):R255–R260 29. Wang D, Sadee W (2016) CYP3A4 intronic SNP rs35599367 (CYP3A4*22) alters RNA splicing. Pharmacogenet Genomics 26:40–43 30. Thomas LF, Sætrom P (2012) Single nucleotide polymorphisms can create alternative polyadenylation signals and affect gene expression through loss of microRNA-regulation. PLoS Comput Biol 8:e1002621

Part II Basics of Long Noncoding RNA Classifications and Functions

2

The Complexity of the Mammalian Transcriptome Sofie R. Salama

Abstract

Draft genome assemblies for multiple mammalian species combined with new technologies to map transcripts from diverse RNA samples to these genomes developed in the early 2000s revealed that the mammalian transcriptome was vastly larger and more complex than previously anticipated. Efforts to comprehensively catalog the identity and features of transcripts present in a variety of species, tissues and cell lines revealed that a large fraction of the mammalian genome is transcribed in at least some settings. A large number of these transcripts encode long non-coding RNAs (lncRNAs). Many lncRNAs overlap or are anti-sense to protein coding genes and others overlap small RNAs. However, a large number are independent of any previously known mRNA or small RNA. While the functions of a majority of these lncRNAs are unknown, many appear to play roles in gene regulation. Many lncRNAs have species-­ specific and cell type specific expression patterns and their evolutionary origins are varied.

S. R. Salama (*) UC Santa Cruz Genomics Institute, Department of Biomolecular Engineering and Howard Hughes Medical Institute, University of California, Santa Cruz, Santa Cruz, CA, USA e-mail: [email protected]

While technological challenges have hindered getting a full picture of the diversity and transcript structure of all of the transcripts arising from lncRNA loci, new technologies including single molecule nanopore sequencing and single cell RNA sequencing promise to generate a comprehensive picture of the mammalian transcriptome. Keywords

Transcriptome · Long noncoding RNA · Genome · ENCODE · Transposable elements

2.1

Overview

The completion of the first draft of the human genome in 2000 [42] revealed that less than 2% of the genome encodes what is traditionally thought of as genes, i.e. segments encoding messenger RNAs (mRNAs) that code for proteins. Shortly thereafter, microarray based methods for interrogating transcription throughout the genome revealed widespread transcription (summarized in [84]). The subsequent advent of RNA sequencing (RNA-Seq) confirmed the prevalence of so-called non-coding RNA (ncRNA) transcripts in a wide variety of metazoans. Many of these transcripts were found to be transcribed by RNA polymerase II, greater than 200 nucleotides and often multi-exonic, a class of transcripts

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_2

11

12

commonly referred to as long non-coding RNAs (lncRNAs). There has been a lot of debate over the importance and functions of this vast non-­ coding transcriptome, but in the last two decades it has become increasingly clear that many if not most of these transcripts play important roles in gene regulation and function. In this chapter, I will review how the advent of whole genome assemblies coupled with microarray and RNA-Sequencing technologies revealed pervasive transcription of the human genome and enabled the identification of thousands of ncRNAs. After reviewing the various classes of RNAs found in these comprehensive studies and their abundance, I will focus on lncRNAs. I will explore their origin, structure and insights as to their function as revealed by comparative genomics analysis. Next, I will explore unique features of lncRNA expression, which could enable their use as important biomarkers in a variety of applications that require the identification of specific cell types. Finally, I will discuss how a new wave of sequencing technologies promise to create a more complete understanding of mammalian transcriptomes.

S. R. Salama

cally altered by two technological innovations in the 1990s and early 2000s, namely whole genome sequencing and assembly as well as microarray technology. The publication of the first draft human genome [42] and rodent genomes [28, 51] shortly thereafter revealed that a very small proportion of mammalian genomes is devoted to protein coding regions (approximately 2%), however a larger proportion was under purifying selection (5–6%) between all three species. Thus, the amount of conserved non-coding DNA exceeded coding DNA. Some of this excess conserved DNA had hallmarks of cis-acting DNA regulatory elements, but much had no clear hallmark of function. Furthermore, the function of the remaining 95% of the genome was even less clear. Some of this so-called dark matter of the genome was likely to serve lineage-specific and species-­ specific roles and some was postulated to have a structural role, but the large amount of genome sequence with unknown function was surprising and humbling. Meanwhile, microarray technology was being developed. This technology enabled an unbiased look at what parts of the genome are being expressed in a given sample of cellular 2.2 Technology to Explore RNA. Microarrays have now been superseded by high throughput short-read sequencing methods, the Transcriptome but were an important milestone in the emerHistorically, ncRNA research focused on so-­ gence of functional genomics methods. Pioneered called housekeeping noncoding RNAs, which in the labs of Pat Brown and Ron Davis at include ribosomal, transfer, small nuclear and Stanford University, the initial microarrays for small nucleolar RNAs and are usually expressed probing gene expression were arrayed spots of constitutively and at high levels. In the 1980s and oligonucleotides complementary to cDNAs of 1990s, work in Arabidopsis and C. elegans lead interest deposited on glass slides. These were to the discovery of short regulatory noncoding then hybridized with fluorescently labeled RNAs including microRNAs, small interfering cDNAs derived from approximately two microRNAs and Piwi-associated RNAs that are grams of total RNA from a sample of interest involved in silencing of gene expression [3]. In [69]. While the initial experiments only probed addition to these small regulatory ncRNAs a for a few dozen genes of interest, the technology handful of lncRNAs were identified through vari- was rapidly scaled up. This was aided by photolious functional studies, including Xist, a critical thography methods that allow for the in situ synregulator of X-inactivation inactivation in female thesis of oligonucleotides on glass slides as mammals [4, 25] and H19, a non-coding RNA developed by Affymetrix (reviewed in [32]). This involved in the regulation of IGF2 and implicated led to the development of high density oligonuin the progression of multiple cancers [1, 26, 58, cleotide microarrays that could tile across 65]. However, our view of lncRNAs was radi- genomic regions of interest or small genomes

2  The Complexity of the Mammalian Transcriptome

like Escherichia coli and yeasts and eventually to complex vertebrate genomes.

13

complexes allowed for comprehensive analysis of promoters. This revealed many promoters associated with these newly identified ncRNAs [7]. In parallel, there was a large effort led by the 2.3 So Much RNA! Riken Institute to map 5′ ends of RNA pol II transcripts and clone full length cDNAs, initially in Initial efforts to reveal the human transcriptome mice [57]. This effort became the FANTOM utilized tiling arrays covering the non-repetitive project (https://fantom.gsc.riken.jp/), which has regions of newly elucidated regions of human comprehensively characterized cDNAs expressed chromosomes 21 and 22. The arrays literally con- in both mouse and human. Their early work in tained thousands of individual spots containing mice also found evidence of pervasive “forests of probes that represented the entire non-repetitive transcription” that included transcripts both sense portion of the draft sequences of these chromo- and anti-sense to previously characterized prosomes at a resolution of 5–35 base pairs. These tein coding transcripts [6]. These efforts lead to arrays were hybridized with cDNA derived from the annotation of over 50,000 loci in the human RNA samples from a variety of human cell lines genome as potential lncRNAs [34]. Currently in [35] or placenta [66]. The results revealed ten-­ its sixth iteration, FANTOM6, the consortium is fold more transcription than predicted by protein currently focusing large-scale functional analysis coding genes [84]. This finding was confirmed as of lncRNAs. Their pilot study highlights the chalmore chromosomes were examined [8] and simi- lenges of large scale functional analysis of a poplar results were found using microarrays for a ulation of transcripts that are often lowly variety of organisms including plants, Arabidopsis expressed and cell type-specific, but demon[76, 86], and the fruit fly, Drosophila melanogas- strates a functional role for many lncRNAs initer [75]. Salient features of the transcriptomes tially identified solely based on expression [64]. revealed in these studies included a large number of previously unannotated transcripts. In addition, many new transcripts were found overlap- 2.4 ENCODE Defines the Nature ping previously annotated genes including of the Human Transcriptome abundant antisense transcripts [87]. Furthermore, these data suggested a large number of alternative The Encyclopedia of DNA Elements (ENCODE), transcript isoforms associated with known pro- funded by the US National Human Genome tein coding genes. In many cases these additional Research Institute (NHGRI), is another large isoforms had limited coding potential suggesting consortium project that reinforced the notion that that they may serve regulatory functions. For widespread transcription in the human genome example, isoforms that utilize alternative exons has functional significance. Begun in 2003, with or retained introns, which would likely target the the stated goal of identifying the complete set of transcript for nonsense mediated decay [43]. functional elements in the human genome, the Together these initial genome wide transcriptome ENCODE project has comprehensively identified studies suggested pervasive transcription of the promoters, enhancers, genes and associated tranhuman genome, but in many cases the functional scripts. The pilot phase, focusing on 1% of the significance of these transcripts was unclear. humans genome and establishing methodology Further large-scale efforts aimed at mapping and standard protocols also revealed pervasive promoters and cloning transcripts reinforced the transcription of the human genome and defined notion that there is a vast number of transcripts physical features of promoters including DNA beyond the protein coding part of the genome. and histone modification patterns, DNA accessiSubsequent use of the tiling arrays described bility, transcription factor binding patterns and above to probe transcription factor binding [48] long-range interactions that were shared among including by RNA polymerase II preinitiation the newly identified ncRNA transcripts and well-­

14

characterized transcripts associated with protein coding genes [18]. When the phase 2, genome-­ wide results were reported in 2012 with even more extensive characterization and annotation of gene regulatory regions as well as transcriptomes from a variety of cell lines using RNA sequencing, a whopping 80% of the genome was suggested to be “functional” [17] by virtue of its DNA regulatory potential or presence in a transcript. Focusing on the human transcriptome, the systematic analysis of the ENCODE Project provided a number of notable observations [12]. Almost 75% of the genome was transcribed in at least 1 of the 15 human cell lines examined in the project with at most half of these transcripts present in any one cell line. A major challenge in these efforts is going from a signal that a genomic region is transcribed to understanding the transcript structure, especially if multiple transcript isoforms arise from a transcribed region. While microarrays gave way to RNA sequencing (RNA-­ Seq), in which cDNA copies of the RNAs present in the sample are directly sequenced, the use of short read high throughput sequencing approaches limits the information about the original RNA transcripts. Standard RNA-Seq library prep methods result in fragments of cDNA of 100– 300  bp. The advent of paired-end sequencing enabled the identification of the sequence at either end of each library clone. However, given that the average length of an exon is 100  bp, a typical paired-end read usually only encodes the relationship between two exons in a transcript. How these two exons relate to other exons in the original transcript is unclear and must be inferred from the relative abundance of the sequencing signal for each exon and the presence of reads spanning exon-exon junctions. Another issue is determining the ends of transcripts. There are a variety of methods for assembling the transcripts present in this data [33, 78], however these methods are better at defining the boundaries of the exons present in a transcribed locus than identifying specific isoforms and their relative abundances. Despite the limitations inherent in defining transcripts based solely on short read RNA-Seq

S. R. Salama

data, many of the transcribed regions appear to encode lncRNAs. Unlike protein-coding RNAs, the majority of lncRNAs were localized to the nucleus, although many show cytoplasmic localization. In general, lncRNAs showed lower expression levels than protein coding transcripts. Further, they showed significantly more cell type specificity, with 29% being expressed in only one cell line compared to 7% for coding transcripts [12]. The ENCODE project also examined small RNAs: small nuclear (sn)RNAs, small nucleolar (sno)RNAs, micro (mi)RNAs and transfer (t) RNAs. About 6% of annotated long transcripts overlap small RNAs and are presumed to be precursors of these small RNAs. These include both protein coding and lncRNA loci. Interestingly, lncRNAs were found to be enriched as hosts for snoRNAs [12]. The ENCODE project also systematically examined the prevalence of enhancer-associated RNAs (eRNAs) at enhancer loci predicted from ENCODE chromatin immunoprecipitation and high throughput sequencing (ChIP-Seq) data [88]. Transcribed enhancers were associated with relatively stronger signals for H3K4 methylation, H3K27 acetylation, H3K79 dimethylation and RNA Polymerase II binding. Further, both these DNA features and transcription were cell type specific. Transcripts initiated within the enhancer, continued outwards on either strand, in some cases, for several kilobases and were observed in both polyadenylated and non-polyadenylated fractions [12]. Further analysis or eRNAs has suggested they can play important roles in regulating transcription from promoters associated with the enhancers from which they arise, although in many settings it remains controversial as to whether the act of transcription or the resulting eRNA/ncRNA plays a functional role (reviewed in [22, 68]). eRNAs have been directly implicated in the recruitment and/or activation of transcription factors such as YY1 [71] and coactivators like BRD4 [63] and CBP [2] as well as in promoting chromatin looping and chromatin accessibility domains through associations with cohesin [45, 50]. Alongside the ENCODE project, the GENCODE project (https://www.gencodegenes.

2  The Complexity of the Mammalian Transcriptome A

15 C

B

Fig. 2.1  Overview of the human transcriptome based on GENCODE v36. (a) Number of RNA transcripts encoding mRNAs (coding), lncRNAs, or small RNAs including, snRNA, snoRNA, rRNA, tRNA, and miRNA. (b) Coverage of the three classes of RNAs in base pairs (bp) in the human genome including intronic sequences. The total coverage of Gencode annotated transcribed

sequences is 1,804,537,538  bp. (c) Overlap between lncRNA loci and protein coding and small RNA loci. LncRNAs that overlap a protein coding gene are shown in blue, those overlapping a small RNA shown in green and lncRNAs that don’t overlap another transcript are shown in pink

org/) aims to annotate all coding and non-coding genes identified in the human and mouse genomes. While taking advantage of automated annotation pipelines and large-scale projects like ENCODE and FANTOM, GENCODE incorporates extensive manual curation to generate a gold standard gene annotation set [31]. At the time of the ENCODE phase 2 release (GENCODE v7), the project had annotated 9277 lncRNA genes and 14,880 transcripts [11, 17]. The most recent human genome release (GENCODE v36, 5/2020) contains 17,958 lncRNA genes and 48,734 lncRNA loci transcripts. For comparison, there were 20,687 protein coding loci in GENCODE v7 and v36 contains 19,962 (Fig. 2.1). In mouse, the other mammalian species with extensive transcriptome data, the most recent annotation set (GENCODE vM25, 4/2020) includes 21,859 protein coding genes and 13,197 lncRNA genes [24]. These current annotation sets continue to be an underestimate of the true diversity of lncRNAs expressed during development and adulthood. Indeed, LNCipedia.org, which aggregates lncRNA gene and transcript annotations from a wide variety of sources lists 56,946 human lncRNA genes in its most recent set (version 5.2). Still, 1.8 billion bases of the human genome are part of the more conservative Gencode annotated transcript set, which including introns represents 56% of our 3.2 billion base genome. Annotated lncRNAs represent almost a third of these transcribed bases.

Taken together, the confluence of nearly complete genome assemblies and high throughput sequencing technologies has enabled a large number of innovative methods to explore the mammalian transcriptome. This work has revealed a complex tapestry of transcription in the human genome and a substantial fraction of these transcripts are lncRNAs.

2.5

 here Do lncRNAs Come W from?

As mentioned above, a majority of lncRNA genes are not highly conserved across significant evolutionary distances such as between human and mouse. Further, even among highly conserved lncRNAs, there is often limited primary sequence conservation in exons and in promoter sequences. Several non-mutually exclusive hypotheses have been proposed for the emergence of lncRNAs: (i) transformation of a protein-coding genes; (ii) duplication of another lncRNA; (iii) de novo origin from sequences previously untranscribed or devoid of exonic sequences, possibly through recombination; (iv) emergence from transposable element (TE) sequences. Individual examples illustrating each of these mechanisms have been described [36, 61]. For example, the highly characterized lncRNA, Xist appears to have evolved from a combination of protein-coding gene decay and transposable element insertion [16].

S. R. Salama

16

An early examination of features of lncRNA genes as a class revealed that a large fraction overlap TE sequences [37]. In their study of 9241 human lncRNAs from a dataset including 28 tissues and cell lines, 83% contained TE-derived sequences. In particular, lncRNAs expressed in pluripotent stem cells were enriched for recent human endogenous retroviruses (HERVs). Many of these lncRNAs use the HERV long terminal repeat (LTR) sequence for transcript initiation suggesting that these retrotransposons provide regulatory signals for newly evolved lncRNAs. Similarly primate-specific LINE 1 elements (L1PAs) are enriched at lncRNA transcriptional start sites, likely providing an antisense promoter to initiate transcription. This study also noted that Alu elements, especially AluY, tend to be about 250 nucleotides downstream from 3′ ends and likely provide a polyadenylation signal. Further examination of transposable elements associated with lncRNAs has reinforced the notion that they contribute to the  genesis and function of lncRNAs. Kapusta and colleagues [36] revealed that TE content in 5′ untranslated regions (UTRs) and 3′ UTRs was significantly higher than in the UTRs of protein coding genes. Additionally, TEs exapted into lncRNAs tended to be older than TEs in the genome overall. While the representation of TEs in lncRNAs tends to mirror their overall representation in the genome, ERVs are overrepresented. For example, the HERVH LTR (LTR7) contains binding sites for the pluripotency transcription factors OCT4 and NANOG and is associated with hundreds of lncRNAs highly expressed in pluripotent stem cell including lncRNA-RoR [46, 55] and lncRNAES3 [52], which both play roles in maintaining pluripotency. Interestingly, RoR is overexpressed in a variety of cancers and tumors often show deregulation of retrotransposons. TEs can also contribute secondary structure potential to lncRNAs. This is often associated with Alu insertions and correlated with A to I editing [36]. A subset of Alu-containing lncRNAs, called 1/2sbs RNAs, can base pair with Alu elements in the 3′ UTR of protein coding genes, thereby creating a binding site for the Staufen1-

mediated RNA decay machinery, which in turn promote post-transcriptional repression of the targeted mRNAs [29]. The importance of lineage specific TEs in lncRNA evolution and function is not restricted to humans and the primate lineage. Similar analysis of mouse and zebrafish lncRNAs reveals a high prevalence of TE sequences in lncRNA genes and enrichment of lineage-specific categories of TEs [36]. Even among lncRNAs conserved across long evolutionary distances, TEs appear to play an important role. An example is cyrano (OIP5-AS1), a lncRNA originally identified in zebrafish that contains an ultraconserved element shared with human and mouse [80]. This element confers binding to mir-7 microRNAs and has been proposed to play roles in neural development and pluripotency through its interactions with mir-7  in mouse and zebrafish [39, 73, 80]. In each species this exon contains lineage specific TEs flanking the ultraconserved sequence suggesting this core mir-7 binding function has diversified in a lineage-specific manner [36]. Indeed in human HeLa cells, cyrano, has been shown to be a competing endogenous (ce)RNA for the RNA binding protein HuR and in that context to repress HuR’s proliferative functions [38]. While there certainly are lncRNAs where TEs do not play a clear role in their evolution or function, they play important roles in the emergence and specialization of many lncRNAs across vertebrates.

2.6

Cell Type Specificity of lncRNAs and Its Implications

While some of the best studied lncRNAs are highly and ubiquitously expressed, a striking property of the lncRNAs identified in the genome-­ wide transcriptome studies mentioned above was their cell type specificity [5, 11, 59, 61]. This finding has held up over time. In particular, as single cell RNA sequencing (scRNA-Seq) methods have come into wide use, it has become clear that many lncRNAs show exquisite cell type specific expression. The finding that lncRNAs tend

2  The Complexity of the Mammalian Transcriptome

to be more lowly expressed than protein coding genes seems to be due, in part, to the fact that they often are expressed in a highly cell type and cell state specific manner, explaining their low overall expression in bulk tissue RNA sequencing experiments. As an example, the human brain is a highly complex tissue with many cell types. Further, for the best understood class of cells, neurons, there are hundreds, if not thousands, of subtypes defined by their neurotransmitter expression, morphology, connectivity, and location in the brain. The development and function of these diverse cell types is a subject of intense study and has been aided in recent years by single cell RNA-Seq atlases of human brain tissue [10, 19, 47, 53]. Further, the development and characterization of pluripotent stem cell derived cerebral organoid cultures has shown them to be an experimentally tractable model that recapitulates many features of the developing brain including the generation of diverse cell types and the timing of their appearance [15, 41, 60]. These studies have revealed that lncRNAs often show robust expression in specific cell types and indeed can be a diagnostic marker of specific cell types and are often dynamically expressed during development [23, 44]. These findings support the notion that our current estimates of lncRNAs continue to be an underestimate and as scRNA-seq studies in diverse tissues proliferate new lncRNAs will be identified. This poses a challenge as current methods for scRNA-Seq data analysis generally rely on pre-existing transcript annotations from trusted sources, like GENCODE which are heavily biased towards transcripts identified from bulk RNA-Seq experiments in cell lines and whole tissues. This concern was born out in a recent study where we specifically looked for lncRNAs expressed during early cerebral cortex development by performing RNA-Seq on weekly time points during the first 5 weeks of cerebral cortex organoid development [23] in multiple primate species including Rhesus macaque, orangutan, chimpanzee and human. We used Cufflinks [78] and the Comparative Annotation Toolkit (CAT)

17

[20] to identify both previously identified and novel transcripts and compare them between species. This analysis found almost 2975 poly-­ exonic lncRNAs of which 503 were human-specific, with the remainder found in at least one other primate (Fig. 2.2a, b). Interestingly, 347 of these loci were previously unannotated. Two thirds of these unannotated transcripts were found in other primates suggesting a conservation of function. We focused on the human lncRNAs that were transiently expressed during the earliest stages of cerebral cortex development in which the neuroepithelium proliferates and gives rise to the pioneer neurons of the cerebral cortex, the Cajal-Retzius neurons, as well as radial glia cells, the cortical stem cells that will give rise to the excitatory projection neurons of the cortical plate and glial cell types (i.e. astrocytes, oligodendrocytes) (Fig. 2.2c, d). We found that that many of these lncRNAs showed similar dynamic expression patterns across the primate species we examined and that the expression of these lncRNAs was cell type specific in scRNA-­ ­ Seq experiments. Furthermore, ectopically expressing these lncRNAs altered the expression of genes associated with these early cell types suggesting a functional role for these highly specific and transiently expressed lncRNAs. Going forward, it will be important to use methodologies that allow for the identification of novel lncRNAs when performing experiments examining understudied cell populations. However, this cell type specificity on lncRNA production opens up the possibility of using lncRNA expression signatures as a diagnostic tool to assess cell populations in both research and clinical settings. An exciting development in this regard is the finding that lncRNAs are found in the cell-free RNA and exosomal RNA fractions of bodily fluids (i.e. blood, urine). As discussed in later chapters these findings hold promise for using lncRNAs as part of minimally invasive tests for patients with diseases associated with the emergence and proliferation of altered cell populations, like cancers (also reviewed in [62, 81, 89]).

S. R. Salama

18

Fig. 2.2  Discovery of lncRNAs expressed during primate cerebral cortex differentiation (a) Data analysis pipeline for discovering lncRNA expressed in early cerebral cortex development across primates. Bulk RNA-Seq generated from RNA harvested from duplicate cerebral cortex organoid cultures served as the input data. (b) Venn diagram showing the number of lncRNAs detected in this data set and their conservation of expression across primates. (c) A heatmap of select lncRNAs expressed in at

2.7

Future Outlook

While much progress has been made, a complete understanding of the diversity of the mammalian transcriptome is still a ways away. However, recent developments promise that a more complete picture will be coming soon. One obstacle has been the fact that we still lack a complete understanding of the human genome and even less so that of other mammalian species. Parts of the genome that are poorly represented include segmental duplications, repetitive DNA around the centromeres and transposable element derived

least 2 species whose expression in human peaks at week 2 selected for functional analysis. All showed cell type specific expression in human week 2 cerebral cortex organoid scRNA-Seq experiments as shown for 4 examples in (d). The left panel shows a TSNE plot where each cell is colored by cell type as assessed by clustering and cell-type specific marker expression. Brown shading in the panels at the right indicates increasing expression of the indicated lncRNA in each cell. (Adapted from [23])

sequences. These missing sequences show high association with human disease and are a major source of species specific genes [13, 21, 54, 67]. Significant progress has been made using a variety of long read sequencing approaches to generate vastly more complete primate genomes [30, 40, 83] than the draft human genome assembly used for the early studies of the human transcriptome. Recently, a telomere to telomere sequence of the human X chromosome was published [49]. This methodology is being applied to generate telomere to telomere assemblies of not just one, but 300 human genomes representing the diversity of humanity (https://humanpangenome.org/).

2  The Complexity of the Mammalian Transcriptome

These more complete genomic resources will enable the discovery of new transcripts, but also discoveries related to their evolution and how they vary within our population. Another obstacle has been that the techniques to precisely determine the structure of transcripts arising from transcribed regions generated in genome wide transcriptome studies have been laborious and difficult to scale. The advent of single molecule long read cDNA sequencing methods using the PacBio [70, 79] and Oxford Nanopore [9, 56, 82] platforms enable sequencing of full length cDNA copies of RNAs which can definitively establish isoform structures. Even better, direct RNA sequencing of mRNAs has been achieved using Oxford Nanopore technology [27, 74, 85], thus eliminating biases introduced by going through a cDNA synthesis step. Furthermore, single molecule direct RNA seq can reveal base modifications to gain a true picture of RNAs as they exist in the cell [72]. Currently, direct RNA sequencing requires large amounts of input RNA and has an accuracy profile significantly below that of DNA sequencing. However these issues will likely be overcome in the near future. Finally, a complete understanding of an organism’s transcriptome requires profiling the diverse cell types that make up the developing and adult tissues of the body. Since the first single cell RNA sequencing paper was published in 2009 [77], there has been rapid progress in the technique. Aided by commercial instruments from Fluidigm [14] and 10X genomics [90], the technique has been embraced by the biomedical research community. Furthermore, there have been significant efforts to catalog scRNA-Seq data, particularly for human and mouse, as well as provide standards and resources for experimental setup and data analysis by organizations like the Human Cell Atlas (https://www.humancellatlas.org/), and the Tabla Muris (https://tabula-­muris.ds.czbiohub.org/) for mouse. The last two decades have brought a revolution in our understanding of the diversity of the mammalian transcriptome. The major contours have been defined in terms of the prevalence and

19

classes or transcripts. However, in the coming decades, I anticipate many new insights. More complete and representative genomes will reveal how variation affects the expression and function of RNA transcripts. Direct RNA sequencing will reveal the extent to which transcripts vary in their isoforms and RNA modifications and enable experiments to explore the consequences of that variation. Finally as we gain a more comprehensive view of the diversity of the cell types present during development and adulthood, we will approach a more comprehensive understanding of the transcriptome. Acknowledgements  The author wishes to thank Andrew Field for helpful discussions to define the scope of this article, and Mark Diekhans for discussions about GENCODE and for assistance generating Fig.  2.1. This work was supported by R01HG010329 and R01MH120295 to SRS.

References 1. Bartolomei MS, Zemel S, Tilghman SM (1991) Parental imprinting of the mouse H19 gene. Nature 351:153–155. https://doi.org/10.1038/351153a0 2. Bose DA, Donahue G, Reinberg D et al (2017) RNA binding to CBP stimulates histone acetylation and transcription. Cell 168:135–149..e22. https://doi. org/10.1016/j.cell.2016.12.020 3. Boyd SD (2008) Everything you wanted to know about small RNA but were afraid to ask. Lab Investig 88:569–578. https://doi.org/10.1038/ labinvest.2008.32 4. Brown CJ, Ballabio A, Rupert JL et  al (1991) A gene from the region of the human X inactivation centre is expressed exclusively from the inactive X chromosome. Nature 349:38–44. https://doi. org/10.1038/349038a0 5. Cabili MN, Trapnell C, Goff L et al (2011) Integrative annotation of human large intergenic noncoding RNAs reveals global properties and specific subclasses. Genes Dev 25:1915–1927. https://doi. org/10.1101/gad.17446611 6. Carninci P, Kasukawa T, Katayama S et al (2005) The transcriptional landscape of the mammalian genome. Science 309:1559–1563. https://doi.org/10.1126/ science.1112014 7. Cawley S, Bekiranov S, Ng HH et  al (2004) Unbiased mapping of transcription factor binding sites along human chromosomes 21 and 22 points to widespread regulation of noncoding RNAs. Cell 116:499–509. https://doi.org/10.1016/ s0092-­8674(04)00127-­8

20 8. Cheng J, Kapranov P, Drenkow J et  al (2005) Transcriptional maps of 10 human chromosomes at 5-nucleotide resolution. Science 308:1149–1154. https://doi.org/10.1126/science.1108625 9. Cole C, Byrne A, Adams M et  al (2020) Complete characterization of the human immune cell transcriptome using accurate full-length cDNA sequencing. Genome Res 30:589–601. https://doi.org/10.1101/ gr.257188.119 10. Darmanis S, Sloan SA, Zhang Y et al (2015) A survey of human brain transcriptome diversity at the single cell level. Proc Natl Acad Sci U S A 112:7285–7290. https://doi.org/10.1073/pnas.1507125112 11. Derrien T, Johnson R, Bussotti G et  al (2012) The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Res 22:1775–1789. https://doi. org/10.1101/gr.132159.111 12. Djebali S, Davis CA, Merkel A et al (2012) Landscape of transcription in human cells. Nature 489:101–108. https://doi.org/10.1038/nature11233 13. Dougherty ML, Nuttle X, Penn O et  al (2017) The birth of a human-specific neural gene by incomplete duplication and gene fusion. Genome Biol 18:49. https://doi.org/10.1186/s13059-­017-­1163-­9 14. Durruthy-Durruthy R, Ray M (2018) Using Fluidigm C1 to generate single-cell full-length cDNA libraries for mRNA sequencing. Methods Mol Biol 1706:199– 221. https://doi.org/10.1007/978-­1-­4939-­7471-­9_11 15. Eiraku M, Watanabe K, Matsuo-Takasaki M et  al (2008) Self-organized formation of polarized cortical tissues from ESCs and its active manipulation by extrinsic signals. Cell Stem Cell 3:519–532. https:// doi.org/10.1016/j.stem.2008.09.002 16. Elisaphenko EA, Kolesnikov NN, Shevchenko AI et  al (2008) A dual origin of the Xist gene from a protein-­coding gene and a set of transposable elements. PLoS One 3:e2521. https://doi.org/10.1371/ journal.pone.0002521 17. ENCODE Project Consortium (2012) An inte grated encyclopedia of DNA elements in the human genome. Nature 489:57–74. https://doi.org/10.1038/ nature11247 18. ENCODE Project Consortium, Birney E, Stamatoyannopoulos JA et  al (2007) Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature 447:799–816. https://doi.org/10.1038/nature05874 19. Eze UC, Bhaduri A, Haeussler M et  al (2021) Single-cell atlas of early human brain development highlights heterogeneity of human neuroepithelial cells and early radial glia. Nat Neurosci. https://doi. org/10.1038/s41593-­020-­00794-­1 20. Fiddes IT, Armstrong J, Diekhans M et  al (2018a) Comparative Annotation Toolkit (CAT)-simultaneous clade and personal genome annotation. Genome Res 28:1029–1038. https://doi.org/10.1101/gr.233460.117 21. Fiddes IT, Lodewijk GA, Mooring M et  al (2018b) Human-specific NOTCH2NL genes affect Notch signaling and cortical neurogenesis. Cell 173:1356– 1369.e22. https://doi.org/10.1016/j.cell.2018.03.051

S. R. Salama 22. Field A, Adelman K (2020) Evaluating enhancer function and transcription. Annu Rev Biochem. https:// doi.org/10.1146/annurev-­biochem-­011420-­095916 23. Field AR, Jacobs FMJ, Fiddes IT et  al (2019) Structurally conserved primate LncRNAs are transiently expressed during human cortical differentiation and influence cell-type-specific genes. Stem Cell Rep 12:245–257. https://doi.org/10.1016/j. stemcr.2018.12.006 24. Frankish A, Diekhans M, Ferreira A-M et  al (2019) GENCODE reference annotation for the human and mouse genomes. Nucleic Acids Res 47:D766–D773. https://doi.org/10.1093/nar/gky955 25. Furlan G, Rougeulle C (2016) Function and evolution of the long noncoding RNA circuitry orchestrating X-chromosome inactivation in mammals. Wiley Interdiscip Rev RNA 7:702–722. https://doi. org/10.1002/wrna.1359 26. Gabory A, Ripoche M-A, Yoshimizu T, Dandolo L (2006) The H19 gene: regulation and function of a non-coding RNA.  Cytogenet Genome Res 113:188– 193. https://doi.org/10.1159/000090831 27. Garalde DR, Snell EA, Jachimowicz D et  al (2018) Highly parallel direct RNA sequencing on an array of nanopores. Nat Methods 15:201–206. https://doi. org/10.1038/nmeth.4577 28. Gibbs RA, Weinstock GM, Metzker ML et al (2004) Genome sequence of the Brown Norway rat yields insights into mammalian evolution. Nature 428:493– 521. https://doi.org/10.1038/nature02426 29. Gong C, Maquat LE (2011) lncRNAs transactivate STAU1-mediated mRNA decay by duplexing with 3’ UTRs via Alu elements. Nature 470:284–288. https:// doi.org/10.1038/nature09701 30. Gordon D, Huddleston J, Chaisson MJP et al (2016) Long-read sequence assembly of the gorilla genome. Science 352:aae0344. https://doi.org/10.1126/science.aae0344 31. Harrow J, Frankish A, Gonzalez JM et  al (2012) GENCODE: the reference human genome annotation for the ENCODE project. Genome Res 22:1760– 1774. https://doi.org/10.1101/gr.135350.111 32. Heller MJ (2002) DNA microarray technology: devices, systems, and applications. Annu Rev Biomed Eng 4:129–153. https://doi.org/10.1146/annurev. bioeng.4.020702.153438 33. Hölzer M, Marz M (2019) De novo transcriptome assembly: a comprehensive cross-species comparison of short-read RNA-Seq assemblers. Gigascience 8. https://doi.org/10.1093/gigascience/giz039 34. Hon C-C, Ramilowski JA, Harshbarger J et al (2017) An atlas of human long non-coding RNAs with accurate 5′ ends. Nature 543:199–204. https://doi. org/10.1038/nature21374 35. Kapranov P, Cawley SE, Drenkow J et  al (2002) Large-scale transcriptional activity in chromosomes 21 and 22. Science 296:916–919. https://doi. org/10.1126/science.1068597 36. Kapusta A, Kronenberg Z, Lynch VJ et  al (2013) Transposable elements are major contributors to the origin, diversification, and regulation of vertebrate

2  The Complexity of the Mammalian Transcriptome long noncoding RNAs. PLoS Genet 9:e1003470. https://doi.org/10.1371/journal.pgen.1003470 37. Kelley D, Rinn J (2012) Transposable elements reveal a stem cell-specific class of long noncoding RNAs. Genome Biol 13:R107. https://doi.org/10.1186/ gb-­2012-­13-­11-­r107 38. Kim J, Abdelmohsen K, Yang X et al (2016) LncRNA OIP5-AS1/cyrano sponges RNA-binding protein HuR.  Nucleic Acids Res 44:2378–2392. https://doi. org/10.1093/nar/gkw017 39. Kleaveland B, Shi CY, Stefano J, Bartel DP (2018) A network of noncoding regulatory RNAs acts in the mammalian brain. Cell 174:350–362.e17. https://doi. org/10.1016/j.cell.2018.05.022 40. Kronenberg ZN, Fiddes IT, Gordon D et  al (2018) High-resolution comparative analysis of great ape genomes. Science 360. https://doi.org/10.1126/science.aar6343 41. Lancaster MA, Knoblich JA (2014) Generation of cerebral organoids from human pluripotent stem cells. Nat Protoc 9:2329–2340. https://doi.org/10.1038/ nprot.2014.158 42. Lander ES, Linton LM, Birren B et al (2001) Initial sequencing and analysis of the human genome. Nature 409:860–921. https://doi.org/10.1038/35057062 43. Lewis BP, Green RE, Brenner SE (2003) Evidence for the widespread coupling of alternative splicing and nonsense-mediated mRNA decay in humans. Proc Natl Acad Sci U S A 100:189–192. https://doi. org/10.1073/pnas.0136770100 44. Liu SJ, Nowakowski TJ, Pollen AA et  al (2016) Single-cell analysis of long non-coding RNAs in the developing human neocortex. Genome Biol 17:67. https://doi.org/10.1186/s13059-­016-­0932-­1 45. Li W, Notani D, Ma Q et al (2013) Functional roles of enhancer RNAs for oestrogen-dependent transcriptional activation. Nature 498:516–520. https://doi. org/10.1038/nature12210 46. Loewer S, Cabili MN, Guttman M et al (2010) Large intergenic non-coding RNA-RoR modulates reprogramming of human induced pluripotent stem cells. Nat Genet 42:1113–1117. https://doi.org/10.1038/ ng.710 47. Maitra M, Nagy C, Turecki G (2019) Sequencing the human brain at single-cell resolution. Curr Behav Neurosci Rep 6:197–208. https://doi.org/10.1007/ s40473-­019-­00192-­3 48. Martone R, Euskirchen G, Bertone P et  al (2003) Distribution of NF-kappaB-binding sites across human chromosome 22. Proc Natl Acad Sci U S A 100:12247–12252. https://doi.org/10.1073/ pnas.2135255100 49. Miga KH, Koren S, Rhie A et  al (2020) Telomere-­ to-­telomere assembly of a complete human X chromosome. Nature 585:79–84. https://doi.org/10.1038/ s41586-­020-­2547-­7 50. Mousavi K, Zare H, Dell’orso S et al (2013) eRNAs promote transcription by establishing chromatin accessibility at defined genomic loci. Mol Cell 51:606–617. https://doi.org/10.1016/j.molcel.2013.07.022

21 51. Mouse Genome Sequencing Consortium, Waterston RH, Lindblad-Toh K et  al (2002) Initial sequencing and comparative analysis of the mouse genome. Nature 420:520–562. https://doi.org/10.1038/ nature01262 52. Ng S-Y, Johnson R, Stanton LW (2012) Human long non-coding RNAs promote pluripotency and neuronal differentiation by association with chromatin modifiers and transcription factors. EMBO J 31:522–533. https://doi.org/10.1038/emboj.2011.459 53. Nowakowski TJ, Bhaduri A, Pollen AA et al (2017) Spatiotemporal gene expression trajectories reveal developmental hierarchies of the human cortex. Science 358:1318–1323. https://doi.org/10.1126/science.aap8809 54. Nuttle X, Giannuzzi G, Duyzend MH et  al (2016) Emergence of a Homo sapiens-specific gene family and chromosome 16p11.2 CNV susceptibility. Nature 536:205–209. https://doi.org/10.1038/nature19075 55. Ohnuki M, Tanabe K, Sutou K et al (2014) Dynamic regulation of human endogenous retroviruses mediates factor-induced reprogramming and differentiation potential. Proc Natl Acad Sci U S A 111:12426– 12431. https://doi.org/10.1073/pnas.1413299111 56. Oikonomopoulos S, Wang YC, Djambazian H et  al (2016) Benchmarking of the Oxford Nanopore MinION sequencing for quantitative and qualitative assessment of cDNA populations. Sci Rep 6:31602. https://doi.org/10.1038/srep31602 57. Okazaki Y, Furuno M, Kasukawa T et  al (2002) Analysis of the mouse transcriptome based on functional annotation of 60,770 full-length cDNAs. Nature 420:563–573. https://doi.org/10.1038/nature01266 58. Pachnis V, Belayew A, Tilghman SM (1984) Locus unlinked to alpha-fetoprotein under the control of the murine raf and Rif genes. Proc Natl Acad Sci U S A 81:5523–5527. https://doi.org/10.1073/ pnas.81.17.5523 59. Pauli A, Valen E, Lin MF et al (2012) Systematic identification of long noncoding RNAs expressed during zebrafish embryogenesis. Genome Res 22:577–591. https://doi.org/10.1101/gr.133009.111 60. Pollen AA, Bhaduri A, Andrews MG et  al (2019) Establishing cerebral organoids as models of human-­ specific brain evolution. Cell 176:743–756.e17. https://doi.org/10.1016/j.cell.2019.01.017 61. Ponting CP, Oliver PL, Reik W (2009) Evolution and functions of long noncoding RNAs. Cell 136:629– 641. https://doi.org/10.1016/j.cell.2009.02.006 62. Rahat B, Ali T, Sapehia D, Mahajan A, Kaur J. Circulating Cell-Free Nucleic Acids as Epigenetic Biomarkers in Precision Medicine. Front Genet. 2020 Aug 11;11:844. https://doi.org/10.3389/ fgene.2020.00844. PMID: 32849827; PMCID: PMC7431953 63. Rahnamoun H, Lee J, Sun Z et  al (2018) RNAs interact with BRD4 to promote enhanced chromatin engagement and transcription activation. Nat Struct Mol Biol 25:687–697. https://doi.org/10.1038/ s41594-­018-­0102-­0

22 64. Ramilowski JA, Yip CW, Agrawal S et  al (2020) Functional annotation of human long noncoding RNAs via molecular phenotyping. Genome Res 30:1060–1072. https://doi.org/10.1101/gr.254219.119 65. Raveh E, Matouk IJ, Gilon M, Hochberg A (2015) The H19 long non-coding RNA in cancer initiation, progression and metastasis  - a proposed unifying theory. Mol Cancer 14:184. https://doi.org/10.1186/ s12943-­015-­0458-­2 66. Rinn JL, Euskirchen G, Bertone P et  al (2003) The transcriptional activity of human Chromosome 22. Genes Dev 17:529–540. https://doi.org/10.1101/ gad.1055203 67. Samonte RV, Eichler EE (2002) Segmental duplications and the evolution of the primate genome. Nat Rev Genet 3:65–72. https://doi.org/10.1038/nrg705 68. Sartorelli V, Lauberth SM (2020) Enhancer RNAs are an important regulatory layer of the epigenome. Nat Struct Mol Biol 27:521–528. https://doi.org/10.1038/ s41594-­020-­0446-­0 69. Schena M, Shalon D, Davis RW, Brown PO (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270:467–470. https://doi.org/10.1126/ science.270.5235.467 70. Sharon D, Tilgner H, Grubert F, Snyder M (2013) A single-molecule long-read survey of the human transcriptome. Nat Biotechnol 31:1009–1014. https://doi. org/10.1038/nbt.2705 71. Sigova AA, Abraham BJ, Ji X et  al (2015) Transcription factor trapping by RNA in gene regulatory elements. Science 350:978–981. https://doi. org/10.1126/science.aad3346 72. Smith AM, Jain M, Mulroney L et al (2019) Reading canonical and modified nucleobases in 16S ribosomal RNA using nanopore native RNA sequencing. PLoS One 14:e0216709. https://doi.org/10.1371/journal. pone.0216709 73. Smith KN, Starmer J, Miller SC et  al (2017) Long noncoding RNA moderates MicroRNA activity to maintain self-renewal in embryonic stem cells. Stem Cell Rep 9:108–121. https://doi.org/10.1016/j. stemcr.2017.05.005 74. Soneson C, Yao Y, Bratus-Neuenschwander A et  al (2019) A comprehensive examination of Nanopore native RNA sequencing for characterization of complex transcriptomes. Nat Commun 10:3359. https:// doi.org/10.1038/s41467-­019-­11272-­z 75. Stolc V, Gauhar Z, Mason C et  al (2004) A gene expression map for the euchromatic genome of Drosophila melanogaster. Science 306:655–660. https://doi.org/10.1126/science.1101312 76. Stolc V, Samanta MP, Tongprasit W et  al (2005) Identification of transcribed sequences in Arabidopsis thaliana by using high-resolution genome tiling arrays. Proc Natl Acad Sci U S A 102:4453–4458. https://doi.org/10.1073/pnas.0408203102 77. Tang F, Barbacioru C, Wang Y et al (2009) mRNA-­ Seq whole-transcriptome analysis of a single cell.

S. R. Salama Nat Methods 6:377–382. https://doi.org/10.1038/ nmeth.1315 78. Trapnell C, Roberts A, Goff L et al (2012) Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc 7:562–578. https://doi.org/10.1038/nprot.2012.016 79. Tseng E, Underwood JG, Tseng E (2013) Full length cDNA sequencing on the PacBio® RS. J Biomol Tech 24:S45 80. Ulitsky I, Shkumatava A, Jan CH et  al (2011) Conserved function of lincRNAs in vertebrate embryonic development despite rapid sequence evolution. Cell 147:1537–1550. https://doi.org/10.1016/j. cell.2011.11.055 81. Vlassov AV, Magdaleno S, Setterquist R, Conrad R (2012) Exosomes: current knowledge of their composition, biological functions, and diagnostic and therapeutic potentials. Biochim Biophys Acta 1820:940–948. https://doi.org/10.1016/j. bbagen.2012.03.017 82. Volden R, Palmer T, Byrne A et al (2018) Improving nanopore read accuracy with the R2C2 method enables the sequencing of highly multiplexed full-length single-­ cell cDNA.  Proc Natl Acad Sci U S A 115:9726–9731. https://doi.org/10.1073/ pnas.1806447115 83. Warren WC, Harris RA, Haukness M et  al (2020) Sequence diversity analyses of an improved rhesus macaque genome enhance its biomedical utility. Science 370. https://doi.org/10.1126/science.abc6617 84. Willingham AT, Gingeras TR (2006) TUF love for “junk” DNA.  Cell 125:1215–1220. https://doi. org/10.1016/j.cell.2006.06.009 85. Workman RE, Tang AD, Tang PS et  al (2019) Nanopore native RNA sequencing of a human poly(A) transcriptome. Nat Methods 16:1297–1305. https://doi.org/10.1038/s41592-­019-­0617-­2 86. Yamada K, Lim J, Dale JM et  al (2003) Empirical analysis of transcriptional activity in the Arabidopsis genome. Science 302:842–846. https://doi. org/10.1126/science.1088305 87. Yelin R, Dahary D, Sorek R et al (2003) Widespread occurrence of antisense transcription in the human genome. Nat Biotechnol 21:379–386. https://doi. org/10.1038/nbt808 88. Yip KY, Cheng C, Bhardwaj N et  al (2012) Classification of human genomic regions based on experimentally determined binding sites of more than 100 transcription-related factors. Genome Biol 13:R48. https://doi.org/10.1186/gb-­2012-­13-­9-­r48 89. Yu S, Cao H, Shen B, Feng J (2015) Tumor-derived exosomes in cancer progression and treatment failure. Oncotarget 6:37151–37168. https://doi.org/10.18632/ oncotarget.6022 90. Zheng GXY, Terry JM, Belgrader P et  al (2017) Massively parallel digital transcriptional profiling of single cells. Nat Commun 8:14049. https://doi. org/10.1038/ncomms14049

3

Towards Molecular Mechanism in Long Non-coding RNAs: Linking Structure and Function Karissa Sanbonmatsu

Abstract

3.1

While long non-coding RNAs play key roles in disease and development, few structural studies have been performed to date for this emerging class of RNAs. Here, we provide a brief review of functional studies of long non-­ coding RNAs, followed by a review of previous structural studies of long non-coding RNAs. We then describe structural studies of other classes of RNAs using chemical probing, nuclear magnetic resonance, small angle X-ray scattering, X-ray crystallography and cryogenic electron microscopy (cryo-EM). Next, we describe the way forward for the structural biology of long non-coding RNAs in terms of cryo-EM.  Finally, we discuss of the roles of long non-coding RNAs in the cell and how structure-function relationships might be used to elucidate further understanding.

Long non-coding RNAs (lncRNAs) have been shown to play important roles in development, epigenetics, stem cell biology, plant biology, RNA processing, hormone response, cancer and brain function [1–17]. Structure-function relationships for these RNAs, however, are not well understood. These RNAs are typically found in mammalian epigenetic systems, exceed 200 nucleotides in length, polyadenylated, alternatively spliced, low in abundance, and display relatively low sequence conservation. Preceded by the identification of non-coding RNAs in general [18, 19], long non-coding RNAs have been shown to have specificity to tissue type and developmental stage [1, 20, 21]. Many genome wide studies have been performed to identify large classes of lncRNAs associated with environmental changes, tissues, and diseases [1]. Loss of function studies have been performed to characterize functional roles of lncRNAs [22]. Biochemical and very low-resolution methods have been used to obtain structural information yielding glimpses of lncRNA structure [23]. High resolution structural biology techniques have been instrumental in determining structure-­ function relationships in other classes of RNA (riboswitches, ribozymes, and ribosomes) [24, 25]. These structure-function relationships enable more precise understanding of mechanism in terms of structural dynamics, thermodynamics,

Keywords

RNA · Long non-coding RNA · Non-coding RNA · RNA structure · RNA biochemistry

K. Sanbonmatsu (*) Los Alamos National Laboratory, Los Alamos, NM, USA e-mail: [email protected]

Introduction

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_3

23

K. Sanbonmatsu

24

kinetics, and Mg2+ effects. Yet, few studies have examined lncRNA mechanism at the atomistic level of detail [23]. Here, we review high resolution structure-function relationships in other RNA systems, describe the state-of-the-art in lncRNA structural and functional studies, and discuss the prospects for higher resolution structure-­function studies in lncRNAs.

3.2

Structure-Function Relationships

Structure-function relationships have been critical in understanding biology at the molecular level of detail. Since the inception of structural biology, 3-D structures of proteins have led to breakthroughs in understanding protein binding, protein complex formation, ligand binding, and self-assembly, all of which are important throughout biology and biomedicine.

3.2.1 Structural Studies of Proteins Often the structure of a protein itself can provide a roadmap of its function. The importance and utility of protein structures spurred development of a variety of structural techniques, culminating in X-ray crystallography and cryo-EM, producing high resolution 3-D structures of a wide range of protein complexes.

3.2.2 S  tructural Studies of RNA Systems As far fewer RNA systems have been studied relative to protein systems, RNA structural biology has considerably lagged behind protein structural biology. However, high resolution structures have been obtained for several classes of RNAs, leading to enormous insights into their structure-function relationships. Self Splicing Introns  Some of the earliest RNA-only systems solved to high resolution are the group I and group II introns [26]. Using X-ray

crystallography, these structures revealed the overall 3-D architecture of the RNA, detailed local RNA-RNA interaction motifs connecting the RNA together, the role of Mg2+ ions in the structure, and how the 2-D secondary structure maps translate into 3-D structures. Importantly, the 3-D structures were critical in determining the mechanism of catalysis for splicing, answering questions that were difficult or impossible to solve using other methods. Riboswitch RNAs  Riboswitch RNAs are regulatory stretches of RNA residing in the 5′-UTR of mRNA in bacterial metabolism-related genes [27]. These RNAs control gene expression by detecting environmental molecules through ligand-binding 3-D folds that alter the regulatory behavior of the RNA.  In a riboswitch, one sequence has two competing secondary structures (and two competing tertiary structures). The presence of ligand shifts the equilibrium to one structure, altering the gene expression ON/OFF state. The majority of riboswitches were discovered with in vitro chemical probing studies revealing the ligand dependence of the secondary structure, supported by in vivo functional studies. These in vitro secondary structures were later validated by in vitro high resolution X-ray crystallographic 3-D structures [28]. The dynamics of these systems have been studied using small angle X-ray scattering (SAXS) experiments and molecular dynamics simulations [29]. SAXS and biochemical studies have also revealed that ligand-free conformations tend to be extended and flexible, whereas ligand-bound conformations tend to be compact and ordered. Ribonucleoprotein Complexes  Structural studies of several ribonucleosome complexes have been studied, including the ribosome, RNA processing complexes, and the spliceosome. The ribosome is perhaps the most extensively studied ribonucleoprotein complex [30]. Structural studies have been attempted since the 1980s, commencing with biochemical studies to determine the secondary structure of the small subunit ribosome RNA (16S) and large subunit ribosomal RNA (23S). Neutron scattering enabled the rough

3  Towards Molecular Mechanism in Long Non-coding RNAs: Linking Structure and Function

placement of proteins in 3-D space relative to the ribosome complex. Early cryo-EM studies yielded the morphologies of the two subunits, the tRNA and mRNA ligands, the ribosomal proteins, and various conformations of the ribosome. Details were filled in with X-ray crystallography structures. High resolution cryo-EM enabled studies of ribosomes in a wide variety of functional states, for a variety of different species. With structures in hand, structural dynamics studies have been performed, integrating cryoEM, single molecule FRET, and large-scale ­ molecular dynamics simulations, providing an comprehensive picture of the molecular mechanism of the ribosome, characterizing the energy landscape and transition rates in the context of the detailed structures of beginning, ending and a plethora of intermediate states for various stages of protein synthesis [31, 32]. RNA Processing  Passmore and co-workers performed cryo-EM studies to obtain high resolution 3-D reconstructions of important ribonucleoprotein complexes that play key roles in mRNA cleavage [33, 34]. Spliceosome  Like the ribosome, the spliceosome has a rich history in mechanism and structural studies [35–38]. Beginning with functional studies, splicing was determined. Secondary structures identified different secondary structure configurations involved. X-ray crystallography studies solved structures of important pieces of complexes. High resolution cryo-EM recently produced several states of the intact spliceosome.

3.3

Studies of Long Non-coding RNAs

Loss of function studies have determined important lncRNAs in terms of functional roles in the cell, including epigenetic sensing and recruitment, sponging, P-bodies, Scaffolding, and RNA processing (lncRNAbnb1/2) and hormone response. Knock down studies also improve understanding. Knock downs of Braveheart

25

showed that this lncRNA is critical for lineage commitment in cardiomyocytes [2]. CRISPR/ Cas9 knock out studies have expanding the number of clear causal roles of lncRNAs. CRISPR/ Cas9 knock out of an 11-nucleotide r-turn RNA motif showed that this structural motif is critical for the overall function of Braveheart [39]. Knockouts had a major reduction in embryoid body beating assays, along with dramatic decreases in normal development. Protein binding studies offer some insight into mechanism. In pull downs and SAXS analysis, Braveheart was shown to bind zinc finger protein CNBP [40]. Several genome wide studies have been performed to identify proteins that bind to Xist [41].

3.3.1 Mechanisms of lncRNAs One of the earliest discovered lncRNAs is Xist (X chromosome inactivation stimulated transcript), responsible for inactivation of the X chromosome during development [42]. More recently, several lncRNAs have been associated with HOX gene systems during development [1]. The 1/2sbs-lncRNA controls mRNA decay by hybridizing with mRNA to form a platform for STAU1 protein binding, triggering degradation of mRNA [6]. Other lncRNAs are required for p21 activation [43], stem cell reprogramming [44] and stress response [45]. LncRNAs with Phenotypes  Although the physiological relevance of many of the reported lncRNAs has not been determined, many lncRNAs have been shown to possess important, visible phenotypes [46]. In addition to Xist, required for dosage compensation, the Braveheart lncRNA has been shown to be required for lineage commitment in cardiomyocytes [2]. FENDRR lncRNA is required for heart, lung and gastrointestinal development [47]. Linc-brn1b is required for neocortex development [47]. The COOLAIR lncRNA is required in A. thaliana for cold-timed flowering [4]. Additionally, the NEAT1 lncRNA has the clear phenotype of being critical for paraspeckle formation [48–50].

26

LncRNA-Protein Interactions  Many studies have been performed to determine protein partners and elucidate their functions [12, 41].

3.3.2 2  -D Structural Studies of lncRNAs: LncRNA Secondary Structure Studies Using Chemical Probing Genome-wide studies of secondary structure have revealed that lncRNAs are more structured than mRNAs, but less structured than ribosomal RNAs [51–57]. Detailed secondary structure studies of complete, intact lncRNA systems show that some lncRNAs are hierarchically structured with sub-domains containing modular RNA secondary structure motifs [58–60]. Studies of Malat-1 and related lncRNAs show that the 3′-end forms a triple helix, protecting it from RNase degradation [14, 61, 62]. Other studies have elucidated lncRNA-protein interactions, emphasizing the need for detailed structural studies and mechanistic studies at the molecular and atomistic level [63, 64]. LncRNAs tend to have low sequence identity and are often described as non-conserved. Some of the most well-studied non-coding RNAs ­(miRNAs and rRNAs) have very high sequence identity (>78% in nucleic acid sequence identity) [65]. In contrast, many other important classes of non-coding RNAs have relatively low sequence identity (nucleic acid sequence identity of ~50%–65%), but secondary structures that are conserved across thousands of sequences. For example, riboswitches, which regulate metabolism in bacteria, typically have sequence identities of only 50%–65%, but have secondary structures conserved across thousands of species [65]. The U2 and U4 spliceosomal RNAs have sequence identities 9000 sequences. The 5S ribosomal RNA has sequence identity of ~60% but secondary structure conserved over 229,000 sequences. The group I intron has decidedly low sequence identity (~36%) but structure conserved across 60,000 species [65].

K. Sanbonmatsu

RNAs with low sequence identity are difficult to find using conventional search algorithms such as BLAST.  However, knowledge of secondary structure dramatically enhances the search success. In the case of riboswitches, the RNA secondary structure was determined for a single species using in vitro chemical probing of the RNA in cell-free reconstituted systems [66–73]. Next, this structure was used as a fingerprint to find the structure in thousands of other species, despite the low sequence identity [74]. The secondary structures determined from cell-free systems by chemical probing were verified by X-ray crystallography [75–79]. To determine the RNA secondary structure of lncRNA molecules, strategies similar to those used to determine the original 16S rRNA secondary structure [80–82] and the riboswitches [83] have been employed. Chemical probing experiments determine nucleotides that are highly mobile and likely to reside in looping regions, as well as those nucleotides with low mobility, likely to participate in Watson-Crick base pairs. To cope with the large RNA size, 3S (Shot-Gun Secondary Structure) can be used, which probes the entire RNA first and then probes shorter segments of the RNA in successive rounds of probing [58, 84]. By matching signals of short segments with full RNA experiments, modular sub-domains are identified, for which a secondary structure is often readily discernable. The resulting secondary structure can be used to improve existing phylogenetic sequence alignments, and, in principle, can be used to find instances of the lncRNA not previously found in other species. An interesting case is the 873 nt steroid receptor RNA activator lncRNA in humans (SRA-1). This lncRNA co-activates the hormone response in human T-47D cells and co-immunoprecipitates with a large number of important proteins, including several hormone receptors (estrogen receptor, progesterone receptor, androgen receptor, glucocorticoid receptor and thyroid receptor) [85–88]. Binding assays in in vitro cell-free reconstituted systems have shown strong binding to the pseudouridinylase Pus1p, estrogen receptor, thyroid receptor, the sex reversal factor DAX-1, and the

3  Towards Molecular Mechanism in Long Non-coding RNAs: Linking Structure and Function

epigenetic factor SHARP.  While the primary function of SRA-1 is to co-activate the hormone response, a speculated secondary function involving the binding of SRA-1 to its cognate protein SRAP has recently been shown not to occur (SRA-1 does not bind to SRAP) [89]. A previous study demonstrated that SRA-1 contains four modular secondary structure sub-­ domains, each containing multiple secondary structure motifs. The secondary structure was consistent with four different probing techniques (SHAPE, DMS, in-line, and RNase V1). Binding studies have shown that SHARP binds to the helix 12 / helix 13 (H12/13) domain [90]. In vitro studies establish the ab initio structure because the probing signal in vivo may to be obfuscated by multiple proteins binding to the RNA [11, 12]. In addition, there are few known cases where an in vitro structure of an intact, individual RNA has been shown to differ from its corresponding in vivo structure. For example, the vast majority of crystallographic structures of RNAs, which are of course determined in vitro, have either (i) been validated in vivo, or (ii) not been disproven in vivo. In the case of riboswitch RNAs, crystallographic data strongly support initial secondary structures determined by chemical probing techniques discussed above. Determination of the precise and detailed secondary structure of lncRNAs allows classification into (i) highly structured RNAs with sub-domains and complex structural motifs, such as multiway junctions, (ii) loosely structured RNAs with multiple stem-loops, but lacking hierarchical domain structure and complex motifs, and (iii) unstructured, disordered RNAs, which lack secondary structure.

3.3.3 3  -D Studies of Long Non-­ coding RNAs at Low Resolution Studies of Tertiary Interactions in Long Non-­ coding RNAs  Pyle and co-workers used UV crosslinking to identify individual tertiary interactions in lncRNA systems [91].

27

Small Angle X-Ray Scattering (SAXS)  Small angle X-ray scattering studies have been used to characterize the 3-D structure of RNA systems that are too flexible to be studied with X-ray crystallography. Often, RNA molecules sample a multitude of conformations. SAXS can characterize the distribution of configurations samples. In addition, SAXS can be a first step towards higher resolution structure determination as the requirements for sample preparation are much less stringent than for X-ray crystallography or for higher resolution cryo-EM.  Recently, low resolution structures of the Braveheart lncRNA and Braveheart-CNBP ribonucleoprotein complex were determined using SAXS [40]. The structures were consistent with 2-D secondary structures determined via chemical probing, with secondary structure domains fairly well-­separated in 3-D physical space. The molecule was found to be somewhat flexible, where multiple all-atom 3-D configurations were consistent with 3-D volume reconstructions consistent with the SAXS data. However, the SAXS data demonstrated compaction upon Mg2+ titration, which is clear evidence of well-defined tertiary structures in the RNA system. This is similar to riboswitch systems, which still sample well-defined 3-D structures, even in their ligand-free states, known to be extended and flexible. Additionally, Braveheart underwent significant reorganization upon protein binding, as evidenced by the substantial change in scattering profiles and corresponding 3-D volume reconstructions as a result of CNBP binding. Atomic Force Microscopy (AFM) Studies of lncRNAs  AFM has been used to characterize the 3-D structure of lncRNA systems without solution. In these freeze-dried experiments, MEG3 displayed tertiary structure consistent with 2-D secondary structures determined by chemical probing [92]. Fluorescence Correlation Spectroscopy (FCS)  FCS has been used to characterize the size, in terms of extended vs. compact, of lncRNAs systems in 3-D.  In one FCS study, lncRNAs (e.g., HOTAIR) were found to be more

K. Sanbonmatsu

28

compact than mRNA transcripts, but less compact than ribosomes [93].

3.3.4 3-D Structural Techniques Used to Study Other Classes of RNAs High resolution techniques have been used to determine structures for a number of other classes of RNA systems, such as riboswitches, ribozymes, introns, ribosomes and spliceosomes. In terms of techniques, nuclear magnetic resonance imaging (NMR) can be used to study quite small systems. This method has the advantage of capturing precise information about the dynamics of the RNA, multiple configurations, and rates of transition between configurations NMR has been used to obtain such information for a variety of riboswitches and regions of viral RNAs, as well as a small region of Xist RepA lncRNA [94]. X-ray crystallography is a traditional form of high-resolution structure determination used for small and medium-sized RNA systems. High resolution structures have been determined for riboswitches, ribozymes, introns and ribosomes. Cryogenic electron microscopy (cryo-EM) can be used to determined high resolution structures for medium-sized and large-sized protein systems and ribonucleoprotein systems. To date, this method has determined a wide variety of structures for ribonucleoprotein complexes, including many ribosome complexes and several spliceosome complexes. Quite recently, the method has been used to obtain medium resolution structures of several RNA-only systems, including riboswitches and regions of viral RNAs [95].

3.3.5 Expansion of Structural Tools to Study Long Noncoding RNAs at High Resolution High resolution structural studies of lncRNA systems will undoubtedly reveal new information about their mechanisms. As early studies present evidence for tertiary contacts, at minimum, cryo­EM studies will likely reveal structured tertiary

motifs surrounded by flexible regions or large swaths of RNA. At the other extreme, these studies may uncover highly structured ribonucleoprotein complexes, or even structured RNA-only systems. The past decade of lncRNA research has clearly shown that lncRNAs represent a highly diverse class of RNAs with a wide range of mechanisms. Thus, a wide range of structure may be observed, ranging from highly dynamic to highly structured. Higher resolution structural studies will be able to shed light on structure-­ function relationships, in terms of specific protein binding partners, RNA binding partners, DNA binding partners, conformational changes, and roles in pathways. These studies may also offer insight into the evolution of lncRNAs. Since lncRNAs often have fairly low sequence identify, structural-function studies will enable analysis of conservation in terms more general measures, such as 2-D structure, 3-D structural RNA motifs, 3-D RNA-protein binding motifs, RNA dynamics and RNA function. Acknowledgements  The author acknowledges generous support by the LANL LDRD program.

References 1. Rinn JL, Chang HY (2012) Genome regulation by long noncoding RNAs. Annu Rev Biochem 81:145–166 2. Klattenhoff CA, Scheuermann JC, Surface LE, Bradley RK, Fields PA, Steinhauser ML, Ding H, Butty VL, Torrey L, Haas S, Abo R, Tabebordbar M, Lee RT, Burge CB, Boyer LA (2013) Braveheart, a long noncoding RNA required for cardiovascular lineage commitment. Cell 152:570–583 3. Mercer TR, Mattick JS (2013) Structure and function of long noncoding RNAs in epigenetic regulation. Nat Struct Mol Biol 20:300–307 4. Swiezewski S, Liu F, Magusin A, Dean C (2009) Cold-­ induced silencing by long antisense transcripts of an Arabidopsis Polycomb target. Nature 462:799–802 5. Ulitsky I, Bartel DP (2013) lincRNAs: genomics, evolution, and mechanisms. Cell 154:26–46 6. Gong C, Maquat LE (2011) lncRNAs transactivate STAU1-mediated mRNA decay by duplexing with 3' UTRs via Alu elements. Nature 470:284–288 7. Kaneko S, Bonasio R, Saldana-Meyer R, Yoshida T, Son J, Nishino K, Umezawa A, Reinberg D (2014) Interactions between JARID2 and noncoding RNAs regulate PRC2 recruitment to chromatin. Mol Cell 53:290–300

3  Towards Molecular Mechanism in Long Non-coding RNAs: Linking Structure and Function 8. Heard E, Mongelard F, Arnaud D, Chureau C, Vourc'h C, Avner P (1999) Human XIST yeast artificial chromosome transgenes show partial X inactivation center function in mouse embryonic stem cells. Proc Natl Acad Sci U S A 96:6841–6846 9. da Rocha ST, Boeva V, Escamilla-Del-Arenal M, Ancelin K, Granier C, Matias NR, Sanulli S, Chow J, Schulz E, Picard C, Kaneko S, Helin K, Reinberg D, Stewart AF, Wutz A, Margueron R, Heard E (2014) Jarid2 is implicated in the initial Xist-induced targeting of PRC2 to the inactive X chromosome. Mol Cell 53:301–316 10. Boumil RM, Lee JT (2001) Forty years of decoding the silence in X-chromosome inactivation. Hum Mol Genet 10:2225–2232 11. Davidovich C, Zheng L, Goodrich KJ, Cech TR (2013) Promiscuous RNA binding by Polycomb repressive complex 2. Nat Struct Mol Biol 20:1250–1257 12. Davidovich C, Wang X, Cifuentes-Rojas C, Goodrich KJ, Gooding AR, Lee JT, Cech TR (2015) Toward a consensus on the binding specificity and promiscuity of PRC2 for RNA. Mol Cell 57:552–558 13. Cech TR, Steitz JA (2014) The noncoding RNA revolution-trashing old rules to forge new ones. Cell 157:77–94 14. Brown JA, Bulkley D, Wang J, Valenstein ML, Yario TA, Steitz TA, Steitz JA (2014) Structural insights into the stabilization of MALAT1 noncoding RNA by a bipartite triple helix. Nat Struct Mol Biol 21:633–640 15. Dharap A, Nakka VP, Vemuganti R (2012) Effect of focal ischemia on long noncoding RNAs. Stroke 43:2800–2802 16. Ponting CP, Oliver PL, Reik W (2009) Evolution and functions of long noncoding RNAs. Cell 136:629–641 17. Derrien T, Johnson R, Bussotti G, Tanzer A, Djebali S, Tilgner H, Guernec G, Martin D, Merkel A, Knowles DG, Lagarde J, Veeravalli L, Ruan X, Ruan Y, Lassmann T, Carninci P, Brown JB, Lipovich L, Gonzalez JM, Thomas M, Davis CA, Shiekhattar R, Gingeras TR, Hubbard TJ, Notredame C, Harrow J, Guigo R (2012) The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Res 22:1775–1789 18. Numata K, Kanai A, Saito R, Kondo S, Adachi J, Wilming LG, Hume DA, Hayashizaki Y, Tomita M, R.G.  Group, G.S.L.  Members (2003) Identification of putative noncoding RNAs among the RIKEN mouse full-length cDNA collection. Genome Res 13:1301–1306 19. Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, Maeda N, Oyama R, Ravasi T, Lenhard B, Wells C, Kodzius R, Shimokawa K, Bajic VB, Brenner SE, Batalov S, Forrest AR, Zavolan M, Davis MJ, Wilming LG, Aidinis V, Allen JE, Ambesi-­ Impiombato A, Apweiler R, Aturaliya RN, Bailey TL, Bansal M, Baxter L, Beisel KW, Bersano T, Bono H, Chalk AM, Chiu KP, Choudhary V, Christoffels A, Clutterbuck DR, Crowe ML, Dalla E, Dalrymple BP, de Bono B,

29

Gatta GD, di Bernardo D, Down T, Engstrom P, Fagiolini M, Faulkner G, Fletcher CF, Fukushima T, Furuno M, Futaki S, Gariboldi M, GeorgiiHemming P, Gingeras TR, Gojobori T, Green RE, Gustincich S, Harbers M, Hayashi Y, Hensch TK, Hirokawa N, Hill D, Huminiecki L, Iacono M, Ikeo K, Iwama A, Ishikawa T, Jakt M, Kanapin A, Katoh M, Kawasawa Y, Kelso J, Kitamura H, Kitano H, Kollias G, Krishnan SP, Kruger A, Kummerfeld SK, Kurochkin IV, Lareau LF, Lazarevic D, Lipovich L, Liu J, Liuni S, McWilliam S, Madan Babu M, Madera M, Marchionni L, Matsuda H, Matsuzawa S, Miki H, Mignone F, Miyake S, Morris K, Mottagui-Tabar S, Mulder N, Nakano N, Nakauchi H, Ng P, Nilsson R, Nishiguchi S, Nishikawa S, Nori F, Ohara O, Okazaki Y, Orlando V, Pang KC, Pavan WJ, Pavesi G, Pesole G, Petrovsky N, Piazza S, Reed J, Reid JF, Ring BZ, Ringwald M, Rost B, Ruan Y, Salzberg SL, Sandelin A, Schneider C, Schonbach C, Sekiguchi K, Semple CA, Seno S, Sessa L, Sheng Y, Shibata Y, Shimada H, Shimada K, Silva D, Sinclair B, Sperling S, Stupka E, Sugiura K, Sultana R, Takenaka Y, Taki K, Tammoja K, Tan SL, Tang S, Taylor MS, Tegner J, Teichmann SA, Ueda HR, van Nimwegen E, Verardo R, Wei CL, Yagi K, Yamanishi H, Zabarovsky E, Zhu S, Zimmer A, Hide W, Bult C, Grimmond SM, Teasdale RD, Liu ET, Brusic V, Quackenbush J, Wahlestedt C, Mattick JS, Hume DA, Kai C, Sasaki D, Tomaru Y, Fukuda S, Kanamori-Katayama M, Suzuki M, Aoki J, Arakawa T, Iida J, Imamura K, Itoh M, Kato T, Kawaji H, Kawagashira N, Kawashima T, Kojima M, Kondo S, Konno H, Nakano K, Ninomiya N, Nishio T, Okada M, Plessy C, Shibata K, Shiraki T, Suzuki S, Tagami M, Waki K, Watahiki A, Okamura-Oho Y, Suzuki H, Kawai J, Hayashizaki Y, F.  Consortium, R.G.E.R.  Group, G.  Genome Science (2005) The transcriptional landscape of the mammalian genome. Science 309:1559–1563 20. Ponjavic J, Ponting CP, Lunter G (2007) Functionality or transcriptional noise? Evidence for selection within long noncoding RNAs. Genome Res 17:556–565 21. Dinger ME, Amaral PP, Mercer TR, Pang KC, Bruce SJ, Gardiner BB, Askarian-Amiri ME, Ru K, Solda G, Simons C, Sunkin SM, Crowe ML, Grimmond SM, Perkins AC, Mattick JS (2008) Long noncoding RNAs in mouse embryonic stem cell pluripotency and differentiation. Genome Res 18:1433–1445 22. Charles Richard JL, Eichhorn PJA (2018) Platforms for investigating LncRNA functions. SLAS Technol 23:493–506 23. Novikova IV, Hennelly SP, Tung CS, Sanbonmatsu KY (2013) Rise of the RNA machines: exploring the structure of long non-coding RNAs. J Mol Biol 425:3731–3746 24. Westhof E (2015) Twenty years of RNA crystallography. RNA 21:486–487 25. Reyes FE, Garst AD, Batey RT (2009) Strategies in RNA crystallography. Methods Enzymol 469:119–139

30 26. Pyle AM (2016) Group II intron self-splicing. Annu Rev Biophys 45:183–205 27. Breaker RR (2011) Prospects for riboswitch discovery and analysis. Mol Cell 43:867–879 28. Serganov A, Patel DJ (2012) Metabolite recognition principles and molecular mechanisms underlying riboswitch function. Annu Rev Biophys 41:343–370 29. Zhang J, Jones CP, Ferre-D'Amare AR (2014) Global analysis of riboswitches by small-angle X-ray scattering and calorimetry. Biochim Biophys Acta 1839:1020–1029 30. Jobe A, Liu Z, Gutierrez-Vargas C, Frank J (2019) New insights into ribosome structure and function. Cold Spring Harb Perspect Biol 11:1–17 31. Sanbonmatsu KY (2012) Computational studies of molecular machines: the ribosome. Curr Opin Struct Biol 22:168–174 32. Sanbonmatsu KY (2019) Large-scale simulations of nucleoprotein complexes: ribosomes, nucleosomes, chromatin, chromosomes and CRISPR.  Curr Opin Struct Biol 55:104–113 33. Kumar A, Clerici M, Muckenfuss LM, Passmore LA, Jinek M (2019) Mechanistic insights into mRNA 3′-end processing. Curr Opin Struct Biol 59:143–150 34. Casanal A, Shakeel S, Passmore LA (2019) Interpretation of medium resolution cryoEM maps of multi-protein complexes. Curr Opin Struct Biol 58:166–174 35. Fica SM, Nagai K (2017) Cryo-electron microscopy snapshots of the spliceosome: structural insights into a dynamic ribonucleoprotein machine. Nat Struct Mol Biol 24:791–799 36. Fica SM (2020) Cryo-EM snapshots of the human spliceosome reveal structural adaptions for splicing regulation. Curr Opin Struct Biol 65:139–148 37. Wilkinson ME, Charenton C, Nagai K (2020) RNA splicing by the spliceosome. Annu Rev Biochem 89:359–388 38. Smathers CM, Robart AR (2019) The mechanism of splicing as told by group II introns: ancestors of the spliceosome. Biochim Biophys Acta Gene Regul Mech 1862:194390 39. Xue Z, Hennelly S, Doyle B, Gulati AA, Novikova IV, Sanbonmatsu KY, Boyer LA (2016) A G-rich motif in the lncRNA Braveheart interacts with a zinc-finger transcription factor to specify the cardiovascular lineage. Mol Cell 64:37–50 40. Kim DN, Thiel BC, Mrozowich T, Hennelly SP, Hofacker IL, Patel TR, Sanbonmatsu KY (2020) Zinc-finger protein CNBP alters the 3-D structure of lncRNA Braveheart in solution. Nat Commun 11:148 41. Minajigi A, Froberg JE, Wei C, Sunwoo H, Kesner B, Colognori D, Lessing D, Payer B, Boukhali M, Haas W, Lee JT (2015) A comprehensive Xist interactome reveals cohesin repulsion and an RNA-directed chromosome conformation. Science (New York, NY) 349:282–294

K. Sanbonmatsu 42. Lee JT, Jaenisch R (1997) The (epi)genetic control of mammalian X-chromosome inactivation. Curr Opin Genet Dev 7:274–280 43. Huarte M, Guttman M, Feldser D, Garber M, Koziol MJ, Kenzelmann-Broz D, Khalil AM, Zuk O, Amit I, Rabani M, Attardi LD, Regev A, Lander ES, Jacks T, Rinn JL (2010) A large intergenic noncoding RNA induced by p53 mediates global gene repression in the p53 response. Cell 142:409–419 44. Guttman M, Donaghey J, Carey BW, Garber M, Grenier JK, Munson G, Young G, Lucas AB, Ach R, Bruhn L, Yang X, Amit I, Meissner A, Regev A, Rinn JL, Root DE, Lander ES (2011) lincRNAs act in the circuitry controlling pluripotency and differentiation. Nature 477:295–300 45. Kino T, Hurt DE, Ichijo T, Nader N, Chrousos GP (2010) Noncoding RNA gas5 is a growth arrest- and starvation-associated repressor of the glucocorticoid receptor. Sci Signal 3:ra8 46. Li L, Chang HY (2014) Physiological roles of long noncoding RNAs: insight from knockout mice. Trends Cell Biol 24:594–602 47. Sauvageau M, Goff LA, Lodato S, Bonev B, Groff AF, Gerhardinger C, Sanchez-Gomez DB, Hacisuleyman E, Li E, Spence M, Liapis SC, Mallard W, Morse M, Swerdel MR, D’Ecclessis MF, Moore JC, Lai V, Gong G, Yancopoulos GD, Frendewey D, Kellis M, Hart RP, Valenzuela DM, Arlotta P, Rinn JL (2013) Multiple knockout mouse models reveal lincRNAs are required for life and brain development. elife 2:e01749 48. Naganuma T, Nakagawa S, Tanigawa A, Sasaki YF, Goshima N, Hirose T (2012) Alternative 3′-end processing of long noncoding RNA initiates construction of nuclear paraspeckles. EMBO J 31:4020–4034 49. Nakagawa S, Hirose T (2012) Paraspeckle nuclear bodies–useful uselessness? Cell Mol Life Sci 69:3027–3036 50. Sasaki YT, Ideue T, Sano M, Mituyama T, Hirose T (2009) MENepsilon/beta noncoding RNAs are essential for structural integrity of nuclear paraspeckles. Proc Natl Acad Sci U S A 106:2525–2530 51. Wan Y, Qu K, Zhang QC, Flynn RA, Manor O, Ouyang Z, Zhang J, Spitale RC, Snyder MP, Segal E, Chang HY (2014) Landscape and variation of RNA secondary structure across the human transcriptome. Nature 505:706–709 52. Wan Y, Qu K, Ouyang Z, Chang HY (2013) Genome-­ wide mapping of RNA structure using nuclease digestion and high-throughput sequencing. Nat Protoc 8:849–869 53. Ouyang Z, Snyder MP, Chang HY (2013) SeqFold: genome-scale reconstruction of RNA secondary structure integrating high-throughput sequencing data. Genome Res 23:377–387 54. Wan Y, Qu K, Ouyang Z, Kertesz M, Li J, Tibshirani R, Makino DL, Nutter RC, Segal E, Chang HY (2012) Genome-wide measurement of RNA folding energies. Mol Cell 48:169–181

3  Towards Molecular Mechanism in Long Non-coding RNAs: Linking Structure and Function 55. Kertesz M, Wan Y, Mazor E, Rinn JL, Nutter RC, Chang HY, Segal E (2010) Genome-wide measurement of RNA secondary structure in yeast. Nature 467:103–107 56. Ding Y, Tang Y, Kwok CK, Zhang Y, Bevilacqua PC, Assmann SM (2014) In vivo genome-wide profiling of RNA secondary structure reveals novel regulatory features. Nature 505:696–700 57. Rouskin S, Zubradt M, Washietl S, Kellis M, Weissman JS (2014) Genome-wide probing of RNA structure reveals active unfolding of mRNA structures in vivo. Nature 505:701–705 58. Novikova IV, Hennelly SP, Sanbonmatsu KY (2012) Structural architecture of the human long non-coding RNA, steroid receptor RNA activator. Nucleic Acids Res 40:5034–5051 59. Ilik IA, Quinn JJ, Georgiev P, Tavares-Cadete F, Maticzka D, Toscano S, Wan Y, Spitale RC, Luscombe N, Backofen R, Chang HY, Akhtar A (2013) Tandem stem-loops in roX RNAs act together to mediate X chromosome dosage compensation in Drosophila. Mol Cell 51:156–173 60. Somarowthu S, Legiewicz M, Chillon I, Marcia M, Liu F, Pyle AM (2015) HOTAIR forms an intricate and modular secondary structure. Mol Cell 58:353–361 61. Wilusz JE, JnBaptiste CK, Lu LY, Kuhn CD, Joshua-­ Tor L, Sharp PA (2012) A triple helix stabilizes the 3′ ends of long noncoding RNAs that lack poly(A) tails. Genes Dev 26:2392–2407 62. Wilusz JE, Freier SM, Spector DL (2008) 3′ end processing of a long nuclear-retained noncoding RNA yields a tRNA-like cytoplasmic RNA.  Cell 135:919–932 63. Chu C, Zhang QC, da Rocha ST, Flynn RA, Bharadwaj M, Calabrese JM, Magnuson T, Heard E, Chang HY (2015) Systematic discovery of Xist RNA binding proteins. Cell 161:404–416 64. Spitale RC, Flynn RA, Zhang QC, Crisalli P, Lee B, Jung JW, Kuchelmeister HY, Batista PJ, Torre EA, Kool ET, Chang HY (2015) Structural imprints in vivo decode RNA regulatory mechanisms. Nature 519:486–490 65. Griffiths-Jones S, Bateman A, Marshall M, Khanna A, Eddy SR (2003) Rfam: an RNA family database. Nucleic Acids Res 31:439–441 66. Regulski EE, Breaker RR (2008) In-line probing analysis of riboswitches. Methods Mol Biol 419:53–67 67. Winkler W, Nahvi A, Breaker RR (2002) Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression. Nature 419:952–956 68. Mandal M, Boese B, Barrick JE, Winkler WC, Breaker RR (2003) Riboswitches control fundamental biochemical pathways in Bacillus subtilis and other bacteria. Cell 113:577–586 69. Mandal M, Lee M, Barrick JE, Weinberg Z, Emilsson GM, Ruzzo WL, Breaker RR (2004) A glycine-­ dependent riboswitch that uses cooperative binding to control gene expression. Science 306:275–279

31

70. Winkler WC, Nahvi A, Roth A, Collins JA, Breaker RR (2004) Control of gene expression by a natural metabolite-responsive ribozyme. Nature 428:281–286 71. Sudarsan N, Hammond MC, Block KF, Welz R, Barrick JE, Roth A, Breaker RR (2006) Tandem riboswitch architectures exhibit complex gene control functions. Science 314:300–304 72. Cheah MT, Wachter A, Sudarsan N, Breaker RR (2007) Control of alternative RNA splicing and gene expression by eukaryotic riboswitches. Nature 447:497–500 73. Sudarsan N, Lee ER, Weinberg Z, Moy RH, Kim JN, Link KH, Breaker RR (2008) Riboswitches in eubacteria sense the second messenger cyclic di-­ GMP. Science 321:411–413 74. Weinberg Z, Barrick JE, Yao Z, Roth A, Kim JN, Gore J, Wang JX, Lee ER, Block KF, Sudarsan N, Neph S, Tompa M, Ruzzo WL, Breaker RR (2007) Identification of 22 candidate structured RNAs in bacteria using the CMfinder comparative genomics pipeline. Nucleic Acids Res 35:4809–4819 75. Batey RT, Gilbert SD, Montange RK (2004) Structure of a natural guanine-responsive riboswitch complexed with the metabolite hypoxanthine. Nature 432:411–415 76. Montange RK, Batey RT (2006) Structure of the S-adenosylmethionine riboswitch regulatory mRNA element. Nature 441:1172–1175 77. Gilbert SD, Rambo RP, Van Tyne D, Batey RT (2008) Structure of the SAM-II riboswitch bound to S-adenosylmethionine. Nat Struct Mol Biol 15:177–182 78. Stoddard CD, Montange RK, Hennelly SP, Rambo RP, Sanbonmatsu KY, Batey RT (2010) Free state conformational sampling of the SAM-I riboswitch aptamer domain. Structure 18:787–797 79. Montange RK, Batey RT (2008) Riboswitches: emerging themes in RNA structure and function. Annu Rev Biophys 37:117–133 80. Woese CR, Magrum LJ, Gupta R, Siegel RB, Stahl DA, Kop J, Crawford N, Brosius J, Gutell R, Hogan JJ, Noller HF (1980) Secondary structure model for bacterial 16S ribosomal RNA: phylogenetic, enzymatic and chemical evidence. Nucleic Acids Res 8:2275–2293 81. Noller HF, Woese CR (1981) Secondary structure of 16S ribosomal RNA. Science 212:403–411 82. Noller HF, Kop J, Wheaton V, Brosius J, Gutell RR, Kopylov AM, Dohme F, Herr W, Stahl DA, Gupta R, Waese CR (1981) Secondary structure model for 23S ribosomal RNA. Nucleic Acids Res 9:6167–6189 83. Winkler WC, Nahvi A, Sudarsan N, Barrick JE, Breaker RR (2003) An mRNA structure that controls gene expression by binding S-adenosylmethionine. Nat Struct Biol 10:701–707 84. Novikova IV, Hennelly SP, Sanbonmatsu KY (2013) 3S: shotgun secondary structure determination for long non-coding RNAs. Methods 63:170–177

32 85. Yao H, Brick K, Evrard Y, Xiao T, Camerini-Otero RD, Felsenfeld G (2010) Mediation of CTCF transcriptional insulation by DEAD-box RNA-binding protein p68 and steroid receptor RNA activator SRA. Genes Dev 24:2543–2555 86. Xu B, Yang WH, Gerin I, Hu CD, Hammer GD, Koenig RJ (2009) Dax-1 and steroid receptor RNA activator (SRA) function as transcriptional coactivators for steroidogenic factor 1 in steroidogenesis. Mol Cell Biol 29:1719–1734 87. Colley SM, Iyer KR, Leedman PJ (2008) The RNA coregulator SRA, its binding proteins and nuclear receptor signaling activity. IUBMB Life 60:159–164 88. Huet T, Miannay FA, Patton JR, Thore S (2014) Steroid receptor RNA activator (SRA) modification by the human pseudouridine synthase 1 (hPus1p): RNA binding, activity, and atomic model. PLoS One 9:e94610 89. McKay DB, Xi L, Barthel KK, Cech TR (2014) Structure and function of steroid receptor RNA activator protein, the proposed partner of SRA noncoding RNA. J Mol Biol 426:1766–1785 90. Arieti F, Gabus C, Tambalo M, Huet T, Round A, Thore S (2014) The crystal structure of the split end

K. Sanbonmatsu protein SHARP adds a new layer of complexity to proteins containing RNA recognition motifs. Nucleic Acids Res 42:6742–6752 91. Liu F, Somarowthu S, Pyle AM (2017) Visualizing the secondary and tertiary architectural domains of lncRNA RepA. Nat Chem Biol 13:282–289 92. Uroda T, Anastasakou E, Rossi A, Teulon JM, Pellequer JL, Annibale P, Pessey O, Inga A, Chillon I, Marcia M (2019) Conserved pseudoknots in lncRNA MEG3 are essential for stimulation of the p53 pathway. Mol Cell 75:982–995.e989 93. Borodavka A, Singaram SW, Stockley PG, Gelbart WM, Ben-Shaul A, Tuma R (2016) Sizes of long RNA molecules are determined by the branching patterns of their secondary structures. Biophys J 111:2077–2085 94. Duszczyk MM, Zanier K, Sattler M (2008) A NMR strategy to unambiguously distinguish nucleic acid hairpin and duplex conformations applied to a Xist RNA A-repeat. Nucleic Acids Res 36:7068–7077 95. Kappel K, Zhang K, Su Z, Watkins AM, Kladwang W, Li S, Pintilie G, Topkar VV, Rangan R, Zheludev IN, Yesselman JD, Chiu W, Das R (2020) Accelerated cryo-EM-guided determination of three-dimensional RNA-only structures. Nat Methods 17:699–707

Part III Long Noncoding RNA and Inflammation

4

Long Non-coding RNAs in Rheumatology Susanne N. Wijesinghe, Mark A. Lindsay, and Simon W. Jones

Abstract

The last decade has seen an enormous increase in long non-coding RNA (lncRNA) research within rheumatology. LncRNAs are arbitrarily classed as non-protein encoding RNA transcripts that exceed 200 nucleotides in length. These transcripts have tissue and cell specific patterns of expression and are implicated in a variety of biological processes. Unsurprisingly, numerous lncRNAs are dysregulated in rheumatoid conditions, correlating with disease activity and cited as potential biomarkers and targets for therapeutic intervention. In this chapter, following an introduction into each condition, we discuss the lncRNAs involved in rheumatoid arthritis, osteoarthritis and systemic lupus erythematosus. These inflammatory joint conditions share several inflammatory signalling pathways and therefore not surprisingly many commonly dysregulated lncRNAs are shared across these conditions. In the interest of translational S. N. Wijesinghe · S. W. Jones (*) Institute of Inflammation and Ageing, MRC Versus Arthritis Centre for Musculoskeletal Ageing Research, University of Birmingham, Birmingham, UK e-mail: [email protected] M. A. Lindsay Department of Pharmacy and Pharmacology, University of Bath, Bath, UK

research only those lncRNAs which are strongly conserved have been addressed. The lncRNAs discussed here have diverse roles in regulating inflammation, proliferation, migration, invasion and apoptosis. Understanding the molecular basis of lncRNA function in rheumatology will be crucial in fully determining the inflammatory mechanisms that drive these conditions. Keywords

Rheumatoid arthritis · Osteoarthritis · Systemic lupus erythematosus · Long noncoding RNA · Inflammation

4.1

Arthritic Diseases

4.1.1 Rheumatoid Arthritis Rheumatoid arthritis (RA) is a chronic inflammatory autoimmune condition resulting in progressive disability and premature death in older adults [1]. It is a lifelong condition mainly effecting the lining of the synovial joint causing pain, stiffness and swelling in and around the effected joints. Unfortunately, up to 1% of the world’s population suffer with this debilitating condition, for which there is no cure. Additionally, with a third of patients unable to work within 2 years of diagnosis, there is a substantial socioeconomic bur-

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_4

35

36

den. RA affects more women than men, with women having a 3.6% lifetime risk of developing RA compared to 1.7% in men [2]. Although the aetiology is not fully clear, a combination of genetic, environmental and lifestyle factors are all associated with RA. Aside from gender, additional RA risk factors include age with a peak disease onset in the 60s, obesity, diabetes, osteoporosis and smoking [3]. Following immune activation, inflammation of the synovial membrane (synovitis) is an initial characteristic presentation of RA. Synovial fibroblasts also termed fibroblast-like synoviocytes (FLS), within the synovial joint membrane, become dysfunctional and hyperplastic forming the pannus. The synovial joint is infiltrated with leukocytes, which interact with FLS inundating the synovial fluid with pro-inflammatory factors [1]. Cells of both the innate and adaptive immune system are thought to be central in RA pathogenesis. Monocytes and macrophages are commonly found to infiltrate the synovium with a polarisation towards the pro-inflammatory (M1) versus anti-inflammatory (M2) macrophage [4]. These cells contribute to a sustained chronic inflammatory state within the joint by releasing pro-­ inflammatory cytokines, such as tumour necrosis factor alpha (TNFα) and interleukin 6 (IL-6) [5]. The pro-inflammatory microenvironment within the synovial joint results in cartilage degradation and bone loss. Synovial hyperplasia causes elevated matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs), which drive joint destruction [1]. Proteoglycans and extracellular matrix (ECM) binding soluble factors are released from damaged cartilage further activating FLS and resulting in a tumour like transformation [6]. These activated FLS express matrix-degrading enzymes such as MMPs, ADAMTs and cathepsin, and activate signalling pathways that regulate growth and apoptosis [6]. Activated FLS together with pro-inflammatory cytokines with pro-osteogenic effects facilitate the differentiation of infiltrating macrophages into osteoclasts, which result in inflammatory cysts, bone resorption, erosion and loss [1, 7].

S. N. Wijesinghe et al.

Synovitis, cartilage damage and bone loss are all detected by radiographs, ultrasonography and magnetic resonance imagining (MRI) [7]. Another early inflammatory marker detected by MRI is seen in the subchondral bone marrow. Like synovitis, the bone marrow is infiltrated by a host of immune cells including macrophages, T lymphocytes, B lymphocytes and osteoclasts [8]. The resulting inflammation is detected by MRI, presenting as bone marrow edema (BME). BME is correlated with disease severity and joint destruction and may develop independently of synovitis. As such, detection of BME in MRIs has 100% accuracy in predicting rapid RA onset [7, 8]. The first joints to be affected by synovitis and BME are the symmetrical joints of the hand and feet, with other joints subsequently becoming diseased [1]. Pro-inflammatory cytokines released by the tissues and cells described above result in dysfunctional intracellular signalling responsible for inflammation, cell survival and apoptosis. Pathways involved in RA include the Janus Kinase/ Signal Transducers and Activators of Transcription (JAK/STAT), the Mitogen-­ Activated Protein Kinase (MAPK), and the Phosphatidylinositide-3-Kinase/AKT/mammalian Target of Rapamycin (PI3K/AKT/mTOR), all of which have been previously reviewed [9]. Notably, elevated interleukins in synovial fluid activates the JAK/STAT signalling pathway, which results in the transcriptional expression of STAT-responsive genes including IL-6, IL-10, interferon gamma (INFγ), Oncostatin M (OSM) and TNFA, which contributes to ECM degradation and joint degeneration [9]. The MAPK signalling pathway consisting of p38 MAP kinases, extracellular signal-regulated protein kinases (ERKs) and C-Jun-N-terminal kinases (JNKs) is involved in cytokine responses, NF-kB activation, cell survival and apoptosis. Immune cell and synoviocyte proliferation, apoptosis and survival are regulated by the PI3K/AKT/mTOR pathway [9]. IL-6 has a fundamental immunoregulatory role in RA pathogenesis, regulating inflammatory pathways in immune cells, synoviocytes and osteoclasts. Elevated IL-6 in RA patient synovial

4  Long Non-coding RNAs in Rheumatology

fluid correlates with disease activity and joint destruction [10, 11]. IL-6 binds the soluble IL-6 receptor (sIL-6R) in the synovial fluid and couples with gp130 subunit in synoviocytes or directly binds the IL-6R on leukocytes and macrophages, which activates the JAK/STAT and Ras-MAPK pathways. In synoviocytes this results in hyperplasia and increased IL-6, IL-1 and Toll-like receptors (TLRs), which promotes a perpetual cycle of inflammation, inducing osteoblasts to produce RANKL, leading to osteoclastogenesis, pro-inflammatory cytokine and MMP production and ultimately bone and cartilage destruction [11, 12]. Synoviocyte secreted RANKL binds RANK receptors on activated macrophages activating the NF-kB, MAPK, NFATc1 and Src signalling pathways and promoting bone resorption. Similarly, TNFα is another important cytokine produced by macrophages, which binds TNF receptors (TNFRs) to activate NF-kB, MAPK and protein kinase B (PKB/AKT) inducing inflammation, tissue degeneration and cell proliferation [11].

4.1.2 Osteoarthritis Globally, osteoarthritis (OA) is the most prevalent degenerative joint disorder affecting 303 million people [13]. In the United States, whilst RA effects 1.3 million adults, OA affects 27 million adults, making OA a significant public health challenge [14]. The debilitating condition affects the entire joint causing loss of articular cartilage mass, subchondral bone sclerosis, joint space narrowing and inflamed synovium [15, 16]. The resulting pain and stiffness of the synovial joints leads to progressive disability and reduced quality of life, amounting to a huge socioeconomic burden costing billions. The Global Burden of Diseases, Injuries and Risk Factors Study (2017) found that incidence and prevalence of OA was up by 8–9% since 1990 and that prevalence not only increased with age but was significantly higher in women [17]. Since age is a significant OA risk factor, with an ageing global population coupled with increased life expectancy, OA prevalence is set to keep increasing [17]. Other risk

37

factors include sex (female), obesity, history of joint injury, abnormal loading, diet and genetics [18]. OA in both weight-bearing and non-weight bearing joints has been linked to obesity, suggesting the impact goes beyond increased biomechanical loading [16, 19]. Adipose tissue is an endocrine organ, which in obesity has increased infiltration of macrophages and secretion of pro-­ inflammatory cytokines known as adipokines, which are likely to have systemic effects on joint integrity [16]. Additionally, central adiposity is strongly associated with OA in women, particularly affecting the knee and hand joints [20]. Menopausal women in particular are at greater of risk of developing hip, knee and hand OA due to hormonal factors [18]. Historically, osteoarthritis was considered a ‘wear and tear’ condition due to ageing. However, it is now known that joint inflammation plays a central role in both the incidence and progression of OA disease. OA pathogenesis involves the degradation of cartilage and remodelling of subchondral bone. This is driven in part by chondrocytes in the articular cartilage that secrete IL-6 into the synovial fluid, where it binds soluble IL-6 receptor (sIL-6R) and couples with membrane bound gp130 on fibroblasts thereby promoting additional FLS IL-6 secretion [16]. This chondrocyte-fibroblast crosstalk is further exacerbated in obese patients with OA, where the adipokine leptin stimulates greater IL-6 secretion from articular chondrocytes [16]. OA chondrocytes also secrete PGE2, MMP3 and MMP13 leading to further articular cartilage degradation [21]. Increased MMPs and aggrecanases ADMATS4 and ADMATS5 contribute to catabolism of integral cartilage matric components including collagen type II resulting in destabilised mechanical properties and structural integrity of both cartilage and bone [22]. Additionally, loading in knee OA increases joint space narrowing resulting in severe mechanical degradation exposing the underlying subchondral bone [22]. OA subchondral bone is hypoxic, which inhibits osteoblast mineralization and bone formation further contributing to joint damage [23]. Synovial immune cells such as IFNy and TNF producing T-cells and synovial derived macro-

38

S. N. Wijesinghe et al.

phages which differentiate into osteoclasts are synovial/ cartilage crosstalk resulting in cartilage also thought to induce ostoclastogenesis and degradation [31]. bone remodelling [24]. Cartilage degradation results in the accumulaSimilar to RA, synovitis is now more widely tion of damage-associated molecular patterns recognised to play a significant role in OA joint (DAMPs) in the synovial joint, which are recogpathology. Synovitis in OA is evidenced by nised by pattern recognition receptors (PRRs) increased infiltration of activated B- and T- cells such as TLRs in surrounding tissue leading to and synovial hypertrophy [25]. Cartilage damage activation of a localised innate immune response. is facilitated by the synovium through secreted TLR1-7 and TLR9 are all upregulated in OA cytokines, growth factors, matrix metalloprote- synovium, whilst the soluble TLR4 is recognised ases and aggrecanases into the synovial fluid [19, as an OA severity biomarker in synovial fluid 24]. FLS from OA patients are more inflamma- [32]. TLR4 is also expressed by osteoblasts and tory compared to non-diseased patient controls may be involved in reduced bone mineralisation with femoral neck fracture, and interestingly in OA.  Activated TLRs, through the NF-kB-­ those that are isolated from obese patients with mediated chemokine release, promote macroOA have an increased inflammatory phenotype. phage and lymphocyte infiltration into OA Inflammatory OA-FLS are also reported to secret synovium. OA damaged articular cartilage and greater levels of pro-inflammatory cytokine IL-6 OA chondrocytes express increased levels of and chemokine CXCL8 [19]. Interestingly, tran- TLRs, which stimulate secretion of catabolic facscriptionally distinct FLS subsets are identified in tors including IL-6, cyclo-oxygense 2 (COX-2) early and late-stage knee OA patients and parapa- and MMP13 [25, 32]. COX-2 is differentially tellar synovitis has been associated with increased expressed in OA joints and regulates the arachipain [26]. Obese OA patients also exhibit a FLS donic inflammatory response pathways [28]. In subset with gene signatures related to immune brief, pro-inflammatory cytokines induce COX-­ cell regulation and inflammatory signalling [27]. 2, which catalyses arachidonic acid into an unstaMany of the major signalling pathways which ble eicosanoid precursor, PGH2. PGH2 is then govern joint inflammation in RA are shared with converted into the major pro-inflammatory and OA, such as the IL-6 mediated JAK/STAT and pain mediating prostaglandin PGE2, which is Ras/MAPK pathways discussed earlier. Similarly, significantly elevated in OA cartilage [33]. the NF-kB signalling pathway is described as the Nitric oxide (NO) and inducible NO synthase master regulator of inflammation and as such (iNOS) are also key mediators of OA cartilage regulates pro-inflammatory cytokines including destruction and chondrocyte apoptosis [25]. Both IL-1β, IL-6, IL-17 and TNFα in both OA and RA, NO and iNOS are elevated in OA cartilage and as well as aggrecanases and MMPs which induce patient serum. The pathogenic effects of IL-1β cartilage degradation in OA [28, 29]. In bone and TNFα are mediated by NO activation. homeostasis, receptor activator of nuclear factor However, conversely some reports suggest innate kappa B (RANK)/ RANKL pathway activates immune suppression in the early stages of OA is NF-kB induced transcription factors that balance NO-associated [34]. In OA, the p38 MAPK pathbone resorption and formation which is deregu- way mediates pro-inflammatory cytokine signal lated in OA.  Additionally, an NF-kB transcrip- transduction. DAMPS, IL-1β and TNFα are all tional target is the hypoxia-inducible factor 2 involved in p38 phosphorylation, which is alpha (HIF-2α) which is elevated in hypoxic OA detected in OA chondrocytes and OA articular subchondral bone and OA articular cartilage cartilage to drive OA pathogenesis [25]. p38 [29]. In OA activated chondrocytes, NF-kB MAPK in OA chondrocytes selectively activates ­signalling regulates ECM remodelling and the MAPK-activated protein kinase 2 (MK2), which production of catabolic enzymes and pro-­ regulates TNF stability and IL-1β induced proinflammatory factors [30].Additionally, NF-kB duction of catabolic factors MMP3, MMP13 and mediated signalling in synovial cells may drive PGE2 [21, 25]. Bioinformatics analysis also finds

4  Long Non-coding RNAs in Rheumatology

that MAPK signal transduction pathway is influential in OA synovitis [35]. Additionally, the MAPK signalling transduction pathways are utilised by many adipokines to elicit pro- and antiinflammatory responses. Through MAPK and PI3K pathways, leptin induces naive T-cell proliferation and IL-2 production [36]., whilst the anti-inflammatory adiponectin through binding to adiponectin receptors attenuates IL-6 and TNFα production by affecting p38-MAPK, JNK and NF-kB signalling pathways [36].

39

overexpression. MALAT1 alleviated LPS-­ induced cell injury through upregulation of miR-­ 19b and suppressing the Wnt/β-catenin and NF-kB pathways [40]. Chondroprotective effects of MALAT1 was also reported in primary rat chondrocytes treated with IL-1β to mimic OA inflammation. Gao et al. [41] report overexpression of MALAT1 promotes proliferation and inhibits apoptosis and ECM degradation through the suppression of the JNK signalling pathway. In contrast, MALAT1 is reported to contribute to OA pathogenesis in several patient studies through its actions on chondrocyte proliferation 4.1.3 Long Non-coding RNAs which is likely due to differences in study context in the Pathogenesis than species dependent functionality. Indeed, as of Arthritis reviewed by Arun et al., MALAT1 has numerous context-dependent molecular mechanisms influ4.1.3.1 Metastasis-Associated Lung encing a myriad of physiological conditions [42]. Adenocarcinoma Transcript 1 In human OA chondrocytes, MALAT1 can (MALAT1) sponge and inhibit miR-127-5p expression leadThe highly-conserved 8.5  kb Metastasis-­ ing to increased osteopontin (OPN) expression Associated Lung Adenocarcinoma Transcript 1 and activation of the PI3K/Akt pathway, which in (MALAT1) was amongst the first cancer-­ turn results in increased chondrocyte proliferaassociated lncRNAs to be discovered [37]. tion [43]. Also, MALAT1 competitively binds MALAT1 is nuclear RNA localized in nuclear miR-150-5p, indirectly promoting AKT3 expresspeckles along with pre-mRNA splicing factors sion and resulting in increased proliferation, and thought to regulate alternative splicing by ECM degradation and suppressed apoptosis in modulating serine/arginine splicing factors [38]. primary chondrocytes [44]. Similarly, MALAT1 Several cancer studies have identified MALAT1 directly binds and inhibits miR-145, which can involvement in molecular signalling pathways no longer suppress ADAMTS5 expression thus including NF-kB, PI3K/AKT, WNT/β-catenin promoting ECM degradation and reduced cell and MAPK/ERK associated with proliferation, viability in IL-1β treated primary chondrocytes apoptosis and inflammation [28, 39]. [45]. Li et  al. [46] found through regulation of MALAT1 studies in OA have largely focused miR-146a that MALAT1 indirectly activated the on articular cartilage tissue or articular chondro- PI3K/AKT pathway, regulating proliferation of cytes and to a lesser extent in synovium or LPS treated chondrocytes isolated from the FLS.  However, the expression of MALAT1 is Sprague Dawley (SD) rat model. Additionally, significantly increased in both OA cartilage and siRNA mediated MALAT1 knockdown in human synovium tissue, as well as in isolated chondro- primary OA chondrocytes silenced IL-6, COX-2 cytes and FLS. MALAT1 expression was found and MMP13 and promoted collagen type II to increase in response to LPS stimulation in the expression (COL2A1) suggesting MALAT1 is murine ATDC5 chondrogenic cell line [40]. Pan pro-inflammatory and pro-degradative [46]. et al. [40] report protective effects of MALAT1, These inflammatory mechanisms have also been since overexpression reversed LPS-induced identified in OA patient FLS. MALAT1 expresinflammatory injury. LPS induced expression and sion is elevated in OA synovial tissue compared secretion of apoptotic and pro-inflammatory fac- to non-OA patient tissue, and even more so in OA tors including Bax, caspase 3 and 9, IL-1B, IL-6, patients who are obese. This increase was correIL-8 and TNFα were all suppressed by MALAT1 lated with pro-inflammatory cytokine levels

40

S. N. Wijesinghe et al.

including IL-6 and CXCL8. Similar to findings in domain interacts with the histone demethylase chondrocytes, LNA-Gapmer silencing of complex LSD1/CoREST/REST [54]. MALAT1  in OA-FLS supressed pro-­ Recent studies indicate that HOTAIR lncRNA inflammatory cytokine expression and inhibited may have a significant role in the pathogenesis of their proliferation [19]. both OA and RA. The differential expression of Interestingly, in RA, MALAT1 expression is HOTAIR has been reported in rheumatic condisignificantly reduced in synovium tissue and in tions particularly in the cartilage tissue of both RA-FLS.  Furthermore, it is one of six lncRNA OA and RA patients. Gain (GOF) and loss (LOF) down-regulated in RA serum exosomes [47–50]. of function studies find HOTAIR to be involved LncRNA screening following treatment with the in cell proliferation, apoptosis and inflammation. dietary anti-oxidant quercetin, identified Chen et al. [55], reported an increase in HOTAIR MALAT1 to be upregulated during quercetin-­ expression in response to LPS induction in C28/ induced apoptosis in immortalised RA-FLS [50]. I2 chondrocytes, which correlated with elevated MALAT1 knockdown reversed quercetin-­pro-inflammatory cytokine profiles of IL-6, IL-8 induced apoptosis, reduced caspase-3 and cas- and TNFα and cell injury. Suppression of pase-­9 expression and activated the PI3K/AKT HOTAIR reduced cell proliferation, apoptosis pathway, enhancing cell proliferation [50]. Li and cytokine expression of C28/I2 articular chonet al. [48] reported that MALAT1 was fundamen- drocytes cells [55]. Mechanistically, this study tal in suppressing the Wnt signalling pathway by found that inflammatory injury was regulated recruiting methyltransferases to the promoter of through HOTAIR mediated down-regulation of the CTNNB1 gene, which encodes the β-catenin miR-17-5p which lead to an increase in ETV1. protein. Silencing of MALAT1 to mimic low Through activation of MAPK/c-Jun and NF-kB expression levels in RA-synovial tissue resulted pathways, ETV1 regulated inflammatory damage in activation of the Wnt/β-atenin signalling path- and cell injury [55]. More recently, the HOTAIR/ way, increased primary RA-FLS proliferation miR-17-5p axis has also been described in priand the secretion of pro-inflammatory cytokines mary human chondrocytes isolated from OA IL-6, IL-10 and TNFα [48]. This in contrast to patient articular cartilage tissue. Hu et  al. [56], MALAT1 silencing in OA-FLS where pro-­ reported increased HOTAIR and reduced miR-­ inflammatory factors and proliferation are inhib- 17-­5p expression in human OA diseased cartiited [19]. It is evident that MALAT1 has a lage, which correlated with chondrocyte significant role in inflammation and cell prolifer- apoptosis and extracellular matrix (ECM) degraation in both conditions, although the disease dation in C28/I2 chondrocyte cell line. RNA specific mechanisms of action and the differences immunoprecipitation assays confirmed HOTAIR noted here leave much to be considered. could bind miR-17-5p, which resulted in the indirect upregulation of FUT2 protein. Additionally, 4.1.3.2 HOX Transcript Antisense RNA FUT2 was found to aggravate ECM degradation (HOTAIR) and chondrocyte apoptosis through the Wnt/B-­HOX transcript antisense RNA (HOTAIR) was catenin pathway [56]. Interestingly, in chondrodiscovered in 2007 by Rin et  al. [51], as a sarcoma SW1353 cells, HOTAIR can directly 2158-nucleotide containing long intergenic non-­ activate the Wnt/β-catenin pathway through coding RNA (lincRNA). HOTAIR is expressed increased H3K27 trimethylation at the promoter from the antisense strand of the HOXC genes of the Wnt inhibitory factor 1 (WIF-1) [57]. located on chromosome 12 [52]. This lincRNA is Other miRNAs that are regulated by HOTAIR in an important epigenetic regulator, which OA include miR-130a-3p and miR-20b [58, 59]. ­selectively binds components of the PRC2 com- Upregulated HOTAIR expression is reported in plex including Suz12 and the histone methyl- knee OA patients with radiographic evidence of transferase EZH2 [52, 53]. Whilst the 5′ region of articular cartilage degradation [58]. Increased HOTAIR associates with PRC2 proteins, the 3′ HOTAIR was found to sponge miR-130a-3p in

4  Long Non-coding RNAs in Rheumatology

41

primary knee OA chondrocytes, reducing miR-­ overexpression of HOTAIR suppressed LPS-­ 130a-­3p levels and resulting in repressed autoph- induced inflammation. HOTAIR was found to agy and cell growth leading to chondrocyte directly target and inhibit miR-138-mediated apoptosis [58]. activation of NF-kB signalling in vivo, resulting In the destabilization of the medial meniscus in the suppression of IL-1β and TNFα [63]. (DMM) OA mouse model, silencing of HOTAIR Interestingly, in RA studies overexpression of reversed cartilage degradation, repressed MMP13 HOTAIR is recognised to be protective, reducing and ADAMTS-5 and activated aggrecan and col- catabolic MMPs and inflammatory cytokines, lagen type II production in cartilage [59]. whilst the opposite is true in OA where HOTAIR HOTAIR was identified as a competing endoge- expression promotes cartilage degradation. These nous RNA (ceRNA), which sponged miR-20b opposing mechanisms of HOTAIR in OA and RA resulting in the upregulation of PTEN, a negative suggests there may be condition specific mecharegulator of the PI3K/AKT signalling pathway nisms coordinated by other regulators which are [59]. These findings support a previous study yet to be determined. where HOTAIR was also found to strongly promote ADAMTS-5 expression in human OA artic- 4.1.3.3 Growth Arrest-Specific 5 (GAS5) ular chondrocytes. Dou et  al. [60], found The growth arrest-specific 5 (GAS5) gene overexpression of HOTAIR stabilized encodes several non-coding RNAs including a ADAMTS-5 mRNA, which could be through lncRNA.  Although the molecular mechanisms miR-20b sponging as described by Chen at el are largely unclear, GAS5 is known to regulate [59]. HOTAIR lncRNA has similar pro-­ apoptosis, proliferation, invasion and metastasis inflammatory functionalities in OA synovium tis- [64]. Interestingly, its secondary structure forms sue. HOTAIR expression has been significantly a stem loop that competitively binds and inhibits noted in the synovial fluid of temporomandibular glucocorticoid receptors, which may be of funcjoint OA (TMJ-OA) patients. This correlated tional relevance in rheumatic conditions [65]. with increased MMP1, MMP3, MMP9 and GAS5 expression in OA cartilage tissue and HOTAIR in rabbit condylar chondrocytes, a tem- chondrocytes is reported to be significantly poromandibular OA model [61]. Additionally, in upregulated [66, 67]. Lentiviral overexpression the ACLT rat model of OA, silencing HOTAIR of GAS5  in primary human OA chondrocytes inhibited the Wnt/β-catenin pathway resulting in inhibited autophagic responses whilst activating reduced synovial inflammation [62] (Tables 4.1, apoptosis and up-regulating expression of several 4.2, and 4.3). MMPs [67]. Song et al. [67] identified a mechaHOTAIR is also described to a lesser extent in nism of reciprocal repression between GAS5 and RA. Song et al. [47] isolated RA patient periph- miR-21, where exogenous GAS5 suppressed eral blood mononuclear cells (PBMCs) and miR-21 resulting in apoptosis and increased serum exosomes to find HOTAIR expression was expression of cartilage MMP13. Lentiviral miR-­ increased by four-fold in these patients. However, 21 injected into mice significantly reduced GAS5 in RA patient FLS, HOTAIR was significantly mRNA levels, DMM-induced cartilage destrucdecreased by threefold. Lentiviral overexpression tion and MMP13 expression. The conditions that of HOTAIR in FLS and osteoclasts significantly regulate this reciprocal inter-regulator repression reduced activation of MMP2 and MMP13. Song between GAS5 and miR-21 requires further et  al. [47] found that LPS-activated monocytic study. More recently, silencing of GAS5 in pricells actively migrated towards RA serum exo- mary chondrocytes promoted proliferation, somes containing high levels of HOTAIR.  This inhibited apoptosis and reduced expression of suggests in vivo circulating HOTAIR-containing pro-inflammatory factors IL-6 and TNFA [68]. exosomes may attract and activate macrophages Double luciferase reporter assays confirmed the inducing immune responses in RA.  More regulatory mechanism of GAS5 lay in the suprecently, in LPS-stimulated rat chondrocytes pression of miR-34a and the subsequent upregu-

S. N. Wijesinghe et al.

42 Table 4.1  Summary of functional lncRNAs in Osteoarthritis

LncRNA MALAT1

Expression (Up ‘+’ / Down ‘–’) +

+

+

HOTAIR

Model Human primary FLS Mouse chondrocyte cell line Rat primary chondrocytes

+

Human primary chondrocytes

+

Human primary chondrocytes

+

Human primary chondrocytes

+

Rat primary chondrocytes

+

Human chondrocyte cell line

+

Human primary chondrocytes

+

Human chondrocyte cell line Human primary chondrocytes

+

+

Mouse primary chondrocytes

+

Human chondrocyte cell line Rabbit primary chondrocytes

+

+

Rat primary synoviocytes

Function Knockdown reduces expression and protein secretion of CXCL8 and IL6 and inhibits the proliferation of FLS Upregulates miR-19b suppressing Wnt/β-catenin and NF-kB pathways and pro-inflammatory factors IL-1β, IL-6, IL-8 and TNFα Prevents activation of JNK signalling pathway supressing IL-1β-induced chondrocyte inflammation, apoptosis and extracellular matrix degradation Acts as a molecular sponge to inhibit miR-127-5p, activating the PI3K/Akt pathway and increasing osteopontin (OPN) expression resulting in increased chondrocyte proliferation Competitively binds miR-150-5p and indirectly promotes AKT3 expression resulting in increased proliferation, ECM degradation and suppressed apoptosis Acts as a molecular sponge to inhibit miR-145, which can no longer suppress ADAMTS5 thus promoting ECM degradation and reduced cell viability Regulates miR-146a which activates the PI3K/ AKT pathway, regulating proliferation and expression of IL-6, COX-2 and MMP13 and COL2A1 Inhibits miR-17-5p mediated suppression of ETV1 which elevates pro-inflammatory cytokines IL-6, IL-8 and TNFα through activation of MAPK/c-Jun and NF-kB pathways Sponging of miR-17-5p upregulates FUT2 increasing ECM degradation and apoptosis through the Wnt/β-­catenin pathway Directly activates the Wnt/β-catenin pathway through increased H3K27 trimethylation at the promoter of the Wnt inhibitory factor 1 Sponges miR-130a-3p reducing miR-130a-3p levels resulting in repressed autophagy and cell growth leading to chondrocyte apoptosis By sponging miR-20b upregulates PTEN, a negative regulator of the PI3K/AKT signalling pathway causing ECM degradation and chondrocyte apoptosis Stabilizes ADAMTS-5 mRNA through miR-20b sponging in chondrocytes Knockdown reverses IL-1β-stimulated expressions of MMP1, MMP3 and MMP9 and significantly decrease apoptosis Silencing inhibits Wnt/β-catenin pathway and reduced inflammation and promoted synoviocytes apoptosis

References [19]

[40]

[41]

[43]

[44]

[45]

[46]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

(continued)

4  Long Non-coding RNAs in Rheumatology

43

Table 4.1 (continued)

LncRNA GAS5

H19

Expression (Up ‘+’ / Down ‘–’) +

+

Human primary chondrocytes

−

Mouse chondrocyte cell line Human primary chondrocytes

+

+

+

− NEAT1

Model Human primary chondrocytes

+

Human chondrocyte cell line Human primary chondrocytes Human chondrocyte cell line Rat primary FLS and chondrocytes Human primary chondrocytes

+

Human primary chondrocytes

+

Mouse and Human chondrocyte cell line Human primary chondrocytes

−

Function Exogenous GAS5 suppresses miR-21 resulting in apoptosis and increased expression of cartilage MMP13 whilst lentiviral miR-21 represses GAS5, MMP13 and cartilage destruction Suppresses miR-34a upregulating apoptotic regulatory protein Bcl-2 increasing apoptosis and expression of pro-inflammatory factors IL-6 and TNFA. Positively regulates KLF2 which suppresses the NF-kB and Notch signalling pathway alleviating LPS-induced inflammation Induced under hypoxic conditions and silenced when stimulated with pro-inflammatory cytokines IL-1β and TNFα Found to sponge miR-130a resulting in LPSinduced apoptosis and inflammation Increased H19 stimulated by IL-1β, inhibits proliferation and induces apoptosis through sponging of miR-106a-5p Suppresses miR-140-5p to regulate cartilage degradation and calcification, increasing MMP1 and MMP13 FLS exosomes containing H19 were responsible for cartilage repair through targeting of miR-106b-5p Sponges miR-193-3p activating SOX5, resulting in elevated IL-6, IL-1B, TNFA and IL-8 expression, increased apoptosis and ECM degradation miR-377-3p sponging by NEAT1 in IL-1β stimulates chondrocytes, increases inflammation, apoptosis and cartilage degradation through elevated ITGA6 expression A ceRNA silencer of miR-16-5p inhibits apoptosis whilst reducing expression of NEAT1 increased apoptosis and inflammatory cytokines Anti-apoptotic and inflammatory ceRNA of miR-181a which regulates GPD1L

References [67]

[68]

[69]

[74]

[75]

[76]

[77]

[78]

[81]

[82]

[83]

[84] (continued)

S. N. Wijesinghe et al.

44 Table 4.1 (continued)

LncRNA XIST

Expression (Up ‘+’ / Down ‘–’) +

+

−

+

MEG3

Human chondrocyte cell line Human and Mouse chondrocyte cell lines Human primary chondrocytes

+

Human chondrocyte cell line

+

Human primary chondrocytes

+

Human primary chondrocytes

−

Rat primary chondrocytes Rat primary chondrocytes

− −

Mouse chondrocyte cell line

−

Rabbit and Human chondrocyte cell line Human primary chondrocytes

− HOTIP

Model Human primary chondrocytes

+ +

Mouse primary chondrocytes Human primary chondrocytes

Function Regulates CXCR4 and downstream MAPK signalling to regulate proliferation and apoptosis through the XIST/ miR-211 axis miR-142-5p/SGTB/XIST axis described to impact on cell growth and apoptosis resulting in increased MMP13 and Bax and suppressed Bcl-2 Overexpression inhibits apoptosis through the miR-­653-­5p/SIRT1 axis

References [95]

Promotes MMP-13 and ADAMTS5 mediated ECM degradation by functioning as a ceRNA of miR-1277-5p. By sponging miR-149-5p, XIST enhanced DNMT3A expression supressing collagen type II and aggrecan production, inhibiting proliferation and promoting apoptosis Recruits DNMT1, DNMT3A and DNMT3B to increase TIMP-3 promoter methylation, thereby silencing TIMP-3 and promoting collagen degradation A ceRNA of miR376c-5p, which is essential for silencing osteopontin known to regulate pro-­ inflammatory cytokines within M1 macrophages, which in turn promotes chondrocyte apoptosis Overexpression is anti-proliferation and proapoptotic through the miR-16/SMAD axis Disrupts the miR-93/TGFBR2 axis activating the TGFβ signalling pathway which regulates ECM degradation A ceRNA of miR-203 whose downstream target, SIRT1, alleviates LPS-induced inflammatory injury through the PI3K/AKT and NF-kB pathways in the absence of MEG3 Overexpression relieves OA-associated pain through suppression of pro-inflammatory cytokines IL-6, TNFA, IL-1B and IL-8

[98]

Targets the miR-361/FOXO1 regulatory axis, which promotes proliferation whilst suppressing apoptosis and ECM degradation Suppresses HoxA13 which regulates integrin-α1 expression and cartilage maintenance HOTTIP targets the miR-455-3p/CCL3 pathway in OA inducing cartilage degradation

[96]

[97]

[99]

[100]

[101]

[105] [106]

[107]

[108]

[109]

[115] [116] (continued)

Table 4.1 (continued)

LncRNA PVT1

Expression (Up ‘+’ / Down ‘–’) + +

+

+

+

TUG1

+

+

UCA1

+

CASC2

+

+

ANRIL

+

Lnc-DILC

−

IGHCy1

+

lincRNA-p21

+

SNHG1

THRIL

+

ZFAS1

−

MIAT

FAS-AS1

Model Human primary chondrocytes Human primary chondrocytes Human chondrocyte cell line Human primary chondrocytes Human chondrocyte cell line Human primary chondrocytes Mouse chondrocyte cell line Human chondrocyte cell line Human chondrocyte cell line Human chondrocyte cell line Human primary FLS Human chondrocyte cell line Human THP-1 cell line Human primary chondrocyte Human chondrocyte cell line Mouse chondrocyte cell line Human primary chondrocytes

Mouse chondrocyte cell line +

Human primary chondrocytes

Function Overexpression of induces apoptosis through sponging of miR-488-3p Silenced IL-1β induced secretion of IL-6, IL-8 and TNFα and expression of MMP3, MMP9 and MMP13 through sponging of miR-149 Knockdown inhibits apoptosis and inflammatory response to IL-1β treatment via up-regulated miR-­27b-­3p targeting TRAF3 Sponging of miR-26b facilitates CTGF expression enhanced cartilage degradation and increases TGF-β1, SMAD3, and MMP-13 Induces TNFA expression and secretion through miR-211-3p sponging facilitating apoptosis

References [119]

Overexpression regulates ECM degradation through the miR-195 suppression and increased MMP-13 expression Upregulation attenuated apoptosis and inflammation by inactivating the Notch and NF-kB signalling pathways Regulates cell survival and matrix synthesis by suppressing the miR-204-5p expression and increasing MMP-13 expression Upregulation promotes apoptosis but is targeted by miR-93-5p for degradation which reverses these effects Overexpression upregulates IL-17 expression, enhances apoptosis and suppresses cell proliferation By sponging miR-122-5p increases DUSP4 expression and regulates proliferation and apoptosis Overexpression supresses IL-6 at the protein level

[127]

ceRNA of miR-6891-3p resulting in increased TLR4 and NF-kB activity promoting IL-6 and TNFα production Sponges and represses miR-451 promoting the apoptosis Acts as a molecular sponge of miR-16-5p to inhibit ERK1/2 and phosphorylated p38 and p65 involved in p38/MAPK and NF-kB signalling pathways Overexpression promotes LPS-induced inflammatory injury by supressing miR-125b thus activating JAK1/STAT3 and NF-kB pathways. Overexpression promotes proliferation and cell migration whilst inhibiting apoptosis and matrix synthesis through suppression of Wnt3a, β-catenin and p53 Silencing attenuates LPS-induced apoptosis and cytokines release by regulating miR-132 expression which inhibits NF-kB and JNK pathways Low expression decreases expression of MMP1 and MMP13, but increases COL2A1 expression, inhibiting cell apoptosis and promote cell proliferation

[120]

[121]

[122]

[123]

[128]

[130]

[134]

[135]

[140]

[144]

[147]

[149] [153]

[155]

[159]

[179]

[181]

S. N. Wijesinghe et al.

46 Table 4.2  Summary of functional lncRNAs in Rheumatoid Arthritis LncRNA MALAT1

HOTAIR

GAS5

Expression (Up ‘+’ / Down ‘–’) Model – Human primary FLS

–

Human FLS cell line

+

Human whole blood

–

Human primary chondrocytes

–

Human primary FLS

–

Human primary FLS

–

Human primary FLS

Function Silencing stimulates β-catenin nucleation, secretion of pro-­ inflammatory cytokines IL-1, IL-10, and TNFα, elevated proliferation and suppressed apoptosis of FLS Knockdown reversed quercetin-induced apoptosis, reduced caspase-3 and caspase-9 expression and activated the PI3K/AKT pathway, enhancing cell proliferation HOTAIR-containing exosomes attract and activate macrophages inducing immune responses suppressing activation of MMP2 and MMP13 Targets and inhibits miR-138-mediated activation of NF-kB signalling in vivo, resulting in increased cell proliferation and suppressed IL-1β and TNFα Silencing reversed Tan IIA effects by down-­ regulating expression of pro-apoptotic caspases 3 and 9 and activating the PI3K/AKT pathway Overexpression downregulated IL-18 expression and promoted apoptosis Inhibiting GAS5 promoter methylation increased GAS5 expression supressing apoptotic regulator HIPK2 and pro-­ inflammatory cytokines TNFA and IL-6

References [48]

[50]

[47]

[63]

[70]

[71]

[72]

(continued)

4  Long Non-coding RNAs in Rheumatology

47

Table 4.2 (continued) LncRNA H19

Expression (Up ‘+’ / Down ‘–’) Model + Human primary FLS and macrophages

+

Human FLS cell line

NEAT1

+

Human whole blood

MEG3

–

Human primary FLS

–

Human primary chondrocytes and FLS

+

Human primary FLS

HOTIP

Function Expression responds to serum starvation, IL-1β, TNFα and PDGF-BB stimulation and is regulated by the MAPK/ ERK1-2 signalling pathway Promotes phosphorylation of TAK1, a MAP3 kinase known to activate the JNK/p38MAPK and NF-kB pathway, resulting in increased IL-6, IL-8 and IL-1β production and increased apoptosis Knockdown prevents CD4+ T-cells from differentiating into pro-inflammatory Th17 cells correlated with RA pathogenesis Suppression promotes proliferation, secretion of inflammatory cytokines IL-6 and IL-8 and invasion, stimulating the STAT3 and PI3K/AKT pathways Overexpression facilitates cell proliferation and inhibited inflammation by downregulating miR-141 and inactivating the AKT/ mTOR pathway Recruits Dnmt3b to facilitate SFRP1 promoter methylation which activates the Wnt signalling pathway, proliferation, invasion, and migration, while supressing apoptosis

References [79]

[80]

[85]

[111]

[110]

[117]

(continued)

S. N. Wijesinghe et al.

48 Table 4.2 (continued) LncRNA PVT1

Expression (Up ‘+’ / Down ‘–’) Model + Human FLS cell line

+

Rat primary FLS

UCA1

–

Human FLS cell line

CASC2

–

Human primary FLS

Lnc-DILC

–

Human primary FLS

lincRNA-p21

–

Human THP-1 cell line

THRIL

+

Human primary FLS

ZFAS1

+

Human primary FLS

Function Promotes proliferation through the miR-543/ SCUBE2 axis whilst PVT1 knockdown results in apoptosis and supressed inflammation Knockdown restores sirt6 expression through decreasing sirt6 methylation thereby alleviating RA Regulates expression of Wnt6 and induces apoptosis Overexpression suppresses IL-17 which promotes apoptosis Overexpression induces apoptosis and supresses IL-6 at the protein level Induced by methotrexate through DNA-protein kinase catalytic subunit dependent mechanisms contributing to NF-kB activation Regulates cell growth and inflammatory response by activating the PI3K/AKT signalling pathway Promotes cell migration and invasion through sponging of miR-27a

References [124]

[125]

[131]

[136]

[145]

[150]

[157]

[159]

lation of the apoptotic regulatory protein Bcl-2. In RA, GAS5 is significantly upregulated in In contrast, effects reported in mouse chondro- peripheral blood but down regulated in RA synogenic ATDC5 cells found LPS-induced inflam- vial tissue and primary RA-FLS [47, 70–72]. mation suppressed GAS5 mRNA levels, which Profiling of blood samples from RA patients promoted apoptosis [69]. Arguably LPS may pro- found GAS5 to be one of several lncRNAs to be mote apoptosis independently of GAS5, however significantly upregulated in RA blood monocyte GAS5 overexpression also alleviated LPS-­ cells [47]. Treatment of primary RA-FLS with induced inflammation suggesting lncRNA mech- the cytotoxic, anti-inflammatory antioxidant anisms may differ between mice and human. Tanshinone IIA (Tan IIA) induced apoptosis and Mechanistically, Li et al. [69] found GAS5 posi- significantly up-regulated GAS5 expression. tively regulated the KLF2 transcription factor Silencing of GAS5 reversed these effects of Tan which in turn suppressed the NF-kB and Notch IIA by down-regulating the expression of pro-­ signalling pathways. apoptotic caspases 3 and 9 and activating the

4  Long Non-coding RNAs in Rheumatology

49

Table 4.3  Summary of lncRNAs in Systemic Lupus Erythematosus LncRNA FAS-AS1

Expression (Up ‘+’ / Down ‘–’) +

MALAT1

+

Human whole blood

+

Human whole blood

–

Human whole blood

+

Human whole blood

GAS5

NEAT1

Model Human whole blood

Human whole blood

+

Human whole blood

+

Human whole blood

+

Human renal cell line

Function Expression is correlated with nephritis and positively correlated with anti-dsDNA antibody levels Silencing reduced expression of IL-21 and SIRT1 Silencing represses all OAS proteins as well as TNFA and IL-1B expression in IFNα-2a treated immune cells. May function as a ceRNA of six miRNAs which target OAS proteins co-expression of GAS5, lnc0640 and lnc5150 may modulate the MAPK and PPAR signalling pathways Elevated in CD4+ T cells of patients with SLE may serve as potential biomarker for diagnosis upregulated in SLE patients identified on whole blood microarray and validated in patient samples an early response lncRNA which selectively regulates TLR4-mediated inflammatory genes through the MAPK pathway Expression in granulocyte MDSCs induces secretion of B-cell activating factor (BAFF), which promoted IFN-signalling activation of B-cells. Silencing alleviates lupus symptoms Contributes to inflammatory cell injury, elevated IL-1β, IL-6, TNFα and IFN-y production and increased apoptosis by sponging of miR-146b and increasing TRAF6 expression which activates NF-kB signalling

References [180]

[187]

[188]

[177]

[190]

[178]

[191]

[192]

[193]

(continued)

S. N. Wijesinghe et al.

50 Table 4.3 (continued) LncRNA XIST

TUG1

UCA1

Expression (Up ‘+’ / Down ‘–’) +

Model Human whole blood

+

Mouse primary B-cells

+

Human whole blood

+

Human whole blood

–

Human kidney cell line

–

Mouse whole kidney

+

Mouse B-cell cell line

Function RNA localization patterns disrupted, evidence of bi-allelic expression and increased transcription of immunity-related genes in SLE lymphocytes B cells of late stage SLE NZB/W F1 mice have decreased localization of Xist RNA to the Xi and increased expression of x-linked genes TLR7 and CXCR3 X-chromosome inactivation maintenance is altered in T cells of SLE patients thus X-linked genes are abnormally upregulated Skewed allelic expression of X-linked genes attributed to high variability of DNA methylation levels which was reversed by XIST knockdown Overexpression targeted the miR-223/SIRT1 axis activating the PI3K/AKT signalling whilst suppressing NF-kB pathway, increasing cell viability and supressing inflammation Inhibition of the NF-kB signalling pathway with PDTC drug mitigated SLE progression and resulted in the up-regulation of TUG1 lncRNA Expression correlated with evidence of active stage and pathological lesions. Overexpression increased B-cell proliferation through activation of the PI3K/ AKT pathway

References [195]

[196]

[197]

[198]

[202]

[203]

[204]

(continued)

4  Long Non-coding RNAs in Rheumatology

51

Table 4.3 (continued) LncRNA THRIL

Expression (Up ‘+’ / Down ‘–’) +

Model Human kidney cell line

PI3K/AKT signalling pathway [70]. In RA patient plasma, GAS5 expression was found to be inversely correlated to concentrations of IL-18, a pro-inflammatory cytokine known to contribute to RA pathogenesis [71]. Overexpression of GAS5  in primary FLS was found to downregulate IL-18 expression and promote apoptosis. Anti-inflammatory effects of GAS5 in RA were echoed in reports that found the GAS5 promoter to be hypermethylated in RA synovial tissue and patient RA-FLS [72]. GAS5 promoter methylation was inhibited with 5-aza-­ 2-deoxycytidine which increased the expression of GAS5 and decreased the expression of the apoptotic regulator HIPK2 and pro-inflammatory cytokines TNFA and IL-6. Collectively, these multiple studies suggest GAS5 has a significant role in regulating apoptosis and inflammation in both RA and OA.

4.1.3.4 H19 Imprinted Maternally Expressed Transcript (H19) The highly evolutionary conserved H19 gene is an imprinted gene which encodes a 2.3  kb lncRNA.  H19 is known for its tumour suppressive effects in cancer where it is associated with cell viability, migration and invasion [73]. Upregulated H19 expression is observed in RA synovial tissue and OA cartilage. Microarray analysis of OA cartilage found H19 was one of 21 up-regulated lncRNAs [66]. Steck et al. [74] found H19 was induced under hypoxic conditions in primary OA chondrocytes and was silenced when stimulated with pro-inflammatory

Function Overexpression increases apoptosis and expression of pro-inflammatory cytokines IL-1B, IL-6, IL-8 and TNFA. Identified as a ceRNA of miR-34a which targets MCP-1 activating the JNK and Wnt/β-catenin signalling pathways

References [205]

cytokines IL-1β and TNFα. In the human chondrogenic cell line C28/I2, elevated H19 was found to sponge miR-130a resulting in LPS-­ induced apoptosis and inflammation [75]. Similarly, elevated H19 in primary human chondrocytes stimulated by IL-1β, inhibited proliferation and induced apoptosis. RNA-immunoprecipitation (RIP) assays confirmed H19 sponging of miR-106a-5p, whose overexpression reversed H19 effects [76]. In HC-A cells, silencing H19 not only facilitated proliferation but also suppressed MMP1 and MMP13 whilst upregulating COL2A1 levels. Yang et al. [77] found H19, through suppression of miR-140-5p, could regulate cartilage degradation and calcification in OA.  In contrast, Tan et al. [78] found primary OA-FLS exosomes containing H19 were responsible for cartilage repair through targeting of miR-106b-5p. They also reported decreased H19 expression in OA cartilage as well as a silencing of H19 in OA chondrocytes in response to IL-1β stimulation [78]. In primary RA-FLS stimulated with IL-1β, H19 was significantly elevated, which was also demonstrated to a lesser extent in primary OA-FLS [79]. Stuhlmuller et al. [79] found H19 expression also responded to serum starvation, TNFα and platlet-­ derived growth factor-BB (PDGF-BB) stimulation and was significantly higher in RA isolated synovial macrophages. Inhibitor assays showed that H19 RNA expression was under the control of the MAPK/ ERK1-2 signalling pathway. Similarly, pro-inflammatory stimulation of RA-FLS MH7A cell line with TNFα increased

52

H19 expression, increased IL-6, IL-8 and IL-1β production and increased apoptosis [80]. Through LOF and GOF studies it was determined that H19 promoted the phosphorylation of TAK1, a MAP3 kinase known to activate the JNK/p38MAPK and NF-kB pathway in RA resulting in cellular inflammation of RA synovial MH7A cells.

4.1.3.5 Nuclear Enriched Abundant Transcript 1 (NEAT1) The Nuclear Enriched Abundant Transcript 1, NEAT1, is found in neighbouring regions of MALAT1 on chromosome 11 and shares several similarities with MALAT1 which was previously known as NEAT2 [53]. Like MALAT1, NEAT1 is found mainly localised in the nucleus and is necessary for the formation of the nuclear paraspeckles, which are ribonucleoprotein (RNP) bodies thought to regulate gene expression. NEAT1 lncRNA is fundamental for maintaining the paraspeckle architecture, where it also influences splicing factors. This lncRNA enables the expression of cytokines and antiviral genes including IL-8 by binding to the SFPG (splicing factor proline/glutamine-rich) RNA-binding protein and sequestering it within the paraspeckles. Removal of SFPG from the IL-8 promoter alleviates repression at this locus allowing IL-8 to be transcribed [54]. NEAT1 expression in OA cartilage tissue and chondrocytes is upregulated and has been described to regulate several miRNAs. Lui et al. [81] found NEAT1 sponged miR-193-3p activating SOX5, resulting in elevated IL-6, IL-1B, TNFA and IL-8 expression, increased apoptosis and promotion of ECM degradation in primary chondrocytes. Similarly, miR-377-3p was also silenced by NEAT1 sponging in IL-1β stimulated primary chondrocytes resulting in increased inflammation, apoptosis and cartilage degradation through elevated ITGA6 expression [82]. Additionally, NEAT1 was identified as a ceRNA silencer of miR-16-5p. However, in mouse ATDC5 chondrocyte cells, this inhibited apoptosis [83]. Similarly, Wang et  al. [84] also report NEAT1 to be anti-apoptotic ceRNA of miR-181a in human chondrocytes suggesting there may be

S. N. Wijesinghe et al.

miRNA specific regulatory mechanisms. Interestingly, NEAT1 expression is down-­ regulated in synovial tissue [84]. In RA, NEAT1 expression is reportedly upregulated in RA blood exosomes, RA PBMCs, and in Th17 cells induced from RA CD4+ T-cells [47, 85]. RA pathogenesis is correlated with elevated levels of pro-­ inflammatory T-helper cells (Th17s) in PBMCs. Shui et  al. [85] found NEAT1 knockdown ­prevented CD4+ T-cells from differentiating into Th17 cells suggesting NEAT1 is involved in RA development.

4.1.3.6 X-Inactive Specific Transcript (XIST) One of the first lncRNAs to be as characterised as many protein-coding transcripts was X-Inactive Specific Transcript (XIST) lncRNA [86]. The X-chromosome consists of numerous immune genes that are silenced through mechanisms of X chromosome inactivation (Xi). Xi is essential for dosage compensation of the X chromosome in female mammals. LncRNA XIST is fundamental in recruiting the PRC2 complex for chromosome wide silencing through H3K27me3 [87]. More recently, XIST has been reported as a microRNA sponge in numerous conditions, although this may very well be a sex-specific regulatory mechanism considering XIST is nearly exclusively expressed in females [88]. Certainly rheumatic conditions are highly prevalent in females possibly due to differential levels of hormones, the ability of women to get pregnant, the health consequences that can manifest as a result of pregnancy and giving birth, as well as the number of X chromosomes present in female cells [89]. Interestingly, Xi-skewing is reported in RA, where three times as many women are affected [90]. Although the functions of XIST lncRNA in RA is poorly defined, YY1 expression and protein levels are elevated. The YY1 transcription factor is fundamental in bridging XIST lncRNA to the inactive X chromosome for silencing. Additionally, inhibition of YY1 reduced IL-6 expression and inflammation in collagen-induced mouse arthritis model [91].

4  Long Non-coding RNAs in Rheumatology

Reports suggest twice as many women as men develop OA of the knee, although there is little differences in the incidence of OA reported in other joints between males and females [92, 93]. As such, in recent years few mechanistic studies have explored these sex specific effects. However, cartilage tissue, chondrocytes and synovium from OA patients all highly express XIST lncRNA and studies largely report an XIST/ miRNA regulatory function [94]. OA pathogenesis is characterised by cartilage degeneration, which involves chondrocyte apoptosis. Through regulation of the chondrocyte apoptosis contributor CXCR4 and downstream MAPK signalling, the XIST/ miR-211 axis was found to regulate proliferation and apoptosis in primary chondrocytes [95]. Similarly, the miR-142-5p/SGTB/ XIST axis was described in IL-1β treated SW1353 chondrocytes to impact on cell growth and apoptosis [96]. Although, one study in CHON-001 and ATDC5 chondrocyte cell lines found overexpression of XIST to inhibit apoptosis through the miR-653-5p/SIRT1 axis [97]. XIST could also promote MMP-13 and ADAMTS5 mediated ECM degradation by functioning as a ceRNA of miR-1277-5p. This was validated in the DMM OA rat model, where downregulation of XIST proved to be protective against ECM degradation [98]. Additionally, by sponging of miR-149-5p, XIST was found to enhance DNMT3A expression supressing collagen type II and aggrecan production, inhibiting proliferation and promoting apoptosis of IL-1β treated CHON-001 chondrocyte cell line [99]. Interestingly, collagen degradation in primary OA chondrocytes is reportedly regulated by MMP inhibitor TIMP-3. XIST was found to recruit DNMT1, DNMT3A and DNMT3B to increase TIMP-3 promoter methylation, thereby silencing TIMP-3 and promoting collagen degradation [100]. OA chondrocyte apoptosis is also regulated by M1 macrophages via the XIST/ miR-376c-5p/OPN axis in co-culture studies [101]. XIST was identified as a ceRNA of miR376c-5p, which was essential for silencing osteopontin (OPN) known to regulate pro-­ inflammatory cytokines within M1 macrophages,

53

which in turn promoted apoptosis in primary chondrocytes.

4.1.3.7 Maternally Expressed Gene 3 (MEG3) The maternally expressed gene 3 (MEG3) lncRNA is a chromatin binding transcript known to interact with the PRC2 complex [102]. MEG3 recognises GA-rich DNA regions within promoter regions of common EZH2 target genes. In this way, it functions as a guide lncRNA for PRC2 and binds chromatin through a RNA-DNA triple helix conformation [102, 103]. MEG3 expression is downregulated across cancers and similar observations are also reported in rheumatic conditions. Functionally, MEG3 is involved in apoptosis and proliferation through modulating the TGFβ and Wnt/β-catenin signalling pathways and the regulation of p53 [102]. MEG3 down regulation is observed in OA cartilage tissue and chondrocytes, although there are some conflicting reports [104–106]. In ATDC5 cells, MEG3 functioned as a ceRNA of miR-203 whose downstream target, SIRT1, could alleviate LPS-induced inflammatory injury through the PI3K/AKT and NF-kB pathways in the absence of MEG3 [107]. Interestingly, treatment of rabbit joints with the pain eliminating nerve inhibitor methylene blue elevated MEG3 expression. Here, MEG3 overexpression was found to relieve OA-associated pain through suppression of pro-inflammatory cytokines IL-6, TNFA, IL-1B and IL-8 [108]. Overexpressed MEG3 was found to be anti-proliferation and pro-apoptotic through the miR-16/SMAD axis in IL-1β treated SD rat chondrocytes [105]. In line with this, a more recent study, using the same IL-1β treated rat OA chondrocytes, also reported MEG3 to be downregulated. However, here overexpression of MEG3 resulted in increased proliferation, suppressed apoptosis and alleviated ECM degradation. Chen et al. [106] found MEG3 to disrupt the miR-93/TGFBR2 axis thus activating the TGFβ signalling pathway which regulates ECM degradation. Although similar findings have been reported in primary chondrocytes isolated from OA patient tissue. Wang et  al. [109]

54

reported MEG3 targeting of miR-361/FOXO1 regulatory axis, which promoted proliferation whilst suppressing apoptosis and ECM degradation. Interestingly, MEG3 is highly expressed in RA synovial tissue and RA-FLS, and in vivo studies in SD rats found this overexpression facilitates cell proliferation and inhibited inflammation by downregulating miR-141 and inactivating the AKT/mTOR pathway [110]. However in a contradictory study, primary RA-FLS MEG3 expression was found to be down regulated and further suppression promoted proliferation and invasion, stimulating the STAT3 and PI3K/AKT pathways [111]. The handful of studies mentioned here utilise various models from primary human FLS to immortalised cell lines as well as several animal models. Lu et  al. 2019, cited trauma patients undergoing joint placement as appropriate controls however on average these patients were 10  years younger than the OA patients [111]. Whilst another study failed to describe the designation of ‘healthy’ control [110]. The many contradictions stipulated here may be attributed to these differences in controls used, studies being underpowered or choice of study model.

4.1.3.8 HOXA Transcript at the Distal Tip (HOTTIP) The HOXA transcript at the distal tip (HOTTIP) transcript is a ~3.8  kb lncRNA that is highly expressed across many cancers and is known to regulate the HOXA locus. Through binding of WDR5 protein and recruitment of the histone methyltransferase protein MLL, HOTTIP drives activation of the HOXA genes through H3K4 methylation [112]. Reports also find HOTTIP can enhance IL-6 expression in ovarian cancer tissue through binding of c-jun. Additionally, HOTTIP enhanced IL-6 secretion in ovarian cancer tissue promoted neutrophil induced inhibition of T-cell activity [113, 114]. These findings may also be functionally relevant in RA and OA where HOTTIP expression is similarly increased in RA-FLS, OA cartilage and chondrocytes and patients present with elevated IL-6 levels. HOTTIP has been linked to the progression of OA through suppression of HoxA13 in chondro-

S. N. Wijesinghe et al.

genic mouse mesenchymal stem cells (MSC), which modulated integrin-α1 expression and cartilage maintenance [115]. Additionally in human chondrogenic MSC, HOTTIP targets the miR-­ 455-­3p/CCL3 pathway in OA inducing cartilage degradation [116]. In primary RA-FLS, HOTTIP is thought to recruit DNA methyltransferase Dnmt3b to silence SFRP1 [117]. Through Dnmt3b HOTTIP could also activate the Wnt signalling pathway leading to inflammation. Overexpression of HOTTIP in the rat adjuvant-­ induced RA model resulted in synovial tissue hyperplasia, increased infiltration of i­ nflammatory cells and elevated IL-6 and IL-8 production and MMP3 expression [117].

4.1.3.9 Plasmacytoma Variant Translocation 1 (PVT1) Plasmacytoma variant translocation 1 (PVT1) is a highly conserved lncRNA transcribed from a prominent cancer-associated region on chromosome 8. PVT1 is a multifaceted lncRNA whose function includes miRNA regulation, epigenetic coordination involving PRC2, cell cycle modulation as well as numerous other signalling pathways [118]. As in cancerous tissues, PVT1 is upregulated in the rheumatic conditions discussed [66]. In OA, PVT1 is largely described as a sponging ceRNA facilitating apoptosis, inflammation and cartilage degradation. Overexpression of PVT1  in OA primary chondrocytes induced apoptosis through sponging of miR-488-3p [119]. Through sponging of miR-149, PVT1 mediates cartilage degradation [120]. PVT1 silencing suppressed primary chondrocyte catabolism and inflammation, where IL-1β induced production of IL-6, IL-8 and TNFα and expression of MMP3, MMP9 and MMP13 were all downregulated, whilst production of anabolic factors, collagen type II and aggrecan, were increased. Similarly, the PVT1/miR-27b-3p/ TRAF3 axis promoted apoptosis and inflammation in C28/I2 cells, whilst the PVT1/miR-26b/ CTGF/TGF-B1 axis enhanced cartilage degradation in primary chondrocytes [121, 122]. Interestingly, PVT1 was also found to induce TNFA expression and secretion through miR-­

4  Long Non-coding RNAs in Rheumatology

211-­3p sponging in TMJ-OA FLS, which in turn facilitated SW982 chondrocyte apoptosis [123]. Although elevated PVT1 expression was found to promote proliferation in RA-FLS through the miR-543/SCUBE2 axis, knockdown resulted in apoptosis and supressed inflammation suggesting tissue specific mechanisms of action [124, 125]. In RA-FLS isolated from Lewis rats injected with complete Freund’s adjuvant, evidence suggests PVT1 facilitated promoter methylation of SIRT6, a stress responsive protein known to supress inflammation and bone destruction in arthritic mice [125].

55

expression modulation although the exact mechanism remains to be described [131].

4.1.3.10 T  aurine Up-Regulated 1 (TUG1) The 7.6 kb Taurine up-regulated 1 (TUG1) transcript is a fundamental cancer regulatory lncRNA involved in a variety of biological processes. Mechanistically, TUG1 regulates transcriptional activity of target genes through its ability to sponge miRNAs and by interacting with the PRC2 compelx [126]. TUG1 is overexpressed in RA patient PBMCs, RA patient serum exosomes and OA patient cartilage [47, 127]. TUG1 overexpression was found to regulate ECM degradation in OA through the miR-195/MMP-13 axis in primary chondrocytes [127]. Interestingly emodin-­ induced TUG1 expression in ATDC5 chondrogenic cells attenuated apoptosis and inflammation by inactivating the Notch and NF-kB signalling pathways [128].

4.1.3.12 Cancer Susceptibility Candidate 2 (CASC2) The cancer susceptibility candidate 2 (CASC2) lncRNA was first recognised in 2004 as an onco-­ suppressor in endometrial cancer cells [132]. CASC2 is a ~3.3 kb lncRNA with three alternative transcripts but no putative protein. In cancer, CASC2 has been identified to regulate proliferation through epigenetic actions and by influencing miRNAs and other regulatory pathways such as STAT3, PI3K/AKT, NF-kB and MAPK [133]. CASC2 is reportedly upregulated in OA chondrocytes and patient plasma [134, 135]. Upregulated CASC2 promoted HC-OA chondrocyte cell apoptosis but was found to be targeted by miR-93-5p for degradation, which reversed these effects [134]. Overexpression of CASC2 in human CHON-001 cells upregulated IL-17 expression, enhanced apoptosis and suppressed cell proliferation [135]. Whilst in OA chondrocytes CASC2 and IL17 expression were positively correlated, in RA patient plasma CASC2 expression was downregulated whilst IL-17 was upregulated [136]. Additionally, in primary RA-FLS, overexpression of CASC2 suppressed IL-17 which promoted apoptosis. These results suggest CASC2 may have disease and tissue specific regulatory mechanisms, which require further investigation.

4.1.3.11 Urothelial Carcinoma-­ Associated 1 (UCA1) The urothelial carcinoma-associated 1 (UCA1) lncRNA was initially identified as upregulated in bladder cancer and subsequently across other cancers. UCA1 gene encodes three variants ranging from 1.4 kb to 2.7 kb although the smallest is the most recognised and well-studied as a miRNA sponge [129]. UCA1 is overexpressed in OA cartilage tissue and through miR-204-5p/MMP-13 axis, suppresses type II and type IV collagen and promotes C28/I2 chondrocyte cell proliferation and MMP13 expression [130]. In RA-FLS cell line, UCA1 expression is significantly reduced and thought to induce apoptosis through Wnt6

4.1.3.13 A  ntisense Non-coding RNA in the INK4 Locus (ANRIL) ANRIL is the antisense non-coding RNA in the INK4 locus on chromosome 9 whose transcript is ~38  kb in length [137]. ANRIL epigenetically regulates gene expression by forming a RNP complex with polycomb repressive complexes that regulate mono- and tri-methylation of H3K27 [138, 139]. ANRIL is known to regulate many biological processes including proliferation and apoptosis. In OA cartilage, ANRIL expression is significantly elevated and downregulation with siRNAs in primary OA-FLS results in cell cycle arrest at GO/G1, inhibited proliferation and enhanced apoptosis [140].

56

ANRIL is able to sponge miR-122-5p resulting in increased DUSP4 expression and the subsequent regulation of proliferation and apoptosis [140]. In RA, there are few functional studies of note although in RA patient PBMCs ANRIL expression is reportedly decreased [47, 141]. Interestingly the ANRIL/miR-125a axis has been shown to exacerbate disease severity and inflammation in bronchial asthma, which could be functionally relevant in RA and SLE where miR-125a expression is similarly downregulated [142].

4.1.3.14 LncRNA Downregulated in Liver Cancer (Lnc-DILC) The lncRNA downregulated in liver cancer stem cells (lnc-DILC) mediates crosstalk between TNFA/NF-kB signalling and IL-6/STAT3 cascade [143]. Lnc-DILC binding sites were also confirmed at the IL-6 promoter in liver cancer stem cells which through lnc-DILC binding blocks IL-6 expression [143, 144]. In both OA and RA patient plasma the lnc-DILC expression is low whilst IL-6 is elevated [145]. In primary RA-FLS, overexpression of lnc-DILC was found to induce apoptosis and supress IL-6 but only at the protein level [145]. Similar overexpression in CHON-001 chondrocytes also inhibited IL-6 production, although had no significant effects on proliferation and apoptosis [144]. In both studies, IL-6 inhibition occurs at the protein rather than mRNA level suggesting lnc-DILC mechanisms effect IL-6 translation. Although the full regulatory mechanisms are poorly defined in RA and OA, lnc-DILC has great therapeutic potential in reducing IL-6 driven inflammation. 4.1.3.15 IGHC Gamma 1 (IGHCy1) IGHCgamma1 (IGHCy1) is a lncRNA transcript significantly upregulated in RA clinical samples and positively correlated with erythrocyte sedimentation rate [146]. IGHCy1 is highly expressed in OA patient PBMCs and in PMA-induced THP-1 macrophages activated with LPS [147]. Silencing with siRNA reduced macrophage cell proliferation. IGHCy1 was identified as a ceRNA of miR-6891-3p resulting in increased TLR4 and NF-kB activity which promoted IL-6 and TNFα production [147].

S. N. Wijesinghe et al.

4.1.3.16 L  ong Intergenic ncRNA p21 (lincRNA-p21) The long intergenic ncRNA p21 (lincRNA-p21) is p53-activated lncRNA that is well characterised in cancer [148]. Modulated by p53, lincRNA-­p21 is a transcriptional repressor involved in triggering apoptosis. Studies also report functions involving protein binding and localisation to chromatin, suppression of targeted mRNA translation as well as cis p21 activation regulating cell cycle [148]. LncRNA-p21 is significantly upregulated in OA patient cartilage tissue [149]. Silencing lncRNA-p21 in primary OA chondrocytes increased cell viability and reduced apoptosis which was reversed by miR-451 overexpression. Tang et  al. [149] found that lncRNA-p21 sponged miR-451 and in this way promoted chondrocyte apoptosis. In RA whole blood, lincRNA-p21 levels were significantly reduced whilst the NF-kB activator p65 was increased [150]. Spurlock et al. [150] found those patients not treated with methotrexate had even lower levels of lincRNA-p21. Methotrexate was found to induce lincRNA-p21 expression through DNA-protein kinase catalytic subunit and contributed to NF-kB activation in THP-1 monocytes. 4.1.3.17 S  mall Nucleolar RNA Host Gene 1 (SNHG1) The small nucleolar RNA host gene 1 (SNHG1) is an lncRNA transcript that can be alternatively spiced into eight snoRNAs [151]. SNGH1 is largely reported as a ceRNA which sponges miRNAs and contributes to cell proliferation, migration and metastasis in cancer [152]. SNHG1 is downregulated in RA patient serum exosomes and in RA patient PBMCs although the biological significance of this in RA is yet to be determined [47]. However, in an IL-1β-induced OA chondrocyte model cell line, SNHG1 overexpression inhibited catabolic and inflammatory factors MMPs, ADMATs, collagen, aggrecans, IL-6, TNFA, COX-2 and PGE2 [153]. SNGH1 was found to sponge miR-16-5p to inhibit ERK1/2, phosphorylated p38 and phosphorylated p65 factors involved in p38/MAPK and NF-kB signalling pathways.

4  Long Non-coding RNAs in Rheumatology

4.1.3.18 T  NF and HNRNPL Related Immunoregulatory LncRNA (THRIL) The THRIL lncRNA was identified in THP-1 macrophages in an RNP-complex with hnRNPL which bind to and suppressed the TNFA promoter, hence its namesake TNF- and HNRNPL-­ related immunoregulatory lncRNA [154]. This lncRNA is reported to also regulate IL-8, CSF1, CCL1 and CXCL10 expression. Interestingly, THRIL expression is elevated in RA and OA patients and in preclinical in vivo models. Pro-­ inflammatory roles are reported in an OA model using ATDC5 cells, where THRIL sponges miR-­ 125b activating the JAK1/STAT3 and NF-kB signalling pathways which induced inflammatory cell injury [155]. Increased THRIL expression is also reported in RA patient T-cells and in primary RA-FLS where THRIL activated the PI3K/AKT signalling pathway modulating cell growth and inflammation [156, 157]. 4.1.3.19 ZNFX1 Anti-Sense 1 (ZFAS1) ZNFX1 antisense RNA1 (ZFAS1) is overexpressed in many cancers and hosts three snoRNAs. ZFAS1 is involved in many cancer-associated biological process, which include increased proliferation, migration, invasion and suppressed apoptosis [158]. Similarly in RA, ZFAS1 is reported to promote cell migration and invasion of patient isolated RA-FLS. ZFAS1 is highly expressed in RA synovial tissue as well as in primary RA-FLS and regulates migration and invasion through sponging of miR-27a [159]. In primary OA chondrocytes, ZFAS1 is downregulated, but its overexpression is reported to promote proliferation and cell migration whilst inhibiting apoptosis and matrix synthesis. Mechanistically, ZFAS1 overexpression was found to significantly suppress Wnt3a, β-catenin and p53 [159].

4.2

Systemic Lupus Erythematosus

Systemic lupus erythematosus (SLE) is another chronic autoimmune disease which leads to inflammation in various parts of the body includ-

57

ing the skin causing rashes, internal organs such as the heart, lungs and kidneys as well as painful and swollen lymph nodes and joints [160]. SLE has an estimated prevalence of 80–100 per 100,000 adults with significant phenotypic heterogeneity. It is one of the leading causes of death in women with a female to male ratio of up to 15:1 [161]. Women also have an earlier peak in disease onset, usually in their 30s–50s, although males with later onset develop more severe comorbidities such as nephritis [160]. Depending on race and ethnicity, those of Black, South/ East Asian and Hispanic decent have significantly increased SLE prevalence with more sever disease activity [162]. Although the cause of SLE is unknown, studies find that SLE heritability is less than 40%. Additionally, several environmental and lifestyle factors are also heavily associated with SLE including smoking, obesity, alcohol consumption, diet and air pollution [160]. The heterogeneity of SLE is such that almost any organ or tissue in the body may be affected with a variety of clinical presentations. In SLE, defective clearance of apoptotic cells and material is central to loss of immune tolerance resulting in the release of nuclear antigens which provoke a cascade of immune responses resulting in auto-reactivity [163]. The pathophysiology is characterised by aberrant immune responses which sustain the production of autoantibodies, driving chronic inflammation [163]. Several effector cells are involved in SLE, including dendritic cells (DCs), T-cells, B-cells, neutrophils, and monocytes. Plasmacytoid dendritic cells (pDC) are activated by neutrophils which undergo a cell death mechanism known as NETosis forming autoantigen containing neutrophil extracellular traps (NETs) [164]. These NETs trigger type-1 IFN production by stimulating TLRs on pDCs, which sustains a positive feedback cycle promoting more NETosis, further pDC activation and enhanced type-1 IFN release. Neutrophils in lupus patients have reduced phagocytic activity, are more apoptotic and prone to NETosis which together stimulates immune activation and tissue damage [164]. SLE myeloid DCs (mDCs), activated by pDC, released IFN-α, secrete pro-­ inflammatory cytokines and activate autoreactive

58

CD8+ T-cells which differentiate into CD4+ T helper cells [165]. Activated pDCs also produce chemokines (CXCL9, CXCL10, CCL3-5), which attract activated T-lymphocytes to sites of inflammation [165]. In SLE, B-cells are influenced by DCs and T-cells to differentiate and produce autoantibodies as a result of failed tolerance checkpoints [166]. More than half of SLE patients present with kidney injury which is a significant contributor to SLE morbidity. The kidney is infiltrated by IL-17 producing T-cells and autoantibody producing B-cells which activate the complement system causing kidney inflammation known as nephritis [167]. Other infiltrating immune cells include pDCS, monocytes, macrophages and platelet aggregates, which bind CD40 on pDCs and monocytes stimulating IFN secretion which facilitates NETosis and further renal tissue damage [163]. The complement system also disrupts the blood-brain barrier resulting in neuronal injury, microglial activation and the infiltration of T-cells [167, 168]. Another common presentation in SLE patients is skin lesions and although not deemed life threatening, cutaneous lupus has a significant contribution in propagating autoimmunity. SLE skin biopsies are abundant in IL-17 secreting T-cells and pDCs, which produce large amounts of IFN-α [167]. SLE shares many of the key inflammatory pathways described in RA and OA including chemokine signalling, T-cell receptor signalling pathway and TLR pathway. As previously mentioned, TLRs, specifically TLR7 and TLR9, trigger type I IFN production in pDCs [169]. TLR signalling stimulates pro-inflammatory cytokine production through MyD88 or IFN-B and IFN-­ inducible genes which act on the NF-kB and MAPK signalling pathways [170]. The IFN signalling pathway is a prominent feature of SLE, which has a central role in SLE pathophysiology. The IFN system consists of ubiquitously expressed IFNα/β receptors (IFNAR) and IFNy (IFNGR) and IFNλ (IFNLR) receptors which are bound by type I, II and III IFN subtypes, respectively, that regulate the expression of 200–2000 genes [169]. A network of cells are involved in the production of IFNs, although the most pro-

S. N. Wijesinghe et al.

lific producer of type I IFN are pDCs [163, 169]. IFN can also act on T-cells to modulate activation, proliferation, differentiation and survival as well as on B-cells to regulate migration, survival, cytokine production and antigen recognition and presentation [171]. T-cells are drawn to sites of inflammation by pDC cytokine production. Pro-inflammatory cytokines such as IL-6, IL-21 and IL-23 activate STAT3, which suppresses IL-2 whilst enhancing transcription of IL-17 and BCL6, which facilitate inflammation and B-cell antibody production [171]. IL-6 can stimulate CD4 T-cells to differentiate into IL-17 producing T-helper cells (Th17). Th17 cells are initiated by IL-21 to produce IL-17 whilst IL-23 maintains sustained expression of IL-17 through the JAK-STAT signalling pathway [172]. SLE T-cells also have elevated serine/threonine protein phosphatase 2A (PP2A), which regulates DNA hypomethylation of IFN-­ regulated loci by suppressing the ERK/DNMT1 pathway [171, 173]. Notably the IL-17 promoter is hypomethylated whilst IL-2 remains methylated and silenced due to a failure in histone deacetylase 1 (HDAC1) recruitment [171]. IL-17 is thought to be a fundamental driver in local tissue damage in SLE patients. Additionally, in SLE, T-cells, macrophages and monocytes secrete TNFα, which acts through TNFR1 and TNFR2 receptors triggering the caspase cascade associated with apoptosis or the activation of NF-kB, JNK and MAPK pro-inflammatory pathways, respectively [172].

4.2.1 Evidence for the Role of lncRNAs in the Pathogenesis of SLE Several lncRNAs have been identified through whole transcriptome profiling of SLE patient samples and many differentially expressed lncRNAs have been validated in SLE patient PBMCs [174–176]. One computational study has used co-expression analysis and ceRNA networks to predict biological significance of some lesser known lncRNAs. Wu et al. [177] found co-­ expression of GAS5, lnc0640 and lnc5150 may

4  Long Non-coding RNAs in Rheumatology

modulate the MAPK and PPAR signalling pathways, contributing to SLE pathogenesis. Additionally, GAS5, lnc0640, lnc3643, lnc7074 and lnc6655 were found to bind miRNAs that targeted genes involved in lncRNA-mRNA co-­ expression networks [177]. These network predictions have yet to be functionally validated in SLE.  MIAT lncRNA is also upregulated in SLE patient serums, although mechanisms have not been established in SLE [178]. However, there are some indications in OA ATDC5 cells where MIAT sponges miR-132 leading to activation of NF-kB and JNK pathways and induction of apoptosis and cytokine release, which may also be functionally relevant in SLE [179]. FAS-AS1 is another lncRNA upregulated in SLE where mechanisms are yet to be determined but its expression is correlated with nephritis and positively correlated with anti-dsDNA antibody levels [180]. Fittingly, in primary OA chondrocytes functional studies find silencing of FAS-AS1 inhibits apoptosis and promotes cell proliferation [181]. Many SLE specific lncRNAs have been correlated with clinical markers such as erythrocyte sedimentation rate (ESR), C reactive protein (CRP), antinuclear antibodies (ANA) and falling complement factors C3 and C4 [182– 186]. Despite identifying these lncRNAs very few have been functionally investigated in SLE to date. Those for which mechanisms have been determined include MALAT1, GAS5, NEAT1, XIST, TUG1, UCA1 and THRIL are all discussed in more detail below.

4.2.1.1 Metastasis-Associated Lung Adenocarcinoma Transcript 1 (MALAT1) Similarly to arthritis, elevated MALAT1 expression is also reported in peripheral blood monocytes (PBMCs), CD19+ B-cells and CD4+ T-cells of SLE patients [187, 188]. Silencing of MALAT1 in primary human monocytes reduced expression of IL-21, an important cytokine in the pathogenesis of SLE.  MALAT1 silencing also suppressed expression of the deacetylase SIRT1 [187]. In another study, MALAT1 expression was positively correlated with type I IFN downstream effectors oligoadenylate synthase (OAS) pro-

59

teins. OAS proteins were differentially expressed in SLE patients with renal disorders (PBMCs: OAS2 and OASL, CD19+ B-cells: OAS3 and OASL, CD4+ T-cells: OAS3) and those with arthritis symptoms (PBMCs and CD19+ B-cells: OAS2 and OAS3, CD4+ T-cells: OAS2). Silencing of MALAT1 repressed all OAS proteins as well as TNFA and IL-1B expression in IFNα-2a treated immune cells. By computation, this study determined that MALAT1 may function as a ceRNA of six miRNAs that all target OAS proteins, although functional validation is required [188].

4.2.1.2 Growth Arrest-Specific 5 (GAS5) In contrast to RA, expression of GAS5 is down regulated in SLE patient plasma [176, 177, 189, 190]. GAS5 was found to be significantly lower in active SLE, which highlighted its potential as a diagnostic marker [189]. LncRNA screening of 240 SLE patients also found GAS5 to be significantly decreased in plasma [177]. GAS5 was one of five proposed lncRNAs that together presented high diagnostic accuracy for SLE.  KEGG pathway analysis of mRNAs associated with SLE found MAPK signalling to be enriched, which correlated with GAS5 lncRNA-mRNA co-­ expression networks as well as ceRNA networks. These predictions together suggest there may be a GAS5/miRNA/MAPK regulatory axis in SLE yet to be characterised. Interestingly, in CD4+ T-cells isolated from SLE patients, GAS5 expression was significantly elevated and presented as a diagnostic marker for SLE patients with ulceration [190]. 4.2.1.3 Nuclear Enriched Abundant Transcript 1 (NEAT1) Whole blood microarrays and qPCR validation find NEAT1 upregulated in SLE patients [178]. Abnormally high levels of NEAT1 lncRNA is also detected in monocytes isolated from SLE patients [191]. Silencing NEAT1 in LPS-induced THP-1 cells down-regulated inflammatory cytokines IL-6, CXCL10 and CCL8. Zhang et  al. [191] determined NEAT1 as an early response gene which selectively regulated TLR4-mediated inflammatory genes through the MAPK pathway.

60

S. N. Wijesinghe et al.

Expansion of myeloid-derived suppressor cells (MDSCs) drives SLE pathogenesis. Through co-­ culture experiments Dong et  al. [192] found NEAT1 expression in granulocyte MDSCs induced the secretion of B-cell activating factor (BAFF), which promoted IFN-signalling activation of B-cells. Furthermore, silencing of NEAT1 alleviated lupus symptoms in lupus-prone MRL/ lpr mouse model. An additional complication of SLE is kidney inflammation known as lupus nephritis effecting ~60% of patients. Elevated NEAT1  in SLE kidney tissues contributed to inflammatory cell injury, which included elevated IL-1β, IL-6, TNFα and IFN-y production as well as increased apoptosis [193]. Mechanistically, it was determined that NEAT1 sponging of miR-­ 146b allowed increased TRAF6 expression and activation of the NF-kB signalling resulting in accelerated cell injury in human renal mesangial cells.

inhibits XIST function by complementary binding of XIST forming a double-stranded RNA complex which is targeted for degradation by the endoribonuclease Dicer. Thus, upregulation of TSIX could be therapeutically protective against the Xi skewing reported in SLE and in tackling cartilage degradation and inflammation in OA as previously described. Intriguingly, the expression levels of TSIX has also been reported to be ­significantly higher in SLE patients compared to healthy donors and found to be highly expressed in female SLE patients compared with males which may be a protective response against elevated XIST [174]. Although the ratio of XIST to TSIX expression levels in SLE has not been determined. As such endogenous TSIX levels may not be sufficient to reverse the effects of XIST which is also known to act locally to repress TSIX on both inactive and active X-chromosomes [200].

4.2.1.4 X-Inactive Specific Transcript (XIST) There is considerable evidence for the role of XIST in the pathogenesis of SLE.  Sex bias strongly drives risk of SLE, with nine times as many woman developing the autoimmune condition [194]. In SLE female patient lymphocytes, XIST localisation patterns are disrupted and the inactive X chromosome becomes partially reactivated leading to the over expression of immunity related genes [195]. In the NZB/W F1 SLE mouse model with female bias, YY1 expression was reduced resulting in poor localisation of XIST lncRNA to the Xi and increased expression of immune regulatory factors TLR7 and CXCR3  in B-cells [196]. Similar disruptions to X-chromosome maintenance is also reported in SLE patient T-cells [197]. Additionally, skewed allelic expression of X-linked genes has also been attributed to high variability of DNA methylation levels in SLE patients, which has been reversed in SLE mouse models by XIST knockdown [198]. Finally, TSIX is the XIST antisense lncRNA which protects the active X chromosome from silencing during X-inactivation of the second X chromosome in females [199]. TSIX

4.2.1.5 Taurine Up-Regulated 1 (TUG1) TUG1 expression is significantly reduced in SLE patient whole blood and may be a clinically relevant biomarker [201]. Xu et al. [201] determined the protective effects of TUG1  in HK-2 renal tubular epithelial cells, to understand lupus nephritis in SLE patients. Overexpression of TUG1 targeted the miR-223/SIRT1 axis activating the PI3K/AKT signalling whilst suppressing NF-kB pathway, increasing cell viability and supressing inflammation [202]. With SLE mice, inhibition of the NF-kB signalling pathway with PDTC drug mitigated SLE progression and resulted in the up-regulation of TUG1 lncRNA expression [203]. 4.2.1.6 Urothelial Carcinoma-­ Associated 1 (UCA1) UCA1 levels in SLE patient plasma was significantly increased along with AKT, particularly in females [204]. Jiang and Li found high UCA1 expression correlated with those patients with evidence of organ involvement suggesting UCA1 could be a biomarker for stratifying SLE patients to distinguish those with and without organ involvement. Gain of function investigations

4  Long Non-coding RNAs in Rheumatology

found that UCA1 overexpression increased cell proliferation through activation of the PI3K/AKT pathway [204].

4.2.1.7 TNF and HNRNPL Related Immunoregulatory LncRNA (THRIL) THRIL expression is elevated in SLE patients and preclinical models. THRIL overexpression in LPS-induced HK2, a SLE model, increased apoptosis and the expression of pro-­inflammatory cytokines IL-1B, IL-6, IL-8 and TNFA. THRIL was identified as a ceRNA of miR-34a which targeted MCP-1, thus THRIL activated the JNK and Wnt/β-catenin signalling pathways which may be crucial in SLE pathogenesis [205].

4.3

Conclusions and Perspectives

The evidence of lncRNA mediated roles in rheumatic conditions has been mounting in recent years and researchers are finally uncovering the diagnostic and therapeutic value of lncRNAs. Numerous lncRNAs have now been identified as central regulators of inflammatory pathways that are relevant to chronic inflammatory rheumatological conditions. This chapter illustrates the diverse role of lncRNAs in regulating inflammation, proliferation, migration, invasion and apoptosis in RA, OA and SLE. Unsurprisingly, since inflammatory diseases share several common pathways, studies have identified lncRNAs that are dysregulated across all three conditions. Although there are still gaps in our knowledge, lncRNA functional characterisation has been best explored in RA and OA and to a lesser extent in SLE, where lncRNAs are still a nascent field. However as inflammatory pathways are shared between conditions it is likely that there will be shared lncRNA functionality amongst respective conditions. These findings will not only add to our understanding of the dysregulation in chronic disease and the involvement of commonly dysregulated pathways, but will also be insightful in identifying therapeutic interventions and at-risk patient populations across these rheumatological conditions.

61

References 1. Guo Q, Wang Y, Xu D, Nossent J, Pavlos NJ, Xu J (2018) Rheumatoid arthritis: pathological mechanisms and modern pharmacologic therapies. Bone Res 6:15. https://doi.org/10.1038/ s41413-­018-­0016-­9 2. Crowson CS, Matteson EL, Myasoedova E et  al (2011) The lifetime risk of adult-onset rheumatoid arthritis and other inflammatory autoimmune rheumatic diseases. Arthritis Rheum 63(3):633–639. https://doi.org/10.1002/art.30155 3. Xu B, Lin J (2017) Characteristics and risk factors of rheumatoid arthritis in the United States: an NHANES analysis. Peer J 5:e4035. https://doi. org/10.7717/peerj.4035 4. Fukui S, Iwamoto N, Takatani A et  al (2017) M1 and M2 monocytes in rheumatoid arthritis: a contribution of imbalance of M1/M2 monocytes to Osteoclastogenesis. Front Immunol 8:1958. https:// doi.org/10.3389/fimmu.2017.01958 5. Kung CC, Dai SP, Chiang H, Huang HS, Sun WH (2020) Temporal expression patterns of distinct cytokines and M1/M2 macrophage polarization regulate rheumatoid arthritis progression. Mol Biol Rep 47(5):3423–3437. https://doi.org/10.1007/ s11033-­020-­05422-­6 6. Pap T, Korb-Pap A (2015) Cartilage damage in osteoarthritis and rheumatoid arthritis--two unequal siblings. Nat Rev Rheumatol 11(10):606–615. https://doi.org/10.1038/nrrheum.2015.95 7. Ostrowska M, Maśliński W, Prochorec-Sobieszek M, Nieciecki M, Sudoł-Szopińska I (2018) Cartilage and bone damage in rheumatoid arthritis. Reumatologia 56(2):111–120. https://doi. org/10.5114/reum.2018.75523 8. Sudoł-Szopińska I, Kontny E, Maśliński W, Prochorec-Sobieszek M, Warczyńska A, Kwiatkowska B (2013) Significance of bone marrow edema in pathogenesis of rheumatoid arthritis. Pol J Radiol 78(1):57–63. https://doi.org/10.12659/ PJR.883768 9. Malemud CJ (2013) Intracellular signaling pathways in rheumatoid arthritis. J Clin Cell Immunol 4:160. https://doi.org/10.4172/2155-­9899.1000160 10. Madhok R, Crilly A, Watson J, Capell HA (1993) Serum interleukin 6 levels in rheumatoid arthritis: correlations with clinical and laboratory indices of disease activity. Ann Rheum Dis 52(3):232–234. https://doi.org/10.1136/ard.52.3.232 11. Noack M, Miossec P (2017) Selected cytokine pathways in rheumatoid arthritis. Semin Immunopathol 39(4):365–383. https://doi.org/10.1007/ s00281-­017-­0619-­z 12. Arleevskaya MI, Larionova RV, Brooks WH, Bettacchioli E, Renaudineau Y (2020) Toll-like receptors, infections, and rheumatoid arthritis. Clin Rev Allergy Immunol 58(2):172–181. https://doi. org/10.1007/s12016-­019-­08742-­z 13. Kloppenburg M, Berenbaum F (2020) Osteoarthritis year in review 2019: epidemiology and therapy.

62 Osteoarthr Cartil 28(3):242–248. https://doi. org/10.1016/j.joca.2020.01.002 14. Hootman JM, Helmick CG, Brady TJ (2012) A public health approach to addressing arthritis in older adults: the most common cause of disability. Am J Public Health 102(3):426–433. https://doi. org/10.2105/AJPH.2011.300423 15. Philp AM, Collier RL, Grover LM, Davis ET, Jones SW (2017) Resistin promotes the abnormal type I collagen phenotype of subchondral bone in obese patients with end stage hip osteoarthritis. Sci Rep 7(1):4042. https://doi.org/10.1038/ s41598-­017-­04119-­4 16. Pearson MJ, Herndler-Brandstetter D, Tariq MA et al (2017) IL-6 secretion in osteoarthritis patients is mediated by chondrocyte-synovial fibroblast cross-­ talk and is enhanced by obesity. Sci Rep 7(1):3451. https://doi.org/10.1038/s41598-­017-­03759-­w 17. Safiri S, Kolahi AA, Smith E et  al (2020) Global, regional and national burden of osteoarthritis 1990-­ 2017: a systematic analysis of the Global Burden of Disease Study 2017. Ann Rheum Dis 79(6):819–828. https://doi.org/10.1136/annrheumdis-­2019-­216515 18. Palazzo C, Nguyen C, Lefevre-Colau M-M, Rannou F, Poiraudeau S (2016) Risk factors and burden of osteoarthritis. Ann Phys Rehabil Med 59(3):134– 138. https://doi.org/10.1016/j.rehab.2016.01.006 19. Nanus DE, Wijesinghe SN, Pearson MJ et al (2020) Regulation of the inflammatory synovial fibroblast phenotype by metastasis-associated lung adenocarcinoma transcript 1 long noncoding RNA in obese patients with osteoarthritis. Arthritis Rheumatol 72(4):609–619. https://doi.org/10.1002/art.41158 20. Tonge DP, Pearson MJ, Jones SW (2014) The hallmarks of osteoarthritis and the potential to develop personalised disease-modifying pharmacological therapeutics. Osteoarthr Cartil 22(5):609–621. https://doi.org/10.1016/j.joca.2014.03.004 21. Jones SW, Brockbank SM, Clements KM et  al (2009) Mitogen-activated protein kinase-activated protein kinase 2 (MK2) modulates key biological pathways associated with OA disease pathology. Osteoarthr Cartil 17(1):124–131. https://doi. org/10.1016/j.joca.2008.05.001 22. Cooke ME, Lawless BM, Jones SW, Grover LM (2018) Matrix degradation in osteoarthritis primes the superficial region of cartilage for mechanical damage. Acta Biomater 78:320–328. https://doi. org/10.1016/j.actbio.2018.07.037 23. Chang J, Jackson SG, Wardale J, Jones SW (2014) Hypoxia modulates the phenotype of osteoblasts isolated from knee osteoarthritis patients, leading to undermineralized bone nodule formation. Arthritis Rheumatol 66(7):1789–1799. https://doi. org/10.1002/art.38403 24. Hügle T, Geurts J (2017) What drives osteoarthritis?synovial versus subchondral bone pathology. Rheumatology (Oxford) 56(9):1461–1471. https:// doi.org/10.1093/rheumatology/kew389

S. N. Wijesinghe et al. 25. Philp AM, Davis ET, Jones SW (2017) Developing anti-inflammatory therapeutics for patients with osteoarthritis. Rheumatology (Oxford) 56(6):869– 881. https://doi.org/10.1093/rheumatology/kew278 26. Nanus DE, Badoume A, Wijesinghe SN et  al (2020) Identification of synovial fibroblasts subsets ­associated with pain and progression of knee osteoarthritis by single cell sequencing. Osteoarthr Cartil 28:S133. https://doi.org/10.1016/j.joca.2020.02.220 27. Wijesinghe SN, Badoume A, Nanus DE, Davis ET, Lindsay MA, Jones SW (2020) Obese synovial fibroblasts exhibit single cell subsets with specific pathological inflammatory functions in osteoarthritis patients. Osteoarthr Cartil 28:S84. https://doi. org/10.1016/j.joca.2020.02.128 28. Pearson MJ, Jones SW (2016) Review: long noncoding RNAs in the regulation of inflammatory pathways in rheumatoid arthritis and osteoarthritis. Arthritis Rheumatol 68(11):2575–2583. https://doi. org/10.1002/art.39759 29. Saito T, Tanaka S (2017) Molecular mechanisms underlying osteoarthritis development: Notch and NF-κB.  Arthritis Res Ther 19(1):94. https://doi. org/10.1186/s13075-­017-­1296-­y 30. Olivotto E, Otero M, Marcu KB, Goldring MB (2015) Pathophysiology of osteoarthritis: canonical NF-κB/ IKKβ-dependent and kinase-independent effects of IKKα in cartilage degradation and chondrocyte differentiation. RMD Open 1(Suppl 1):e000061. https://doi.org/10.1136/rmdopen-­2015-­000061 31. Ahmed AS, Gedin P, Hugo A et al (2018) Activation of NF-κB in synovium versus cartilage from patients with advanced knee osteoarthritis: a potential contributor to inflammatory aspects of disease progression. J Immunol 201(7):1918. https://doi. org/10.4049/jimmunol.1800486 32. Barreto G, Manninen M, Eklund K (2020) Osteoarthritis and Toll-Like receptors: when innate immunity meets chondrocyte apoptosis. Biology (Basel) 9(4). https://doi.org/10.3390/ biology9040065 33. Bar-Or D, Rael LT, Thomas GW, Brody EN (2015) Inflammatory pathways in knee osteoarthritis: potential targets for treatment. Curr Rheumatol Rev 11(1):50–58. https://doi.org/10.2174/157339711166 6150522094131 34. Hsu CC, Lin CL, Jou IM, Wang PH, Lee JS (2017) The protective role of nitric oxide-dependent innate immunosuppression in the early stage of cartilage damage in rats: role of nitric oxide in cartilage damage. Bone Joint Res 6(4):253–258. https://doi. org/10.1302/2046-­3758.64.BJJ-­2016-­0161.R1 35. Huang H, Zheng J, Shen N et al (2018) Identification of pathways and genes associated with synovitis in osteoarthritis using bioinformatics analyses. Sci Rep 8(1):10050. https://doi.org/10.1038/ s41598-­018-­28280-­6 36. Azamar-Llamas D, Hernández-Molina G, Ramos-­ Ávalos B, Furuzawa-Carballeda J (2017) Adipokine contribution to the pathogenesis of osteoarthri-

4  Long Non-coding RNAs in Rheumatology tis. Mediat Inflamm 2017:5468023. https://doi. org/10.1155/2017/5468023 37. Ji P, Diederichs S, Wang W et al (2003) MALAT-1, a novel noncoding RNA, and thymosin beta4 predict metastasis and survival in early-stage non-small cell lung cancer. Oncogene 22(39):8031–8041. https:// doi.org/10.1038/sj.onc.1206928 38. Tripathi V, Ellis JD, Shen Z et al (2010) The nuclear-­ retained noncoding RNA MALAT1 regulates alternative splicing by modulating SR splicing factor phosphorylation. Mol Cell 39(6):925–938. https:// doi.org/10.1016/j.molcel.2010.08.011 39. Li ZX, Zhu QN, Zhang HB, Hu Y, Wang G, Zhu YS (2018) MALAT1: a potential biomarker in cancer. Cancer Manag Res 10:6757–6768. https://doi. org/10.2147/CMAR.S169406 40. Pan L, Liu D, Zhao L, Wang L, Xin M, Li X (2018) Long noncoding RNA MALAT1 alleviates lipopolysaccharide-­induced inflammatory injury by upregulating microRNA-19b in murine chondrogenic ATDC5 cells. J Cell Biochem 119(12):10165– 10175. https://doi.org/10.1002/jcb.27357 41. Gao GC, Cheng XG, Wei QQ, Chen WC, Huang WZ (2019) Long noncoding RNA MALAT-1 inhibits apoptosis and matrix metabolism disorder in interleukin-1β-induced inflammation in articular chondrocytes via the JNK signaling pathway. J Cell Biochem 120(10):17167–17179. https://doi. org/10.1002/jcb.28977 42. Arun G, Aggarwal D, Spector DL (2020) Long non-­ coding RNA: functional implications. Noncoding RNA 6(2). https://doi.org/10.3390/ncrna6020022 43. Liang J, Xu L, Zhou F et al (2018) MALAT1/miR-­ 127-­5p regulates Osteopontin (OPN)-mediated proliferation of human chondrocytes through PI3K/Akt pathway. J Cell Biochem 119(1):431–439. https:// doi.org/10.1002/jcb.26200 44. Zhang Y, Wang F, Chen G, He R, Yang L (2019) LncRNA MALAT1 promotes osteoarthritis by modulating miR-150-5p/AKT3 axis. Cell Biosci 9:54. https://doi.org/10.1186/s13578-­019-­0302-­2 45. Liu C, Ren S, Zhao S, Wang Y (2019) LncRNA MALAT1/MiR-145 adjusts IL-1β-induced chondrocytes viability and cartilage matrix degradation by regulating ADAMTS5  in human osteoarthritis. Yonsei Med J 60(11):1081–1092. https://doi. org/10.3349/ymj.2019.60.11.1081 46. Li H, Xie S, Zhang R, Zhang H (2020) LncRNA MALAT1 mediates proliferation of LPS treated-­ articular chondrocytes by targeting the miR-146a-­ PI3K/Akt/mTOR axis. Life Sci 254:116801. https:// doi.org/10.1016/j.lfs.2019.116801 47. Song J, Kim D, Han J, Kim Y, Lee M, Jin EJ (2015) PBMC and exosome-derived Hotair is a critical regulator and potent marker for rheumatoid arthritis. Clin Exp Med 15(1):121–126. https://doi.org/10.1007/ s10238-­013-­0271-­4 48. Li GQ, Fang YX, Liu Y et  al (2019) MALAT1driven inhibition of Wnt signal impedes proliferation and inflammation in fibroblast-like Synoviocytes

63 through CTNNB1 promoter methylation in rheumatoid arthritis. Hum Gene Ther 30(8):1008–1022. https://doi.org/10.1089/hum.2018.212 49. Wan L, Liu J, Huang C et al (2020) Decreased long-­ chain non-coding RNA MALAT1 expression and increased hsa-miR155-3p expression involved in Notch signaling pathway regulation in rheumatoid arthritis patients. Xi Bao Yu Fen Zi Mian Yi Xue Za Zhi 36(6):535–541 50. Pan F, Zhu L, Lv H, Pei C (2016) Quercetin promotes the apoptosis of fibroblast-like synoviocytes in rheumatoid arthritis by upregulating lncRNA MALAT1. Int J Mol Med 38(5):1507–1514. https:// doi.org/10.3892/ijmm.2016.2755 51. Rinn JL, Kertesz M, Wang JK et al (2007) Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 129(7):1311–1323. https://doi.org/10.1016/j. cell.2007.05.022 52. Rinn JL, Kertesz M, Wang JK et al (2007) Functional demarcation of active and silent chromatin domains in human HOX loci by Noncoding RNAs. Cell 129(7):1311–1323. https://doi.org/10.1016/j. cell.2007.05.022 53. Bond CS, Fox AH (2009) Paraspeckles: nuclear bodies built on long noncoding RNA.  J Cell Biol 186(5):637–644. https://doi.org/10.1083/ jcb.200906113 54. Imamura K, Imamachi N, Akizuki G et  al (2014) Long noncoding RNA NEAT1-dependent SFPQ relocation from promoter region to paraspeckle mediates IL8 expression upon immune stimuli. Mol Cell 53(3):393–406. https://doi.org/10.1016/j. molcel.2014.01.009 55. Chen H, Qi J, Bi Q, Zhang S (2017) Expression profile of long noncoding RNA (HOTAIR) and its predicted target miR-17-3p in LPS-induced inflammatory injury in human articular chondrocyte C28/ I2 cells. Int J Clin Exp Pathol 10(9):9146–9157 56. Hu J, Wang Z, Shan Y, Pan Y, Ma J, Jia L (2018) Long non-coding RNA HOTAIR promotes osteoarthritis progression via miR-17-5p/FUT2/β-catenin axis. Cell Death Dis 9(7):711. https://doi.org/10.1038/ s41419-­018-­0746-­z 57. Yang Y, Xing D, Wang Y, Jia H, Li B, Li JJ (2020) A long non-coding RNA, HOTAIR, promotes cartilage degradation in osteoarthritis by inhibiting WIF-1 expression and activating Wnt pathway. BMC Mol Cell Biol 21(1):53. https://doi.org/10.1186/ s12860-­020-­00299-­6 58. He B, Jiang D (2020) HOTAIR-induced apoptosis is mediated by sponging miR-130a-3p to repress chondrocyte autophagy in knee osteoarthritis. Cell Biol Int 44(2):524–535. https://doi.org/10.1002/ cbin.11253 59. Chen Y, Zhang L, Li E et  al (2020) Long-chain non-coding RNA HOTAIR promotes the progression of osteoarthritis via sponging miR-20b/PTEN axis. Life Sci 253:117685. https://doi.org/10.1016/j. lfs.2020.117685

64 60. Dou P, Hu R, Zhu W et  al (2017) Long noncoding RNA HOTAIR promotes expression of ADAMTS-5 in human osteoarthritic articular chondrocytes. Pharmazie 72(2):113–117. https://doi. org/10.1691/ph.2017.6649 61. Zhang C, Wang P, Jiang P et al (2016) Upregulation of lncRNA HOTAIR contributes to IL-1β-induced MMP overexpression and chondrocytes apoptosis in temporomandibular joint osteoarthritis. Gene 586(2):248–253. https://doi.org/10.1016/j. gene.2016.04.016 62. Mao T, He C, Wu H, Yang B, Li X (2019) Silencing lncRNA HOTAIR declines synovial inflammation and synoviocyte proliferation and promotes synoviocyte apoptosis in osteoarthritis rats by inhibiting Wnt/β-catenin signaling pathway. Cell Cycle 18(22):3189–3205. https://doi.org/10.1080/153841 01.2019.1671716 63. Zhang HJ, Wei QF, Wang SJ et al (2017) LncRNA HOTAIR alleviates rheumatoid arthritis by targeting miR-138 and inactivating NF-κB pathway. Int Immunopharmacol 50:283–290. https://doi. org/10.1016/j.intimp.2017.06.021 64. Ji J, Dai X, Yeung SJ, He X (2019) The role of long non-coding RNA GAS5  in cancers. Cancer Manag Res 11:2729–2737. https://doi.org/10.2147/CMAR. S189052 65. Kino T, Hurt DE, Ichijo T, Nader N, Chrousos GP (2010) Noncoding RNA gas5 is a growth arrest- and starvation-associated repressor of the glucocorticoid receptor. Sci Signal 3(107):ra8. https://doi. org/10.1126/scisignal.2000568 66. Xing D, Liang JQ, Li Y et al (2014) Identification of long noncoding RNA associated with osteoarthritis in humans. Orthop Surg 6(4):288–293. https://doi. org/10.1111/os.12147 67. Song J, Ahn C, Chun CH, Jin EJ (2014) A long non-coding RNA, GAS5, plays a critical role in the regulation of miR-21 during osteoarthritis. J Orthop Res 32(12):1628–1635. https://doi.org/10.1002/ jor.22718 68. Ji Q, Qiao X, Liu Y, Wang D, Yan J (2020) Silencing of long-chain non-coding RNA GAS5  in osteoarthritic chondrocytes is mediated by targeting the miR-34a/Bcl-2 axis. Mol Med Rep 21(3):1310– 1319. https://doi.org/10.3892/mmr.2019.10900 69. Li F, Sun J, Huang S, Su G, Pi G (2018) LncRNA GAS5 overexpression reverses LPS-induced inflammatory injury and apoptosis through up-regulating KLF2 expression in ATDC5 chondrocytes. Cell Physiol Biochem 45(3):1241–1251. https://doi. org/10.1159/000487455 70. Li G, Liu Y, Meng F et al (2018) Tanshinone IIA promotes the apoptosis of fibroblast-like synoviocytes in rheumatoid arthritis by up-regulating lncRNA GAS5. Biosci Rep 38(5). https://doi.org/10.1042/ BSR20180626 71. Ma C, Wang W, Li P (2019) LncRNA GAS5 overexpression downregulates IL-18 and induces the apoptosis of fibroblast-like synoviocytes. Clin Rheumatol

S. N. Wijesinghe et al. 38(11):3275–3280. https://doi.org/10.1007/ s10067-­019-­04691-­2 72. Li M, Wang N, Shen Z, Yan J (2020) Long non-­ coding RNA growth arrest-specific transcript 5 regulates rheumatoid arthritis by targeting homeodomain-­interacting protein kinase 2. Clin Exp Rheumatol 38:1145 73. Ghafouri-Fard S, Esmaeili M, Taheri M (2020) H19 lncRNA: roles in tumorigenesis. Biomed Pharmacother 123:109774. https://doi.org/10.1016/j. biopha.2019.109774 74. Steck E, Boeuf S, Gabler J et al (2012) Regulation of H19 and its encoded microRNA-675 in osteoarthritis and under anabolic and catabolic in vitro conditions. J Mol Med (Berl) 90(10):1185–1195. https://doi. org/10.1007/s00109-­012-­0895-­y 75. Hu Y, Li S, Zou Y (2019) Knockdown of LncRNA H19 relieves LPS-induced damage by modulating miR-130a in osteoarthritis. Yonsei Med J 60(4):381– 388. https://doi.org/10.3349/ymj.2019.60.4.381 76. Zhang X, Liu X, Ni X, Feng P, Wang YU (2019) Long non-coding RNA H19 modulates proliferation and apoptosis in osteoarthritis via regulating miR-­ 106a-­5p. J Biosci 44(6):128 77. Yang B, Xu L, Wang S (2020) Regulation of lncRNA-H19/miR-140-5p in cartilage matrix degradation and calcification in osteoarthritis. Ann Palliat Med 9(4):1896–1904. https://doi.org/10.21037/ apm-­20-­929 78. Tan F, Wang D, Yuan Z (2020) The fibroblastlike Synoviocyte derived Exosomal long non-­ coding RNA H19 alleviates osteoarthritis progression through the miR-106b-5p/TIMP2 Axis. Inflammation 43(4):1498–1509. https://doi. org/10.1007/s10753-­020-­01227-­8 79. Stuhlmüller B, Kunisch E, Franz J et  al (2003) Detection of oncofetal h19 RNA in rheumatoid arthritis synovial tissue. Am J Pathol 163(3):901–911. https://doi.org/10.1016/S0002-­9440(10)63450-­5 80. Yang J, Li Y, Wang L, Zhang Z, Li Z, Jia Q (2020) LncRNA H19 aggravates TNF-α-induced inflammatory injury via TAK1 pathway in MH7A cells. Biofactors. https://doi.org/10.1002/biof.1659 81. Liu F, Liu X, Yang Y et  al (2020) NEAT1/miR193a-­ 3p/SOX5 axis regulates cartilage matrix degradation in human osteoarthritis. Cell Biol Int 44(4):947–957. https://doi.org/10.1002/cbin.11291 82. Tu Y, Ma T, Wen T et al (2020) MicroRNA-377-3p alleviates IL-1β-caused chondrocyte apoptosis and cartilage degradation in osteoarthritis in part by downregulating ITGA6. Biochem Biophys Res Commun 523(1):46–53. https://doi.org/10.1016/j. bbrc.2019.11.186 83. Li D, Sun Y, Wan Y, Wu X, Yang W (2020) LncRNA NEAT1 promotes proliferation of chondrocytes via down-regulation of miR-16-5p in osteoarthritis. J Gene Med:e3203. https://doi.org/10.1002/jgm.3203 84. Wang Z, Hao J, Chen D (2019) Long noncoding RNA Nuclear enriched abundant transcript 1 (NEAT1) regulates proliferation, apoptosis,

4  Long Non-coding RNAs in Rheumatology and inflammation of chondrocytes via the miR-­ 181a/Glycerol-3-phosphate dehydrogenase 1-like (GPD1L) Axis. Med Sci Monit 25:8084–8094. https://doi.org/10.12659/MSM.918416 85. Shui X, Chen S, Lin J, Kong J, Zhou C, Wu J (2019) Knockdown of lncRNA NEAT1 inhibits Th17/CD4. J Cell Physiol 234(12):22477–22484. https://doi. org/10.1002/jcp.28811 86. Borsani G, Tonlorenzi R, Simmler MC et al (1991) Characterization of a murine gene expressed from the inactive X chromosome. Nature 351(6324):325– 329. https://doi.org/10.1038/351325a0 87. Plath K, Mlynarczyk-Evans S, Nusinow DA, Panning B (2002) Xist RNA and the mechanism of X chromosome inactivation. Annu Rev Genet 36:233–278. https://doi.org/10.1146/annurev. genet.36.042902.092433 88. Marshall EA, Stewart GL, Sage AP, Lam WL, Brown CJ (2019) Beyond sequence homology: cellular biology limits the potential of XIST to act as a miRNA sponge. PLoS One 14(8):e0221371. https:// doi.org/10.1371/journal.pone.0221371 89. Lambert NC (2019) Nonendocrine mechanisms of sex bias in rheumatic diseases. Nat Rev Rheumatol 15(11):673–686. https://doi.org/10.1038/ s41584-­019-­0307-­6 90. Kanaan SB, Onat OE, Balandraud N et  al (2016) Evaluation of X chromosome inactivation with respect to HLA genetic susceptibility in rheumatoid arthritis and systemic sclerosis. PLoS One 11(6):e0158550. https://doi.org/10.1371/journal. pone.0158550 91. Lin J, He Y, Chen J, Zeng Z, Yang B, Ou Q (2017) A critical role of transcription factor YY1  in rheumatoid arthritis by regulation of interleukin-6. J Autoimmun 77:67–75. https://doi.org/10.1016/j. jaut.2016.10.008 92. Srikanth VK, Fryer JL, Zhai G, Winzenberg TM, Hosmer D, Jones G (2005) A meta-analysis of sex differences prevalence, incidence and severity of osteoarthritis. Osteoarthr Cartil 13(9):769–781. https://doi.org/10.1016/j.joca.2005.04.014 93. Hame SL, Alexander RA (2013) Knee osteoarthritis in women. Curr Rev Musculoskelet Med 6(2):182– 187. https://doi.org/10.1007/s12178-­013-­9164-­0 94. Xiang S, Li Z, Bian Y, Weng X (2019) Identification of changed expression of mRNAs and lncRNAs in osteoarthritic synovium by RNA-sequencing. Gene 685:55–61. https://doi.org/10.1016/j. gene.2018.10.076 95. Li L, Lv G, Wang B, Kuang L (2018) The role of lncRNA XIST/miR-211 axis in modulating the proliferation and apoptosis of osteoarthritis chondrocytes through CXCR4 and MAPK signaling. Biochem Biophys Res Commun 503(4):2555–2562. https://doi.org/10.1016/j.bbrc.2018.07.015 96. Sun P, Wu Y, Li X, Jia Y (2020) miR-142-5p protects against osteoarthritis through competing with lncRNA XIST. J Gene Med 22(4):e3158. https://doi. org/10.1002/jgm.3158

65 97. Lian LP, Xi XY (2020) Long non-coding RNA XIST protects chondrocytes ATDC5 and CHON-001 from IL-1β-induced injury via regulating m ­ iR-653-5p/ SIRT1 axis. J Biol Regul Homeost Agents 34(2):379– 391. https://doi.org/10.23812/19-­549-­A-­65 98. Wang T, Liu Y, Wang Y, Huang X, Zhao W, Zhao Z (2019) Long non-coding RNA XIST promotes extracellular matrix degradation by functioning as a competing endogenous RNA of miR-1277-5p in osteoarthritis. Int J Mol Med 44(2):630–642. https:// doi.org/10.3892/ijmm.2019.4240 99. Liu Y, Liu K, Tang C, Shi Z, Jing K, Zheng J (2020) Long non-coding RNA XIST contributes to osteoarthritis progression via miR-149-5p/DNMT3A axis. Biomed Pharmacother 128:110349. https://doi. org/10.1016/j.biopha.2020.110349 100. Chen H, Yang S, Shao R (2019) Long non-coding XIST raises methylation of TIMP-3 promoter to regulate collagen degradation in osteoarthritic chondrocytes after tibial plateau fracture. Arthritis Res Ther 21(1):271. https://doi.org/10.1186/ s13075-­019-­2033-­5 101. Li L, Lv G, Wang B, Kuang L (2020) XIST/miR376c-­5p/OPN axis modulates the influence of proinflammatory M1 macrophages on osteoarthritis chondrocyte apoptosis. J Cell Physiol 235(1):281– 293. https://doi.org/10.1002/jcp.28968 102. Ghafouri-Fard S, Taheri M (2019) Maternally expressed gene 3 (MEG3): a tumor suppressor long non coding RNA. Biomed Pharmacother 118:109129. https://doi.org/10.1016/j.biopha.2019.109129 103. Mondal T, Subhash S, Vaid R et  al (2015) MEG3 long noncoding RNA regulates the TGF-β pathway genes through formation of RNA-DNA triplex structures. Nat Commun 6:7743. https://doi.org/10.1038/ ncomms8743 104. Su W, Xie W, Shang Q, Su B (2015) The long noncoding RNA MEG3 is downregulated and inversely associated with VEGF levels in osteoarthritis. Biomed Res Int 2015:356893. https://doi. org/10.1155/2015/356893 105. Xu J, Xu Y (2017) The lncRNA MEG3 downregulation leads to osteoarthritis progression via miR-16/SMAD7 axis. Cell Biosci 7:69. https://doi. org/10.1186/s13578-­017-­0195-­x 106. Chen K, Zhu H, Zheng MQ, Dong QR (2019) LncRNA MEG3 inhibits the degradation of the extracellular matrix of chondrocytes in osteoarthritis via targeting miR-93/TGFBR2 Axis. Cartilage 1947603519855759. https://doi. org/10.1177/1947603519855759 107. Wang Z, Chi X, Liu L et  al (2018) Long noncoding RNA maternally expressed gene 3 knockdown alleviates lipopolysaccharide-induced inflammatory injury by up-regulation of miR-203 in ATDC5 cells. Biomed Pharmacother 100:240–249. https://doi. org/10.1016/j.biopha.2018.02.018 108. Li X, Tang C, Wang J et al (2018) Methylene blue relieves the development of osteoarthritis by upregu-

66 lating lncRNA MEG3. Exp Ther Med 15(4):3856– 3864. https://doi.org/10.3892/etm.2018.5918 109. Wang A, Hu N, Zhang Y et  al (2019) MEG3 promotes proliferation and inhibits apoptosis in osteoarthritis chondrocytes by miR-361-5p/FOXO1 axis. BMC Med Genet 12(1):201. https://doi.org/10.1186/ s12920-­019-­0649-­6 110. Li G, Liu Y, Meng F et  al (2019) LncRNA MEG3 inhibits rheumatoid arthritis through miR-141 and inactivation of AKT/mTOR signalling pathway. J Cell Mol Med 23(10):7116–7120. https://doi. org/10.1111/jcmm.14591 111. Lu X, Qian J (2019) Downregulated MEG3 participates in rheumatoid arthritis via promoting proliferation of fibroblast-like synoviocytes. Exp Ther Med 17(3):1637–1642. https://doi.org/10.3892/ etm.2018.7100 112. Wang KC, Yang YW, Liu B et  al (2011) A long noncoding RNA maintains active chromatin to coordinate homeotic gene expression. Nature 472(7341):120–124. https://doi.org/10.1038/ nature09819 113. Shang A, Wang W, Gu C et  al (2019) Long noncoding RNA HOTTIP enhances IL-6 expression to potentiate immune escape of ovarian cancer cells by upregulating the expression of PD-L1 in neutrophils. J Exp Clin Cancer Res 38(1):411. https://doi. org/10.1186/s13046-­019-­1394-­6 114. Ghafouri-Fard S, Dashti S, Taheri M (2020) The HOTTIP (HOXA transcript at the distal tip) lncRNA: review of oncogenic roles in human. Biomed Pharmacother 127:110158. https://doi. org/10.1016/j.biopha.2020.110158 115. Kim D, Song J, Han J, Kim Y, Chun CH, Jin EJ (2013) Two non-coding RNAs, MicroRNA-101 and HOTTIP contribute cartilage integrity by epigenetic and homeotic regulation of integrin-α1. Cell Signal 25(12):2878–2887. https://doi.org/10.1016/j. cellsig.2013.08.034 116. Mao G, Kang Y, Lin R et al (2019) Long non-coding RNA HOTTIP promotes CCL3 expression and induces cartilage degradation by sponging miR-­ 455-­ 3p. Front Cell Dev Biol 7:161. https://doi. org/10.3389/fcell.2019.00161 117. Hu X, Tang J, Bao P et al (2020) Silencing of long non-coding RNA HOTTIP reduces inflammation in rheumatoid arthritis by demethylation of SFRP1. Mol Ther Nucleic Acids 19:468–481. https://doi. org/10.1016/j.omtn.2019.11.015 118. Onagoruwa OT, Pal G, Ochu C, Ogunwobi OO (2020) Oncogenic role of PVT1 and therapeutic implications. Front Oncol 10:17. https://doi. org/10.3389/fonc.2020.00017 119. Li Y, Li S, Luo Y, Liu Y, Yu N (2017) LncRNA PVT1 regulates chondrocyte apoptosis in osteoarthritis by acting as a sponge for miR-488-3p. DNA Cell Biol 36(7):571–580. https://doi.org/10.1089/ dna.2017.3678 120. Zhao Y, Zhao J, Guo X, She J, Liu Y (2018) Long non-coding RNA PVT1, a molecular sponge for

S. N. Wijesinghe et al. miR-149, contributes aberrant metabolic dysfunction and inflammation in IL-1β-simulated osteoarthritic chondrocytes. Biosci Rep 38(5). https://doi. org/10.1042/BSR20180576 121. Lu X, Yu Y, Yin F et  al (2020) Knockdown of PVT1 inhibits IL-1β-induced injury in chondrocytes by regulating miR-27b-3p/TRAF3 axis. Int Immunopharmacol 79:106052. https://doi. org/10.1016/j.intimp.2019.106052 122. Ding LB, Li Y, Liu GY et al (2020) Long non-coding RNA PVT1, a molecular sponge of miR-26b, is involved in the progression of hyperglycemia-­ induced collagen degradation in human chondrocytes by targeting CTGF/TGF. Innate Immun 26(3):204– 214. https://doi.org/10.1177/1753425919881778 123. Xu K, Meng Z, Xian XM et  al (2020) LncRNA PVT1 induces chondrocyte apoptosis through upregulation of TNF-α in synoviocytes by sponging miR-211-3p. Mol Cell Probes 52:101560. https:// doi.org/10.1016/j.mcp.2020.101560 124. Wang J, Kong X, Hu H, Shi S (2020) Knockdown of long non-coding RNA PVT1 induces apoptosis of fibroblast-like synoviocytes through modulating miR-543-dependent SCUBE2  in rheumatoid arthritis. J Orthop Surg Res 15(1):142. https://doi. org/10.1186/s13018-­020-­01641-­6 125. Zhang CW, Wu X, Liu D et al (2019) Long non-coding RNA PVT1 knockdown suppresses fibroblast-­ like synoviocyte inflammation and induces apoptosis in rheumatoid arthritis through demethylation of. J Biol Eng 13:60. https://doi.org/10.1186/ s13036-­019-­0184-­1 126. Zhou H, Sun L, Wan F (2019) Molecular mechanisms of TUG1  in the proliferation, apoptosis, migration and invasion of cancer cells. Oncol Lett 18(5):4393– 4402. https://doi.org/10.3892/ol.2019.10848 127. Tang LP, Ding JB, Liu ZH, Zhou GJ (2018) LncRNA TUG1 promotes osteoarthritis-induced degradation of chondrocyte extracellular matrix via miR-195/MMP-13 axis. Eur Rev Med Pharmacol Sci 22(24):8574–8581. https://doi.org/10.26355/ eurrev_201812_16620 128. Liang Z, Ren C (2018) Emodin attenuates apoptosis and inflammation induced by LPS through up-­ regulating lncRNA TUG1  in murine chondrogenic ATDC5 cells. Biomed Pharmacother 103:897–902. https://doi.org/10.1016/j.biopha.2018.04.085 129. Ghafouri-Fard S, Taheri M (2019) UCA1 long non-coding RNA: an update on its roles in malignant behavior of cancers. Biomed Pharmacother 120:109459. https://doi.org/10.1016/j. biopha.2019.109459 130. Wang G, Bu X, Zhang Y et  al (2017) LncRNAUCA1 enhances MMP-13 expression by inhibiting miR-204-5p in human chondrocytes. Oncotarget 8(53):91281–91290. https://doi.org/10.18632/ oncotarget.20108 131. Yan ZF, Zhao XY, Liu W, Liu XP (2018) UCA1 impacts progress of rheumatoid arthritis by inducing the apoptosis of fibroblast-like synoviocyte. Eur

4  Long Non-coding RNAs in Rheumatology Rev Med Pharmacol Sci 22(4):914–920. https://doi. org/10.26355/eurrev_201802_14370 132. Baldinu P, Cossu A, Manca A et  al (2004) Identification of a novel candidate gene, CASC2, in a region of common allelic loss at chromosome 10q26  in human endometrial cancer. Hum Mutat 23(4):318–326. https://doi.org/10.1002/humu.20015 133. Palmieri G, Paliogiannis P, Sini MC et  al (2017) Long non-coding RNA CASC2  in human cancer. Crit Rev Oncol Hematol 111:31–38. https://doi. org/10.1016/j.critrevonc.2017.01.003 134. Sun Y, Kang S, Pei S, Sang C, Huang Y (2020) MiR93-5p inhibits chondrocyte apoptosis in osteoarthritis by targeting lncRNA CASC2. BMC Musculoskelet Disord 21(1):26. https://doi. org/10.1186/s12891-­019-­3025-­y 135. Huang T, Wang J, Zhou Y, Zhao Y, Hang D, Cao Y (2019) LncRNA CASC2 is up-regulated in osteoarthritis and participates in the regulation of IL-17 expression and chondrocyte proliferation and apoptosis. Biosci Rep 39(5). https://doi.org/10.1042/ BSR20182454 136. Liu C, Guo X, Bai S, Zeng G, Wang H (2020) lncRNA CASC2 downregulation participates in rheumatoid arthritis, and CASC2 overexpression promotes the apoptosis of fibroblast-like synoviocytes by downregulating IL-17. Mol Med Rep 21(5):2131–2137. https://doi.org/10.3892/mmr.2020.11018 137. Congrains A, Kamide K, Ohishi M, Rakugi H (2013) ANRIL: molecular mechanisms and implications in human health. Int J Mol Sci 14(1):1278–1292. https://doi.org/10.3390/ijms14011278 138. Kotake Y, Nakagawa T, Kitagawa K et  al (2011) Long non-coding RNA ANRIL is required for the PRC2 recruitment to and silencing of p15(INK4B) tumor suppressor gene. Oncogene 30(16):1956– 1962. https://doi.org/10.1038/onc.2010.568 139. Kong Y, Hsieh CH, Alonso LC (2018) A lncRNA at the CDKN2A/B locus with roles in cancer and metabolic disease. Front Endocrinol (Lausanne) 9:405. https://doi.org/10.3389/fendo.2018.00405 140. Li X, Huang TL, Zhang GD, Jiang JT, Guo PY (2019) LncRNA ANRIL impacts the progress of osteoarthritis via regulating proliferation and apoptosis of osteoarthritis synoviocytes. Eur Rev Med Pharmacol Sci 23(22):9729–9737. https://doi. org/10.26355/eurrev_201911_19535 141. Zhang TP, Zhu BQ, Tao SS et  al (2019) Long non-coding RNAs genes polymorphisms and their expression levels in patients with rheumatoid arthritis. Front Immunol 10:2529. https://doi.org/10.3389/ fimmu.2019.02529 142. Ye S, Zhu S, Feng L (2020) LncRNA ANRIL/miR125a axis exhibits potential as a biomarker for disease exacerbation, severity, and inflammation in bronchial asthma. J Clin Lab Anal 34(3):e23092. https://doi.org/10.1002/jcla.23092 143. Wang X, Sun W, Shen W et  al (2016) Long noncoding RNA DILC regulates liver cancer stem cells

67 via IL-6/STAT3 axis. J Hepatol 64(6):1283–1294. https://doi.org/10.1016/j.jhep.2016.01.019 144. Huang J, Liu L, Yang J, Ding J, Xu X (2019) lncRNA DILC is downregulated in osteoarthritis and regulates IL-6 expression in chondrocytes. J Cell Biochem 120(9):16019–16024. https://doi. org/10.1002/jcb.28880 145. Wang G, Tang L, Zhang X, Li Y (2019) LncRNA DILC participates in rheumatoid arthritis by inducing apoptosis of fibroblast-like synoviocytes and down-regulating IL-6. Biosci Rep 39(5). https://doi. org/10.1042/BSR20182374 146. Xu D, Jiang Y, Yang L et al (2017) Long noncoding RNAs expression profile and functional networks in rheumatoid arthritis. Oncotarget 8(56):95280– 95292. https://doi.org/10.18632/oncotarget.20036 147. Zhang P, Sun J, Liang C et  al (2020) lncRNA IGHC.  Mediat Inflamm 2020:9743037. https://doi. org/10.1155/2020/9743037 148. Chen S, Liang H, Yang H et  al (2017) LincRNa-p21: function and mechanism in cancer. Med Oncol 34(5):98. https://doi.org/10.1007/ s12032-­017-­0959-­5 149. Tang L, Ding J, Zhou G, Liu Z (2018) LncRNA-p21 promotes chondrocyte apoptosis in osteoarthritis by acting as a sponge for miR-451. Mol Med Rep 18(6):5295–5301. https://doi.org/10.3892/ mmr.2018.9506 150. Spurlock CF, Tossberg JT, Matlock BK, Olsen NJ, Aune TM (2014) Methotrexate inhibits NF-κB activity via long intergenic (noncoding) RNA-p21 induction. Arthritis Rheumatol 66(11):2947–2957. https:// doi.org/10.1002/art.38805 151. Zimta AA, Tigu AB, Braicu C, Stefan C, Ionescu C, Berindan-Neagoe I (2020) An emerging class of long non-coding RNA with oncogenic role arises from the snoRNA host genes. Front Oncol 10:389. https://doi.org/10.3389/fonc.2020.00389 152. Thin KZ, Tu JC, Raveendran S (2019) Long noncoding SNHG1 in cancer. Clin Chim Acta 494:38– 47. https://doi.org/10.1016/j.cca.2019.03.002 153. Lei J, Fu Y, Zhuang Y, Zhang K, Lu D (2019) LncRNA SNHG1 alleviates IL-1β-induced osteoarthritis by inhibiting miR-16-5p-mediated p38 MAPK and NF-κB signaling pathways. Biosci Rep 39(9). https://doi.org/10.1042/BSR20191523 154. Li Z, Chao TC, Chang KY et  al (2014) The long noncoding RNA THRIL regulates TNFα expression through its interaction with hnRNPL.  Proc Natl Acad Sci U S A 111(3):1002–1007. https://doi. org/10.1073/pnas.1313768111 155. Liu G, Wang Y, Zhang M, Zhang Q (2019) Long noncoding RNA THRIL promotes LPS-induced inflammatory injury by down-regulating microRNA-125b in ATDC5 cells. Int Immunopharmacol 66:354–361. https://doi.org/10.1016/j.intimp.2018.11.038 156. Moharamoghli M, Hassan-Zadeh V, Dolatshahi E, Alizadeh Z, Farazmand A (2019) The expression of GAS5, THRIL, and RMRP lncRNAs is increased in T cells of patients with rheumatoid arthritis.

68 Clin Rheumatol 38(11):3073–3080. https://doi. org/10.1007/s10067-­019-­04694-­z 1 57. Liang Y, Li H, Gong X, Ding C (2020) Long non-coding RNA THRIL mediates cell growth and inflammatory response of fibroblast-like Synoviocytes by activating PI3K/AKT signals in rheumatoid arthritis. Inflammation 43(3):1044–1053. https://doi.org/10.1007/ s10753-­020-­01189-­x 158. Dong D, Mu Z, Zhao C (2018) Sun M. : a novel tumor-related long non-coding RNA. Cancer Cell Int 18:125. https://doi.org/10.1186/s12935-­018-­0623-­y 159. Ye Y, Gao X, Yang N (2018) LncRNA ZFAS1 promotes cell migration and invasion of fibroblast-like synoviocytes by suppression of miR-27a in rheumatoid arthritis. Hum Cell 31(1):14–21. https://doi. org/10.1007/s13577-­017-­0179-­5 160. Gergianaki I, Bortoluzzi A, Bertsias G (2018) Update on the epidemiology, risk factors, and disease outcomes of systemic lupus erythematosus. Best Pract Res Clin Rheumatol 32(2):188–205. https://doi.org/10.1016/j.berh.2018.09.004 161. Margery-Muir AA, Bundell C, Nelson D, Groth DM, Wetherall JD (2017) Gender balance in patients with systemic lupus erythematosus. Autoimmun Rev 16(3):258–268. https://doi.org/10.1016/j. autrev.2017.01.007 162. Lewis MJ, Jawad AS (2017) The effect of ethnicity and genetic ancestry on the epidemiology, clinical features and outcome of systemic lupus erythematosus. Rheumatology (Oxford) 56(suppl_1):i67–i77. https://doi.org/10.1093/rheumatology/kew399 163. Ferretti C, La Cava A (2016) Chapter 8 - overview of the pathogenesis of systemic lupus erythematosus. In: Tsokos GC (ed) Systemic lupus erythematosus. Academic, pp 55–62 164. Sawalha AH (2016) Chapter 15 - neutrophils in systemic lupus erythematosus. In: Basic, applied and clinical aspects 165. Jacquemin C, Blanco P (2016) Chapter 16 - the role of dendritic cells in systemic lupus erythematosus. In: Tsokos GC (ed) Systemic lupus erythematosus. Academic, pp 131–136 166. Drijvers JM, Pillai S (2016) Chapter 14  - integrating current thinking on peripheral B-cell tolerance in lupus. In: Tsokos GC (ed) Systemic lupus erythematosus. Academic, pp 121–126 167. Tsokos GC (2020) Autoimmunity and organ damage in systemic lupus erythematosus. Nat Immunol 21(6):605–614. https://doi.org/10.1038/ s41590-­020-­0677-­6 168. Tsokos GC, Lo MS, Costa Reis P, Sullivan KE (2016) New insights into the immunopathogenesis of systemic lupus erythematosus. Nat Rev Rheumatol 12(12):716–730. https://doi.org/10.1038/ nrrheum.2016.186 169. Hagberg N, Rönnblom L (2016) Chapter 19  - the interferon system in lupus erythematosus. In: Tsokos GC (ed) Systemic lupus erythematosus. Academic, pp 153–158

S. N. Wijesinghe et al. 170. Luan M, Shang Z, Teng Y et al (2017) The shared and specific mechanism of four autoimmune diseases. Oncotarget 8(65):108355–108374. https://doi. org/10.18632/oncotarget.19383 171. Moulton VR, Suarez-Fueyo A, Meidan E, Li H, Mizui M, Tsokos GC (2017) Pathogenesis of human systemic lupus erythematosus: a cellular perspective. Trends Mol Med 23(7):615–635. https://doi. org/10.1016/j.molmed.2017.05.006 172. Moulton VR (2016) Chapter 17  - Cytokines. In: Tsokos GC (ed) Systemic lupus erythematosus. Academic, pp 137–141 173. Sunahori K, Nagpal K, Hedrich CM, Mizui M, Fitzgerald LM, Tsokos GC (2013) The catalytic subunit of protein phosphatase 2A (PP2Ac) promotes DNA hypomethylation by suppressing the phosphorylated mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) kinase (MEK)/phosphorylated ERK/DNMT1 protein pathway in T-cells from controls and systemic lupus erythematosus patients. J Biol Chem 288(30):21936–21944. https:// doi.org/10.1074/jbc.M113.467266 174. Ye H, Wang X, Wang L et  al (2019) Full highthroughput sequencing analysis of differences in expression profiles of long noncoding RNAs and their mechanisms of action in systemic lupus erythematosus. Arthritis Res Ther 21(1):70. https://doi. org/10.1186/s13075-­019-­1853-­7 175. Luo Q, Li X, Xu C et al (2018) Integrative analysis of long non-coding RNAs and messenger RNA expression profiles in systemic lupus erythematosus. Mol Med Rep 17(3):3489–3496. https://doi.org/10.3892/ mmr.2017.8344 176. Li J, Wu GC, Zhang TP et al (2017) Association of long noncoding RNAs expression levels and their gene polymorphisms with systemic lupus erythematosus. Sci Rep 7(1):15119. https://doi.org/10.1038/ s41598-­017-­15156-­4 177. Wu GC, Hu Y, Guan SY, Ye DQ, Pan HF (2019) Differential plasma expression profiles of long non-coding RNAs reveal potential biomarkers for systemic lupus erythematosus. Biomol Ther 9(6). https://doi.org/10.3390/biom9060206 178. Xu H, Chen W, Zheng F et al (2020) Reconstruction and analysis of the aberrant lncRNA-miRNA-­ mRNA network in systemic lupus erythematosus. Lupus 29(4):398–406. https://doi. org/10.1177/0961203320908927 179. Li C, Pan S, Song Y, Li Y, Qu J (2019) Silence of lncRNA MIAT protects ATDC5 cells against lipopolysaccharides challenge via up-­ regulating miR-132. Artif Cells Nanomed Biotechnol 47(1):2521–2527. https://doi.org/10.1080/2169140 1.2019.1626410 180. Ali MA, Shaker OG, Khalefa AA et al (2020) Serum long noncoding RNAs FAS-AS1 & PVT1 are novel biomarkers for systemic lupus erythematous. Br J Biomed Sci:1–5. https://doi.org/10.1080/09674845. 2020.1765459

4  Long Non-coding RNAs in Rheumatology 181. Zhu JK, He TD, Wei ZX, Wang YM (2018) LncRNA FAS-AS1 promotes the degradation of extracellular matrix of cartilage in osteoarthritis. Eur Rev Med Pharmacol Sci 22(10):2966–2972. https://doi. org/10.26355/eurrev_201805_15051 182. Wang JB, Li J, Zhang TP et  al (2019) Expression of several long noncoding RNAs in peripheral blood mononuclear cells of patients with systemic lupus erythematosus. Adv Med Sci 64(2):430–436. https:// doi.org/10.1016/j.advms.2019.08.002 183. Wang Y, Chen S, Du J et  al (2018) Long noncoding RNA expression profile and association with SLEDAI score in monocyte-derived dendritic cells from patients with systematic lupus erythematosus. Arthritis Res Ther 20(1):138. https://doi. org/10.1186/s13075-­018-­1640-­x 184. Xuan J, Xiong Y, Shi L, Aramini B, Wang H (2019) Do lncRNAs and circRNAs expression profiles influence discoid lupus erythematosus progression?-a comprehensive analysis. Ann Transl Med 7(23):728. https://doi.org/10.21037/atm.2019.12.10 185. Guo G, Chen A, Ye L et al (2020) TCONS_00483150 as a novel diagnostic biomarker of systemic lupus erythematosus. Epigenomics 12(11):973–988. https://doi.org/10.2217/epi-­2019-­0265 186. Li L-J, Zhao W, Tao S-S et al (2017) Comprehensive long non-coding RNA expression profiling reveals their potential roles in systemic lupus erythematosus. Cell Immunol 319:17–27. https://doi.org/10.1016/j. cellimm.2017.06.004 187. Yang H, Liang N, Wang M et al (2017) Long noncoding RNA MALAT-1 is a novel inflammatory regulator in human systemic lupus erythematosus. Oncotarget 8(44):77400–77406. https://doi. org/10.18632/oncotarget.20490 188. Gao F, Tan Y, Luo H (2020) MALAT1 is involved in type I IFNs-mediated systemic lupus erythematosus by up-regulating OAS2, OAS3, and OASL.  Braz J Med Biol Res 53(5):e9292. https:// doi.org/10.1590/1414-­431X20209292 189. Wu GC, Li J, Leng RX et al (2017) Identification of long non-coding RNAs GAS5, linc0597 and lnc-DC in plasma as novel biomarkers for systemic lupus erythematosus. Oncotarget 8(14):23650–23663. https://doi.org/10.18632/oncotarget.15569 190. Suo QF, Sheng J, Qiang FY, Tang ZS, Yang YY (2018) Association of long non-coding RNA GAS5 and miR-21 levels in CD4. Exp Ther Med 15(1):345– 350. https://doi.org/10.3892/etm.2017.5429 191. Zhang F, Wu L, Qian J et al (2016) Identification of the long noncoding RNA NEAT1 as a novel inflammatory regulator acting through MAPK pathway in human lupus. J Autoimmun 75:96–104. https://doi. org/10.1016/j.jaut.2016.07.012 192. Dong G, Yang Y, Li X et  al (2020) Granulocytic myeloid-derived suppressor cells contribute to IFN-I signaling activation of B cells and disease progression through the lncRNA NEAT1-BAFF axis in systemic lupus erythematosus. Biochim Biophys Acta Mol basis Dis 1866(1):165554. https://doi. org/10.1016/j.bbadis.2019.165554

69 193. Zhang LH, Xiao B, Zhong M et al (2020) LncRNA NEAT1 accelerates renal mesangial cell injury via modulating the miR-146b/TRAF6/NF-κB axis in lupus nephritis. Cell Tissue Res. https://doi. org/10.1007/s00441-­020-­03248-­z 194. Rider V, Abdou NI, Kimler BF, Lu N, Brown S, Fridley BL (2018) Gender bias in human systemic lupus erythematosus: a problem of steroid receptor action? Front Immunol 9:611. https://doi. org/10.3389/fimmu.2018.00611 195. Wang J, Syrett CM, Kramer MC, Basu A, Atchison ML, Anguera MC (2016) Unusual maintenance of X chromosome inactivation predisposes female lymphocytes for increased expression from the inactive X. Proc Natl Acad Sci U S A 113(14):E2029–E2038. https://doi.org/10.1073/pnas.1520113113 196. Martin AR, Syrett CM, Myles A, Atchison ML, Anguera MC (2018) Atypical Xist RNA localization to the inactive X in a female-biased murine model of systemic lupus erythematosus. J Immunol 200(1 Supplement):40.18–40.18 197. Syrett CM, Paneru B, Sandoval-Heglund D et  al (2019) Altered X-chromosome inactivation in T cells may promote sex-biased autoimmune diseases. JCI Insight 4(7). https://doi.org/10.1172/jci. insight.126751 198. Zhang Y, Li X, Gibson A, Edberg J, Kimberly RP, Absher DM (2020) Skewed allelic expression on X-chromosome associated with aberrant expression of XIST on systemic lupus erythematosus lymphocytes. Hum Mol Genet. https://doi.org/10.1093/hmg/ ddaa131 199. Lee JT, Davidow LS, Warshawsky D (1999) Tsix, a gene antisense to Xist at the X-inactivation Centre. Nat Genet 21(4):400–404. https://doi. org/10.1038/7734 200. Loos F, Maduro C, Loda A et  al (2016) Xist and Tsix transcription dynamics is regulated by the X-to-autosome ratio and Semistable transcriptional states. Mol Cell Biol 36(21):2656–2667. https://doi. org/10.1128/MCB.00183-­16 201. Cao HY, Li D, Wang YP, Lu HX, Sun J, Li HB (2020) Clinical significance of reduced expression of lncRNA TUG1  in the peripheral blood of systemic lupus erythematosus patients. Int J Rheum Dis 23(3):428–434. https://doi. org/10.1111/1756-­185X.13786 202. Xu Y, Deng W, Zhang W (2018) Long non-coding RNA TUG1 protects renal tubular epithelial cells against injury induced by lipopolysaccharide via regulating microRNA-223. Biomed Pharmacother 104:509–519. https://doi.org/10.1016/j. biopha.2018.05.069 203. Cao HY, Li D, Wang YP, Lu HX, Sun J, Li HB (2020) The protection of NF-κB inhibition on kidney injury of systemic lupus erythematosus mice may be correlated with lncRNA TUG1. Kaohsiung J Med Sci 36(5):354–362. https://doi.org/10.1002/kjm2.12183 204. Jiang CR, Li TH (2018) Circulating UCA1 is highly expressed in patients with systemic lupus erythematosus and promotes the progression through

70 the AKT pathway. Eur Rev Med Pharmacol Sci 22(8):2364–2371. https://doi.org/10.26355/ eurrev_201804_14828 205. Deng Y, Luan S, Zhang Q, Xiao Y (2018) Long noncoding RNA THRIL contributes in lipopolysaccharide-­ induced HK-2 cells injury by

S. N. Wijesinghe et al. sponging miR-34a. J Cell Biochem. https://doi. org/10.1002/jcb.27354

5

LncRNAs and Cardiovascular Disease Elizabeth J. Hennessy

Abstract

A novel class of RNA molecule emerged from human transcriptome sequencing studies termed long non-coding RNAs. These RNA molecules differ from other classes of non-­ coding RNAs such as microRNAs in their sizes, sequence motifs and structures. Studies have demonstrated that long non-coding RNAs play a prominent role in the development and progression of cardiovascular disease. They provide the cell with tiered levels of gene regulation interacting with DNA, other RNA molecules or proteins acting in various capacities to control a variety of cellular mechanisms. Cell specificity is a hallmark of lncRNA studies and they have been identified in macrophages, smooth muscle cells, endothelial cells and hepatocytes. Recent lncRNA studies have uncovered functional micropeptides encoded within lncRNA genes that can have a different function to the lncRNA. Disease associated mutations in the genome tend to occur in non-coding regions signifying the importance of non-coding genes in disease associations. There is a great deal of work to be done in the non-coding E. J. Hennessy (*) University of Pennsylvania, Perelman School of Medicine, Institute for Translational Medicine and Therapeutics (ITMAT), Philadelphia, PA, USA e-mail: [email protected]

RNA field and tremendous therapeutic potential due to their cell type specificity. A better understanding of the functions and interactions of lncRNAs will inevitably have clinical implications. Keywords

Non-coding RNA · Cardiovascular disease · Cholesterol · SNP · Micropeptide · Biomarker · Therapeutic

5.1

Introduction

Cardiovascular disease (CVD) has been the number one cause of death worldwide for nearly a century regardless of economic status accounting for 17.3 million deaths per year, a number that will grow to more than 23.6 million by 2030 and cost $320.1 billion annually (heart.org). The World Health Organization classifies CVD as a group of disorders of the heart and vasculature that includes congenital heart disease (CHD), which occurs when there are defects to the heart that exist at birth due to errors that occurred during development. Other disorders associated with CVD include, diseases of the arteries that supply the heart with blood such as coronary artery disease (CAD), atherosclerosis, hypertension, thrombosis, aortic aneurysm, cardiac fibrosis, myocardial infarction (MI) and ischemic stroke.

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_5

71

E. J. Hennessy

72

Current therapies focus on addressing the role of lifestyle factors and cholesterol whereas recent genetic studies have shown that a set of 100 genetic variants or single nucleotide polymorphisms (SNPs) influence CVD risk [1]. Despite advances in treatments for the pathologies associated with CVD, including widespread use of statins to lower intracellular cholesterol synthesis there is still a significant need for a better understanding of the processes involved in the progression of the disease and new therapeutics to prevent it. When the human transcriptome was sequenced over a decade ago [2], it was revealed that the majority of transcription was occurring in regions where there was no evidence of annotated protein-­coding genes; these regions were referred to as “dark matter of the genome”. Approximately 85% of the genome is actively transcribed into several classes of non-coding RNA molecules including microRNAs (miRNAs) and long non-­ coding RNAs (lncRNAs) [3]. LncRNAs have been identified in all model organisms [4–6]. The importance of non-coding RNA molecules is evident by the fact that as the complexity of an organism increases, the abundance of non-coding RNA sequences found in its genome also grows illustrating the requirement for more sophisticated transcriptional regulation in more evolved eukaryotic species [7]. They have no strict sequence conservation restraints like protein-­ coding genes and they evolve quite rapidly. The conservation of lncRNAs can be at the sequence level, through synteny of their genomic regions, shared functions, or at the structural level. There are three main methods lncRNAs are thought to have evolved; via transformation of a protein coding gene, chromosomal rearrangement and the movement of transposable elements (TEs) throughout the genome [8]. The vast majority of human lncRNAs are non-conserved and many are most likely tissue specific. Lack of conservation often suggests a lack of function but there is emerging evidence that non-conserved lncRNAs are clearly functional and it is not their sequences that deem them conserved but their secondary structures [9]. One of the first lncRNAs to be described, Xist regulates X-chromosome inacti-

vation, and it has undergone rapid sequence evolution, while preserving its function. Despite moderate sequence conservation, Xist displays conserved RNA secondary structure between various species [10]. It is clear from this example that we do not yet understand the mode of lncRNA conservation. There are four general methods by which lncRNAs execute their functions; as signals, decoys, guides or scaffolds [11]. They exhibit specific and regulated patterns of expression in cells and tissues, which can help when identifying them and trying to determine their function [12, 13]. LncRNAs can control transcriptional and post-transcriptional gene regulation as well as mRNA translation. The interactions of nuclear lncRNAs with other molecules can result in cellular epigenetic modifications such as changes to DNA methylation status and modifications to histones as well as the remodeling of chromatin, which culminate in transcriptional activation or repression of target gene expression [14, 15]. Cytoplasmic lncRNAs can interact with miRNAs to post-transcriptionally regulate gene expression as well as act as molecular scaffolds for RNA-­ protein complexes [16, 17]. There is much to be uncovered about the roles of lncRNAs in the cell and the mechanisms they use to exert their functions.

5.2

 he Role of Long Non-coding T RNAs in the Regulation of CVD

LncRNAs contribute to the regulation of essential cellular processes in a cell and tissue specific manner, such as differentiation and the control of disease networks. They have been featured in studies involving cellular development and differentiation as well as exhibited prominent roles in cancer progression. Although several thousand lncRNAs have been identified in the genome [18–21], the function of only a limited number has thus far been described. Recent studies indicate that non-coding RNAs play important roles in the cell specific regulation of genes involved in cardiovascular development and disease. Several

5  LncRNAs and Cardiovascular Disease

key examples include Fendrr (FOXF1 Adjacent Non-Coding Developmental Regulatory RNA) and Braveheart that are required for proper embryonic development of the mouse and cardiovascular lineage commitment [22–25]. My work in the lab of Dr. Kathryn Moore, characterized a primate specific lncRNA called CHROME (Cholesterol Homeostasis Regulator of MiRNA Expression aka CHROMR) that was identified to have a key role in regulating cholesterol homeostasis through the targeting of genes involved in this process including ABCA1 (ATP-binding cassette transporter) and OSBPL6 (oxysterol binding protein-like 6) [26]. LeXis (Liver-expressed LXR-induced sequence) and MeXis (macrophage-­expressed LXR-induced sequence) exhibit coordinated crosstalk between the key cholesterol metabolism pathways SREBP and LXR [27, 28]. Because these lncRNAs are distinctly expressed in either mouse or human they will be interesting to study in order to understand the evolution of lncRNA genes. Because of their lack of sequence conservation, human specific lncRNAs are difficult to study because there is not a mouse model to elucidate functional consequences of the expression of the lncRNA. Only a few lncRNAs implicated in diseases in vitro have yielded phenotypes in knockout mouse models [22]. Regardless, it will be of great value to develop a more complete understanding of both mouse and human genomes and the regulation of cell type specific gene pathways leading to pathologies associated with CVD.

5.2.1 LncRNAs in Cardiovascular Lineage Commitment and Development The defective development of cardiac tissues has been linked to susceptibility to cardiovascular disease later in life [29–32]. During vertebrate embryogenesis, the heart is the first organ to function so proper heart development is essential for the correct development of the entire organism. Cardiogenesis occurs when pluripotent stem cells expand and differentiate into mesodermal and cardiac cells types like cardiomyocytes,

73

smooth muscle cells and endothelial cells. The expression of transcription factors that drive cellular differentiation is controlled by the induction and patterning of morphogens, chemical cues produced by source cells that diffuse through tissues in an embryo setting up concentration gradients that direct cells to become specialized. An increasing number of lncRNAs have been implicated in embryonic cardiac development. LncRNAs are differentially expressed in different tissues implying that they regulate cell lineage commitment during development and differentiation. Some of the first studies demonstrating this were done using zebrasfish embryos. A loss of function study examined two lncRNAs, cyrano and megamind, that resulted in developmental defects of the embryo [33]. Mouse studies then identified lncRNAs that were found to be required for proper embryo development. HOTTIP is a lncRNA implicated in mouse limb formation during embryogenesis [34]. Another mouse study demonstrated that a set of lncRNAs expressed in embryonic stem cells physically associates with different chromatin regulatory proteins leading to alterations in global lineage programs and their pluripotent state [35]. These studies suggest that lncRNAs may regulate the epigenetic patterning of the genome through direct interactions with chromatin to control cell fate. It is imperative to develop a deeper understanding of the pathways regulated by lncRNAs during lineage commitment. Fendrr is required for the proper development of the mouse and this was shown using two different knockout models [22–24]. Loss of Fendrr in mice is lethal. It is transcribed bidirectionally from the protein coding gene Foxf1a. It interacts with the polycomb repressor complex 2 (PRC2) bringing it to target gene promoters including Foxf1 and Pitx2, increasing H3K27 trimethylation at these sites and repressing transcription of these genes. A human orthologue of Fendrr has been identified and it is highly induced in IFNγ stimulated macrophages suggesting a role in macrophage polarization [36]. Braveheart like Fendrr is required for cardiovascular lineage commitment through its interaction with SUZ12 a core component of PRC2 to regulate gene regu-

74

latory networks including key transcription factors (MesP1, Gata4, Hand2, Nkx2.5 and Tbx5) [25]. Using SHAPE (selective 2′ hydroxyl acylation analyzed by primer extension) probing, the secondary structure of Braveheart was determined and it was found that a 5′ asymmetric G-rich internal loop in Braveheart interacts with the zinc finger protein CNBP (cellular nucleic acid binding protein) to control cardiomyocyte differentiation [37, 38]. A more recent study has identified the 3-D structure of Braveheart and its complex with CNBP using small angle X-ray scattering (SAXS) [38]. CARMEN (Cardiac mesoderm enhancer associated noncoding RNA) was identified in a study detecting human lncRNAs involved in cardiac development and cardiac precursor cell differentiation [39]. It is located upstream of two microRNAs associated with cardiac precursor cell differentiation miRNA-143 and miRNA-145 suggesting that the locus would have important implications for cardiogenesis. It is expressed during pathological remodeling in mouse and human hearts and is required to maintain differentiated cardiomyocytes. CARMEN interacts with two components of PRC2, SUZ12 and EZH2 bringing them to target gene promoters to regulate stress marker genes Nppa, Nppb, Myh6, Myh7, Tgfb1 and Col1a [39]. The lncRNAs described regulate components of the same pathways involved in cardiogenesis through interactions with the polycomb repressor complex. The more that is understood about the molecular switches that control cardiac lineage commitment and development the more successful therapies will be that target cardiac related diseases.

5.2.2 LncRNAs Regulating Macrophage Lipid Metabolism CVD involves a network of different cell types, including macrophages, vascular smooth muscle cells (VSMCs), endothelial cells (ECs) and hepatocytes. Because lncRNA expression is notoriously tissue and cell type specific [13], it is

E. J. Hennessy

imperative to individually examine the cell types involved in the pathologies associated with CVD and the lncRNAs potentially regulating the pathways responsible. Macrophages are one of the key cell types involved in the development and progression of CVD (Fig. 5.1a). Resident tissue macrophages like those found in cardiac tissues are seeded during embryonic development and they self-renew from circulating monocytes [40]. Macrophages are responsible for the inflammatory response generated in reaction to excess lipid accumulation. The maintenance of proper lipid levels is a tightly regulated process involving various signaling pathways where disruptions in a pathway such as from causal SNPs result in cholesterol related pathologies such as hyperlipidemia and atherosclerosis that contribute to CVD. Extracellular cholesterol uptake and efflux along with endogenous fatty acid and cholesterol biosynthesis are pathways responsible for regulating lipid levels in the cell and are controlled by the nuclear transcription factors LXR (liver X receptor) and SREBP (sterol regulatory element-­ binding proteins). LXR promotes the removal of cholesterol via the efflux pathway where cholesterol is exported from the cell to the acceptor molecule HDL (high-density lipoprotein) and transported to the liver, converted to bile acids and excreted [41]. Agonists of LXR stimulate fatty acid synthesis and elevate triglycerides by inducing SREBP dependent genes. SREBPs regulate genes involved in the synthesis and uptake of cholesterol and fatty acids essential components of the cell [42, 43]. They are activated under low sterol conditions where they get translocated from the membrane of the endoplasmic reticulum to the nucleus and bind to specific sterol regulatory element DNA sequence motifs in the promoters of genes. Treatment of cells with statins reduces cholesterol content by inhibiting endogenous cholesterol synthesis and induces the transcription of SREBP [44]. Excess sterol in the cell inhibits SREBP translocation to the nucleus by directly binding to its partner molecule SCAP indicating a negative feedback mechanism, whereas the LXR transcription factor is activated

5  LncRNAs and Cardiovascular Disease

75

Fig. 5.1  The development of an atherosclerotic plaque involves macrophages (a) smooth muscle cells (b) endothelial cells (c) and hepatocytes (d). Macrophages located in the intima engulf LDL cholesterol but when there is defective cholesterol efflux due to excess cholesterol or decreased circulating HDL from the liver macrophages can progress to foam cells and accumulate in the intima where they can become necrotic. The response of smooth

muscle cells to LDL cholesterol is to proliferate, migrate and secrete extracellular matrix proteins into the intima region. The plaque becomes more stable and eventually after there is enough necrosis it is less stable and can rupture. Endothelial cells respond to excess LDL cholesterol by remodeling the intimal space by proliferating and sprouting. (Figure made using biorender)

in high sterol conditions [45, 46]. It acts as a cholesterol sensor and protects the cell from the deleterious consequences of excess cholesterol. The LXR and SREBP transcription factors are crucial for the maintenance of cholesterol homeostasis. Their aberrant expression can affect their

downstream signaling events disrupt the fine balance of cellular lipid levels. A greater understanding of how these signaling pathways are controlled particularly by lncRNAs will ultimately contribute to novel therapeutics and the prevention of CVD.

76

CHROME is a primate specific lncRNA with a key role in regulating liver and macrophage cholesterol homeostasis through the targeting of genes involved in this process [26]. Dietary cholesterol and cellular cholesterol loading induce the expression of CHROME via the LXR transcription factors to mediate the response to excess cholesterol. CHROME promotes cholesterol efflux and biogenesis of HDL by sequestering a set of miRNAs from binding to their target mRNAs, which have key roles in these processes including ABCA1 and OSBPL6 [47, 48]. Cells lacking CHROME exhibit reduced expression of ABCA1 and OSBPL6 because the microRNAs are free to bind to their 3′UTRs and degrade them. CHROME is a central component in the regulation of cholesterol homeostasis in humans (Fig.  5.2a). MeXis is a lncRNA found on the same locus as Abca1 and exhibits coordinated crosstalk between the SREBP and LXR pathways [28]. Its expression is induced in response to LXR activation by the agonist GW3965 as well as oxidized or acetylated LDL in macrophages. Knockout of MeXis in mouse bone marrow derived cells resulted in altered chromosome architecture at the Abca1 locus resulting in impaired cholesterol efflux and accelerated the development of atherosclerosis. MeXis guides the transcriptional coactivator DDX17 to the ­promoter of Abca1 modulating its expression in a cell type specific fashion (Fig. 5.2b). Experiments performed in human macrophages demonstrated that the LXR-MeXis-ABCA1 axis is conserved and may be relevant for studying this pathway in human disease. Several studies have identified lncRNAs that are regulated during foam cell formation by treating THP-1 derived macrophages with oxidized LDL (oxLDL). A human lncRNA array identified that oxLDL induced the expression of lncRNA-­ DYNLRB2-­2 (named for nearest annotated gene, dynein, light chain, roadblock-type 2) in THP-1 macrophage-derived foam cells [49]. lncRNA-­ DYNLRB2-­2 promotes cholesterol efflux via the upregulation of ABCA1 by the G-protein coupled receptor 119 (GPR119) and reduced the production of inflammatory cytokines TNF-α, IL-1β, and IL-6 found in the serum of ApoE−/− mice

E. J. Hennessy

[50]. It remains unknown what the mechanism is resulting in the oxLDL induction of lncRNA-­ DYNLRB2-­2 but future studies of this lncRNA could provide additional evidence of a novel pathway leading to atheroprotection [51]. A well characterized lncRNA that is essential for paraspeckle formation, NEAT1 has also been implicated in the macrophage response to oxLDL [52]. A study investigating the role of paraspeckles in lipid uptake demonstrated that NEAT1 is induced in response to oxLDL. It localizes to these subnuclear organelles where it interacts with the structural protein NONO.  Paraspeckles are thought to regulate gene expression by sequestering proteins or mRNAs, protecting them from degradation so they are available when the cell needs them [53]. Transfecting siRNA against NEAT1 into human THP-1 derived macrophages, the authors show that NEAT1 inhibits lipid uptake by suppressing the receptor CD36 and promotes TNF-α secretion by activating p65 phosphorylation and NFκB signaling. A detailed mechanism is lacking from this study, but it is interesting to observe a well-studied lncRNA demonstrating roles in different cell types and pathologies [54]. A microarray analysis of THP-1 derived macrophages versus THP-1 derived macrophage foam cells identified RP5-833A20.1 to be upregulated in response to oxLDL whereas its host gene the transcription factor NFIA is downregulated. The authors use both in vitro and an in vivo ApoE−/− mouse model to uncover a mechanism by which RP5-833A20.1 induces the expression of the miRNA hsa-miR-382-5p assisting the miRNA in binding to its target gene NFIA and reducing its expression. The decrease in NFIA promotes THP-1 macrophage derived foam cell formation through an unknown mechanism [55]. E330013P06 is a proinflammatory lncRNA that is upregulated in monocytes from type 2 diabetic patients and in macrophages from db/db (leptin deficient) and diet induced insulin resistant type 2 diabetic mice [56]. The overexpression of this lncRNA in macrophages induced inflammatory genes and increased foam cell formation whereas a small interfering RNA (siRNA) against E330013P06 inhibited the same set of

5  LncRNAs and Cardiovascular Disease

77

ABCA1

OSBPL6

ATP8B1

MeXis guides DDX17 to ABCA1 promoter to induce ABCA1 and increase cholestral efflux

Fig. 5.2  LncRNAs bind to and interact with proteins to control signaling pathways involved in CVD. CHROME (a) is expressed in macrophages and hepatocytes. It is expressed in atherosclerotic plaques, induced in response to high cholesterol diet and LXR activation. It binds to Ago2 and its associated miRNAs to sequester them away from their target mRNAs involved in cholesterol efflux. MeXis (b) is induced in macrophages in response to oxLDL, acLDL and LXR activation. It binds to the transcriptional coactivator DDX17 bringing it to the promoter of Abca1 to increase its expression and increase cholesterol efflux. LeXis (b) is located on the same locus as MeXis but it is liver specific. It is also induced in response to excess cholesterol and LXR activation. It binds to the RNA polymerase II partner Raly to prevent RNA poly-

merase II from binding to target gene promoters such as Srebf2, Hmgcr, Cyp51 and Fdps and results in a reduction of cholesterol and triglycerides. ApoA1-AS (c) is a liver specific lncRNA found within the apolipoprotein gene locus and is induced in obese conditions and fatty lover disease. It binds to the PRC2 component Suz12 bringing it to the promoter regions of ApoA1, ApoC3 and ApoA4 silencing their transcription. ApoA4-AS (c) is also liver specific and found on the same strand of DNA as ApoA1-AS.  It is induced in high cholesterol conditions like obesity to bind to the mRNA stabilizing protein HuR to protect ApoA4 mRNA and increasing secretion of cholesterol and triglycerides into the plasma. (Protein structures taken from Wikipedia, RNAFold used to generate secondary structures of lncRNAs)

78

E. J. Hennessy

inflammatory genes. The relevance for this 5.2.3 LncRNAs Involved in Smooth Muscle Cell Proliferation lncRNA in vivo remains unknown. Interestingly, and Migration the E330013P06 gene overlaps the gene encoding the lncRNA CARMEN described in the section on cardiovascular development (Sect. 5.2.1) The walls of blood vessels are comprised of suggesting a potential coordinated role in regu- VSMCs and ECs, each having a role in maintainlating the development of cardiovascular cells ing vascular homeostasis. VSMCs are normally and their roles in CVD progression. LncRNA quiescent and contractile but upon vascular injury H19 is highly conserved and has established such as the progression of atherosclerosis, functions in regulation of body weight, cell pro- VSMCs release cytokines, growth factors, and liferation and development of some cancers [57– deposit extracellular matrix proteins, resulting in 59]. Mutations in this lncRNA can cause genomic the aggregation of platelets and recruitment of imprinting associated diseases Beckwith-­inflammatory cells. The cumulative effect is a Wiedemann and Prader Willi syndromes [60, 61]. phenotypic switch of the VSMCs to a more synH19 is expressed from chromosome 11 in human thetic and pro-proliferative and migratory state and chromosome 7  in mouse and exclusively [68, 69]. Figure 5.1b illustrates the early events transcribed from the maternally inherited allele that lead to the development of the atherosclein a process called imprinting [62]. H19 overex- rotic plaque including the proliferation, migrapression in a model of rat cardiomyopathy tion and accumulation of VSMCs. VSMCs can showed reduced inflammatory cytokine produc- express lipid uptake receptors and form foam-­ tion in the myocardial tissues [63] and LPS treat- like cells participating in the early accumulation ment of a chondrocyte cell line upregulated H19 of lipids alongside macrophages resulting in and decreased miR-130a. Knockdown of H19 in plaque formation. VSMCs play a crucial role in the same cell type increased cell viability and vascular remodeling and atherosclerosis. apoptosis and reduced pro-inflammatory cytoSeveral studies have shown that the lncRNA kines IL-1β, IL-6 and TNF-α [64]. The authors H19 has a role in regulating the VSMC response show that the H19 dependent regulation of in CVD. A study examining lncRNA expression inflammation is via a molecular sponging mecha- in mouse and human abdominal aortic aneurysm nism miRNA miR-130a. The miR-130 family of (AAA), a chronic inflammatory condition of the miRNAs has previously been implicated in the vasculature found that H19 was the most signifinegative regulation of inflammatory metabolism cantly upregulated lncRNA in through NFκB signaling [65]. The expression of AAA.  Overexpressing H19  in mice enhanced H19 was upregulated in blood samples from AAA formation and increased Il-6 and Mcp-1 patients with atherosclerosis [66]and in oxLDL while also enhancing macrophage infiltration. treated Raw264.7 macrophage cells while expres- Knockdown of H19  in mice using shRNA sion of miR-130b was decreased in oxLDL decreased AAA formation and Il-6 and Mcp-1 treated cells [67]. Knockdown of H19 with following AAA inducing angiotensin II infusion shRNA in Raw264.7 treated with oxLDL, showed [70]. H19 competes with the miRNA let-7a to decreased IL-1β, and TNF-α and increased IL-4, regulate Il-6 expression. In a study of both IL-10 and miR-130b expression demonstrating a VSMCs and ECs, overexpression of H19 role for H19  in atherosclerosis progression. increased proliferation and decreased apoptosis Additional studies examining H19  in vascular of both cell types [71]. The paraspeckle localized cell types including VSMCs and ECs have fur- lncRNA NEAT1 is expressed in VSMCs and is ther implicated H19 in the inflammatory response induced during phenotype switching, when and in the pathologies associated with CVD. VSMCs move from a contractile to synthetic phenotype in response to vascular injury such as in the atherosclerotic plaque [72]. Silencing NEAT1 with siRNA enhanced the expression of

5  LncRNAs and Cardiovascular Disease

smooth muscle cell specific genes and reduced proliferation and migration. NEAT1 knockout mice exhibited decreased neointima formation following injury due to the attenuated VSMC proliferation. NEAT1 interacts with Wdr5 (WD repeat domain 5) where it sequesters it from binding to smooth muscle cell specific genes resulting in their downregulation [72]. lincRNA­p21 was measured in atherosclerotic plaques from ApoE−/− mice and its expression was significantly lower compared to plaques isolated from control wild-type mice fed a high fat diet [73]. siRNA silencing of lincRNA-p21 in human VSMCs increased cell proliferation, decreased the expression of the p53 pathway downstream gene Mdm2 (mouse double minute 2) and decreased apoptosis. Mdm2 interacts with lincRNA-­p21 to relieve its repression on p53 and enhance p53 activation. p53 has a well-­established role in the development of atherosclerosis, its inhibition accelerates atherosclerosis progression [74]. RNA-seq of primary human carotid aortic SMCs (HCASMCs) revealed that the lncRNA SENCR (Smooth muscle and Endothelial Cell eNriched migration/differentiation associated lncRNA) is expressed and could control SMC phenotypic switching [75]. Knockdown of SENCR in SMCs revealed decreased expression of contractile associated genes and increased ­pro-­migratory genes that increased SMC migration. The mechanism of SENCR regulation of SMC migration remains unknown but because it localizes to the cytoplasm it is unlikely to interact with DNA elements in the nucleus and likely to participate in post-transcriptional regulation of the target genes. Another lncRNA involved in VSMC phenotype switching is SMILR (Smooth Muscle Induced lncRNA). This lncRNA was identified in an RNA-seq analysis of human VSMCs treated with IL-1α or PDGF (platelet derived growth factor) key regulators in the VSMC response to injury [76]. Overexpression of SMILR caused an increase in VSMC proliferation. Knockdown of SMILR reduced the expression of its neighboring gene HAS2, which synthesizes hyaluronic acid, a component of the extracellular matrix that accumulates in athero-

79

sclerotic plaques. Some of the VSMC expressed lncRNAs described are expressed in other cell types and specific mechanisms and binding partners have been described, for the sake of this review we wanted to focus on their role in proliferation and migration of VSMCs.

5.2.4 LncRNAs Regulating Endothelial Cell Growth and Sprouting Like VSMCs, ECs play a crucial role in the response to atherosclerosis. The endothelial lining of blood vessels undergoes remodeling in response to stressors like excess LDL cholesterol and this results in impaired nitric oxide dependent vasodilation and glucose uptake ultimately leading to enhanced oxidative stress and inflammation. Dysfunction of the endothelial lining of the arterial vasculature is an important contributor to the pathobiology of atherosclerosis (Fig.  5.1c). Activated ECs upregulate adhesion molecules like E-selectin, P-selectin, VCAM-1 (vascular cell adhesion molecule-1) and MCP-1 [77]. Like macrophages and VSMCs, ECs also express lncRNAs that regulate their response to CVD and the mutually expressed lncRNA H19 plays a prominent role. The overexpression of H19 in HUVECs (human umbilical vein endothelial cells), upregulates ACP5 (acid phosphatase 5), a gene associated with cancer progression, this upregulation promotes EC proliferation and apoptosis. The authors use data obtained from GEO dataset GSE76741 containing differentially expressed genes in HUVECs with H19 knockdown by LNA GapmeR to determine ACP5 is a target of H19 [78]. H19 and ACP5 were both significantly increased in serum of atherosclerotic patients suggesting a positive regulatory role for H19 on ACP5 expression. In another study, the inhibition of H19 with antisense LNA GapmeRs resulted in decreased EC growth and the ability of the ECs to form capillary like structures signifying a decrease in angiogenesis [79]. Microarray analysis revealed an enrichment of genes related to cell cycle including cyclin/CDK inhibitors.

80

This study did not identify a mechanism for the H19 dependent effects on cell growth but future studies could examine previously identified targets of H19  in other cell types, including miR-­ 130a in macrophages. A more recent study examined H19 in ECs and found overexpression of H19 promoted cell growth, increased VEGF (vascular endothelial growth factor) and eNOS (endothelial nitric oxide synthase) protein levels and capillary tube like structure formation or “sprouting” [80]. H19 was found to downregulate miRNA-181a leading to activation of the JNK and AMP kinase pathways and subsequent angiogenic effects. An additional study added to this mechanism by demonstrating that H19 regulates VEGF by inhibiting the phosphorylation of STAT3 signaling [81]. LncRNA MALAT1 (metastasis-associated lung adenocarcinoma transcript 1 – aka NEAT2) is a conserved lncRNA expressed in mouse and human. Its expression is increased in response to hypoxia and silencing of MALAT1 with siRNAs or GapmeRs induced migration and sprouting but inhibited proliferation [82]. In vivo deletion of MALAT1 further showed the inhibition of proliferation and neointima formation. Microarray analysis of MALAT1 siRNA transfected HUVECs showed an impact of MALAT1 on cell cycle genes including downregulation of cyclinA2, cyclinB1, cyclinB2 and CDK1 (cyclin dependent kinase 1) and an induction of the cell cycle inhibitor p21. This study demonstrates the balancing act MALAT1 is performing between the proliferative to migratory phenotypes.

E. J. Hennessy

(cholesterol 7a-hydroxylase) an enzyme critical for the conversion of cholesterol into bile acids which are excreted from the body and the APO (apolipoprotein) gene locus whose lipoprotein products act as precursor molecules to the cholesterol acceptor molecule HDL. The cholesterol transporter ABCA1 is also expressed by hepatocytes and facilitates the production of HDL and release into the circulation where it is used in cholesterol efflux by macrophages and hepatocytes as well as reverse cholesterol transport [83]. Two antisense lncRNAs, APOA1-AS and APOA4-AS were identified within the APO gene cluster on chromosome 11q23.3 encoding APOA1, APOC3, APOA4 and APOA5. APOA1-AS and APOA4-AS cis-regulate the formation and function of the plasma lipoproteins by acting as switches to enable the interaction of repressive chromatin modifying complexes with the locus [84, 85]. APOA1-AS is expressed only in primates and interacts with a component of PRC2, SUZ12 to mediate gene silencing on the APO genes in the locus (Fig. 5.2c). Silencing of APOA1-AS in hepatic cells resulted in the upregulation of APOA1, APOC3 and APOA4. In vivo delivery of an antisense oligonucleotide (ASO) against APOA1-AS to African Green monkeys increased the HDL precursor APOA1 mRNA in the liver and circulating APOA1 after 1  week confirming the role of APOA1-AS in the HDL biogenesis pathway. APOA4-AS is conserved between mouse and human and is elevated along with APOA4 in mouse models of obesity and in humans with fatty liver disease. APOA4-AS interacts with the mRNA stabilizing protein HuR to stabilize APOA4 mRNA (Fig.  5.2c). Studies 5.2.5 Hepatocyte Expressed knocking down APOA4-AS on obese mice using LncRNAs and the Maintenance shRNA demonstrated decreased hepatic APOA4 of Cholesterol Homeostasis mRNA levels and plasma total cholesterol and triglyceride levels but no change in liver triglycThe liver works in a coordinated effort with the eride levels indicating the primary role of APOA4 cells of the vasculature to maintain cholesterol is in triglyceride secretion from the liver. homeostasis operating a tightly regulated netLeXis is encoded in the same locus as the work of pathways controlling cholesterol uptake, macrophage specific lncRNA MeXis and Abca1 synthesis, transport and excretion (Fig.  5.1d). and exhibits coordinated crosstalk between the The liver expresses HMGCR (3-hydroxy-3-­ SREBP and LXR pathways (Fig. 5.2b) [27, 28]. methylglutaryl-coenzyme A reductase), which is LeXis is expressed in the livers of mice fed a essential for cholesterol synthesis, CYP7A1 western diet or following pharmacological acti-

5  LncRNAs and Cardiovascular Disease

vation of LXR by its agonist GW3965. Overexpression of LeXis caused a reduction in serum and liver cholesterol levels, however this reduction was lost when LeXis was overexpressed in mice deficient in the Srebp-2 adaptor Scap signifying a role for LeXis in the SREBP-2 pathway. LeXis reduces binding of RNA polymerase II to the promoters of Srebf2 and its target genes Hmgcr, Cyp51 and Fdps by binding to the ribonucleoprotein Raly and sequestering it away from the promoter regions and acting as an emergency response to cholesterol overload. A recent study showed that Raly binds specifically to the promoter region of Srebp2 and not its counterpart Srebp1 via direct interaction with Nuclear Transcription factor Y (Nfy) to regulate cholesterol-­modulating genes [86]. A potential human orthologue to LeXis exists signifying a lncRNA that could potentially be used to treat human metabolic disorders. Lnc-HC was identified in a study of liver lncRNAs expressed in rats with high fat diet induced metabolic syndrome [87, 88]. Lnc-HC localizes to the nucleus of hepatocytes where it acts as a negative regulator of cholesterol metabolism by interacting with the heterogeneous nuclear ribonucleoprotein hnRNPA2B1 to regulate the expression of cholesterol metabolism genes Cyp7a1 and Abca1. Knocking down of Lnc-HC with shRNA in a rat model fed a high cholesterol diet showed improved serum parameters including total cholesterol, triglycerides and HDL levels. Following stimulation with excess cholesterol, lnc-HC interacts with hnRNPA2B1 where it forms a complex that interacts with Cyp7a1 and Abca1 mRNAs resulting in their degradation [88]. A follow up study used in vitro and in vivo assays to demonstrate that lnc-HC negatively regulates PPARγ expression by regulating miR-130-3p [89]. PPARγ is a important transcription factor in lipid metabolism pathways. This study further shows the potential of lnc-HC to regulate key pathways involved in CVD. The role of NEAT1 in macrophage lipid uptake and VSMC phenotype switching was detailed in Sects. 5.2.2 and 5.2.3, but NEAT1 also functions in lipid metabolism in the liver. Several

81

studies have shown its ability to act as a miRNA sponge regulating various pathways leading to liver cancer and NAFLD (nonalcoholic fatty liver disease). It was first implicated in lipolysis in a HCC (hepatocellular carcinoma) study [90]. NEAT1 is overexpressed in HCC and disrupts lipolysis through targeting ATGL (adipose triglyceride lipase) and decreasing its products DAG (diacylglycerol) and FFA (free fatty acid) leading to uncontrolled cell proliferation. The study further showed that the NEAT1 mediated promotion of HCC cell growth is through a miR-­ 124-­3p/ATGL/DAG-FFA/PPARα pathway. An in vitro model of NAFLD used the BRL3A rat liver cell line treated with FFA to show that NEAT1 is upregulated during the progression of NAFLD. miRNA-506 was downregulated and its target Gli3, a transcription factor previously implicated in NAFLD was upregulated, indicating a competing endogenous RNA scenario [91]. An in vivo mouse model of high fat diet induced NAFLD demonstrated that NEAT1 promotes hepatic lipid accumulation through its inhibition of anti-­ inflammatory miRNA-146a-5p [92] and upregulation of its target Rock1, a previously implicated player in the progression of NAFLD [93, 94]. The lncRNA lncARSR was first identified in renal cell carcinoma cells that were resistant to Sunitinib, a receptor tyrosine kinase inhibitor approved for treatment of renal cancer and has since been implicated in the pathogenesis of several cancers including hepatic [95, 96]. lncARSR also has a role in hepatic lipogenesis and like NEAT1 it is induced in the HepG2 hepatocyte cell line treated with the fatty acids oleic acid, palmitic acid and stearic acid, in livers of NAFLD patients and the livers of mice fed either a high fat diet or methionine-choline deficient diet inducing NAFLD [97, 98]. YAP1 (Yes-associated protein) is a co-activator of the Hippo signaling pathway and has been implicated in NAFLD by activating Akt signaling and influencing IRS2 (insulin receptor) expression [99]. lncARSR binds to YAP1, promoting its nuclear translocation and inhibiting its phosphorylation. Once in the nucleus, YAP1 phosphorylates Akt and activates insulin signaling via IRS2. Silencing of lncARSR in NAFLD mice decreased Irs2 expression, lipid

82

accumulation and triglycerides alleviating the disease. A related study found that SREBP-1c is an additional target of lncARSR. Similarly, overexpression of lncARSR induced the phosphorylation of Akt and subsequently increased SREBP-1c activity while knockdown of lncARSR using shRNA decreased SREBP-1c. Treatment of cells with a PI3 kinase (PI3K) inhibitor abolished the lncARSR effect on SREBP-1c signifying that lncARSR regulates SREBP-1c by activating the PI3K/Akt pathway [97]. Several other liver expressed lncRNAs have been implicated in regulation of the SREBP pathway. LncHR1 (HCV regulated 1) is a conserved lncRNA that is upregulated in hepatocytes infected with hepatitis C virus (HCV) [100]. It is well established that HCV infection upregulates SREBP and lipid accumulation in hepatocytes [101]. Overexpression of lncHR1 suppresses SREBP-1c expression and decreased lipid droplet formation whereas shRNA silencing of lncHR1 enhanced SREBP-1c expression. A transgenic mouse model overexpressing lncHR1 and fed a high fat diet displayed decreased hepatic SREBP-1c, FAS (fatty acid synthase) and ACCα (acetyl-CoA carboxylase-α). The precise mechanism that lncHR1 uses to suppress SREBP-1c and the levels of lipid in the cell remains unknown but it was established that lncHR1 decreases the phosphorylation of the PDK1/AKT/FoxO1 pathway thus suppressing SREBP-1c [102]. LncHR1 acts as a negative regulator of SREBP-1c during HCV infection to regulate lipid levels. Because lncRNA MALAT1 was implicated in liver cell proliferation and type 2 diabetes [103–105], its role in hepatic lipid accumulation was next examined. Its expression is increased in hepatocytes treated with the fatty acid palmitate and in the livers of obese mice (ob/ ob) [106]. Knockdown of MALAT1 with siRNA in human HepG2 hepatocytes inhibited palmitate induced lipid accumulation and decreased SREBP-1c protein levels while having no effect on its mRNA indicating its post-transcriptional regulation of SREBP-1c. An in vivo model of ob/ ob mice treated with siRNA against MALAT1 showed decreased lipid accumulation in the liver

E. J. Hennessy

and increased Srebp-1c. The study goes on to show that overexpression of MALAT1 prevented the ubiquitination of SREBP-1c protein affecting the stability of the protein. The study found that MALAT1 interacts with SREBP-1c protein in the nucleus promoting its stability and interaction with target gene promoters including ACLY and FAS. The lncRNA Gm16551 is exclusively expressed in the liver. Histone marks surrounding the gene indicate it is actively transcribed in the liver and repressed in the heart. Its expression was reduced in the livers of mice both upon fasting for 24 hours and in ob/ob obese mice compared to lean littermate control animals [107]. Srebp-1c induces Gm16551 and activates a negative feedback loop regulating Srebp-1c activity in the liver by affecting other Srebp-1c regulated lipogenic genes including Acly (ATP citrase lyase), Fas and Scd (stearoyl-coenzyme A desaturase 1) and plasma triglyceride levels [107]. Like Gm16551, lncLSTR (liver-specific triglyceride regulator) was identified in livers of mice subjected to fasting and refeeding [108]. Its expression sharply declined after a 24 hour fast and quickly recovered upon refeeding. shRNA knockdown of lncLSTR resulted in reduced plasma triglyceride levels likely due to enhanced tissue triglyceride clearance. Knockdown of lncLSTR altered the expression of ApoC2, Lpl (lipoprotein lipase) and Cyp8b1 a key regulator of the bile acid pathway by interacting with the DNA binding protein TDP-43 and reducing its occupancy of FXR sites in their promoters (Table 5.1). Only a few lncRNAs have shown significant regulation of metabolic homeostasis in vivo and interestingly a few have a common target SREBP-1c signifying the importance of this pathway. Genome-wide screens of different cardiovascular cell types and tissues have established a comprehensive catalog of lncRNAs implicated in metabolic regulation in CVD associated pathologies. Researchers need to continue developing bioinformatics and experimental tools to advance the roadmap of characterizing functional lncRNAs in lipid homeostasis.

5  LncRNAs and Cardiovascular Disease

83

Table 5.1  LncRNA studies by cell type, species, target and function LncRNA Macrophages CHROME

MeXis LncRNA-­ DYNLRB2-­2 NEAT1 RP5-833A20.1

Species

Target

Function

References

Human

miR-33a/b miR-27b miR-128 DDX17

Promotes cholesterol efflux

[26]

Promotes cholesterol efflux

[28]

GPR119

Promotes cholesterol efflux

[49–51]

NONO miR-382-5p

Suppresses lipid uptake Promotes foam cell formation

[52–54] [55]

Promotes foam cell formation

[56]

miR-130a/b

Promotes inflammation

[62–67]

let-7a

Promotes IL-6 and AAA

[70, 71]

WDR5 MDM2

Promotes phenotype switching Enhances p53 activation

[72] [73, 74]

Affects contractility and migration

[75]

Promotes proliferation

[76]

Promotes cell growth and sprouting

[78–81]

Suppresses migration and sprouting; Promotes proliferation

[82]

SUZ12 HuR

Targets APO gene cluster Promotes cholesterol efflux

[85] [84]

Raly hnRNPA2B1 miR-130-3p miR-124-3p miR-506 miR-146a-5p YAP1

Decrease cholesterol and TGs Suppresses cholesterol metabolism

[27] [87–89]

Promotes cell proliferation

[90–94]

Promotes lipid accumulation

[95–97]

Decreases lipid accumulation

Mouse Human Mouse

Human Mouse Human E330013P06 Mouse Human H19 Mouse Human Smooth muscle cells H19 Mouse Human NEAT1 Mouse LincRNA-p21 Mouse Human SENCR Mouse Human SMILR Human Endothelial cells H19 Human MALAT1

Hepatocytes ApoA1-AS ApoA4-AS LeXis Lnc-HC

Mouse Human

Human Mouse Human Mouse Rat

NEAT1

Mouse Rat Human

LncARSR

Mouse Human Mouse Human Mouse Human Mouse Human Mouse

LncHR1 MALAT1 Gm16551 LncLSTR

ACP5 Cyclin/CDK inhibitors VEGF CyclinA2/B 1/B2 CDK1

SREBP-1c

Promotes lipid accumulation

SREBP-1c

Decreases plasma TGs

[100– 102] [103– 105] [107]

TDP-43

Promotes plasma TGs

[108]

E. J. Hennessy

84

5.3

LncRNA Encoded Micropeptides and Their Role in CVD

detection of peptides smaller than 10 kilodaltons (kDa). There is also a limit to the sensitivity of mass spectrometry leading to bias against the detection of lowly expressed genes which Recent studies have demonstrated that RNA tran- lncRNA genes often are. In vitro translation scripts once thought to be non-coding for pro- assays and western blotting further detects and teins do in fact encode short open reading frames confirms the presence of a peptide. Ribosome (ORFs) that can be non-canonically translated profiling and mass spectrometry studies have into small functional peptides termed micropep- identified thousands of previously unannotated tides. These small proteins are often overlooked CDSs across various species [113–115]. but can be highly functional. The first eukaryotic High throughput methods have been used to micropeptide was identified in soybean and char- identify novel micropeptides [116, 117]. A study acterized in 2002 [109]. Pseudogenes, which from 2017 identified the “hidden proteome” by were thought to be defunct relics of protein-­ examining eukaryotic transcriptomes to classify coding genes, contain signals that encode short unannotated ORFs or alternative ORFS (altORFs) stretches of amino acids. These newly identified that were greater than 30 amino acids in length. micropeptides call into question what functional The authors identify 539,134 altORFs compared proteins look like, how are they produced and to 68,264 annotated coding regions in the human what can they can do? Because arbitrary mini- transcriptome. They used conservation analysis mum size limits have been imposed on the length across species, ribosomal profiling, examination of coding sequences (CDSs), many RNAs have of protein domains and mass spectrometry and been mistakenly annotated as non-coding in expression in cells to determine the functional many databases. The size and complexity of the impact of the altORFs [116]. A 2019 study examproteome is likely to be greatly underestimated. ined a snapshot of translation in 80 hearts using To determine the protein coding potential of RNA-sequencing (RNA-seq) and ribosome proan RNA transcript, an AUG start codon signify- filing and dissected processes that are under traning translation initiation (also known as the scriptional versus translational control [117]. The Kozak sequence) is identified and amino acid authors show that upstream ORFs influence the spans less than 100 amino acids are classified as efficiency of translation and that there is extena micropeptide [110]. Ribosomal footprinting/ sive translation of lncRNAs with known funcprofiling assesses the binding potential of ribo- tions in human heart, liver, and kidney producing somes to an RNA transcript. Inhibitors of microproteins in vivo. translation are used to stall ribosomes on the ­ It is well established that there is pervasive RNA and deep sequencing of ribosome protected translation outside of canonical coding sequences fragments shows what parts of the transcript are [116, 117]. A 2020 study describes a strategy that bound by ribosomes [111]. Canonical translation combines ribosome profiling, MS-based prooccurs when the 40S ribosomal subunit binds to teomics, microscopy, and CRISPR-based genetic the 5′ cap of an RNA molecule and scans along screens to discover and characterize the widethe transcript until an AUG initiation codon is spread translation of functional microproteins. recognized. The 60S subunit then combines with Human mRNAs can be bicistronic with large and the 40S subunit to form the 80S ribosome, which small proteins functionally cooperating [118]. A translates the ORF into its amino acid sequence subset of lncRNAs was identified that can encode and protein. If the 40S subunit does not dissociate stable, functional proteins, suggesting that they from the RNA it can be used to translate upstream may be misannotated. Some lncRNAs potentially ORF (uORF) to produce a short peptide [112]. have dual roles at the RNA and at the protein levMass spectrometry confirms the presence of pre- els. The authors provide examples of uORFs dicted peptides but there are limitations to the encoding functional peptides, highlighting the method including experimental bias against the diverse cellular roles that uORFs may play

5  LncRNAs and Cardiovascular Disease

beyond translational control of the downstream protein. It is of note that CRISPR mediated gene disruption used to study gene function of the uORF might interfere with the translation or function of the canonical protein translated from the main mRNA. However, when ribosome profiling is combined with the ability of CRISPR to precisely disrupt protein-coding regions we will be able to define and characterize the functional protein-coding capacity of any genome. A family of micropeptides was identified that controls Ca2+ handling by inhibiting SERCA (sarcoplasmic reticulum calcium ATPase) activity in muscle. Ca2+ handling plays a key role in regulating muscle contractility particularly in the heart. PLN (Phospholamban) (5  kDa) and SLN (Sarcolipin) (4–6 kDa) are small peptides that are expressed in cardiac and slow skeletal muscle. Because PLN and SLN are absent in fast skeletal muscle this suggests there is an unidentified factor regulating Ca2+ handling and contractility. MLN (Myoregulin) (5  kDa) is a micropeptide encoded in the third exon of an annotated lncRNA and it is exclusively expressed in fast and slow skeletal muscle. The authors use in vitro translation assays to show the coding potential of MLN and a CRISPR/Cas9 based FLAG tagged knockin approach to show the endogenous protein ­ expression of MLN in a C2C12 myoblast cell line. MLN and SERCA colocalize within the SR membrane where MLN directly binds to SERCA. MLN−/− mice showed no morphological abnormalities but the mice were able to run for a longer amount of time and further as well as have more Ca2+ available in SR, showing improved exercise performance and Ca2+ handling by the muscle. MLN regulates Ca2+ signaling by inhibiting the pump activity of SERCA restricting the import of Ca2+ into the cell. SERCA removes cytosolic Ca2+ from contractile proteins, sequestering and storing it in the lumen of the sarcoplasmic reticulum (SR) to allow muscle relaxation. MLN acts to reduce muscle performance by inhibiting the activity of the key calcium pump SERCA.  There is mouse and human homology between both MLN and the lncRNA that it is found within. It is possible that some high performance athletes have SNPs in the MLN gene pre-

85

venting it from binding to SERCA to inhibit its activity resulting in better muscle performance. Using PhyloCSF to identify codon substitution frequencies, another muscle specific micropeptide was identified encoded within a larger lncRNA gene called Dworf. Converse to MLN, Dworf is a 34 amino acid micropeptide that acts as an activator of SERCA within the heart. Dworf has a stronger binding affinity for SERCA than the inhibitor peptide PLN.  Overexpression of Dworf mitigates the contractile dysfunction seen with elevated PLN levels. The studies characterizing MLN and Dworf demonstrate that some annotated lncRNAs can encode micropeptides that are essential for regulating cardiac processes. A lncRNA that encodes a small ORF in its second exon, called SPAAR (small regulatory polypeptide of amino acid response) was identified in the differentiation of human embryonic stem cell into ECs [119]. The lncRNA and SPAAR exhibit opposing roles in angiogenesis (growth of new blood vessels) with the lncRNA being anti-­ angiogenic and SPAAR pro-angiogenic. SPAAR overexpression results in reduction in endothelial barrier integrity. It binds to the actin binding protein SYNE1 suggesting its role in promoting angiogenesis is mediated through control of the actin cytoskeleton. The systematic identification of functional short CDSs remains challenging and the cellular functions of identified peptide products remains relatively unexplored. It is clear that the microproteome is a largely unannotated reservoir full of biological insights that can be applied to the pathways associated with CVD.

5.4

 ncRNA SNPs Associated L with CVD

Non-coding RNAs play important roles in the regulation of genes involved in CVD [22–25, 120–122] and this can be influenced by genetic variation. The vast majority of SNPs are found in non-coding regions of the genome including within lncRNA genes, promoters, enhancers, splice sites and translation initiation sites [123]. This is thought to be due to the rate at which these non-coding regions evolve, since they do

86

not need to maintain an open reading frame for translation, they are not subject to the pressures of conservation. It is thought that what links these non-coding regions to phenotypes are processes that control how genes are regulated. A SNP may potentially perturb transcription factor binding or structural motifs and disrupt its function thereby leading to disease. lncRNAs have been implicated in modifying the epigenetic landscape of the cell and SNPs in lncRNA genes could potentially result in abnormal epigenetic modulations and affect the development of CVD. Differential histone methylation and acetylation patterns have been identified in patients with CVD [124]. The holy grail of genome wide association studies (GWAS) is to identify SNPs in people susceptible to CVD using DNA sequencing before they have an adverse event and intervening treatments can be administered. ANRIL (Antisense Noncoding RNA in the INK4 Locus  – also known as CDKN2B-AS or CDKN2B-AS1) is a 3.8 kb lncRNA expressed in endothelial, smooth muscle and mononuclear inflammatory cells like macrophages. It was identified through GWAS that linked a susceptibility locus on human chromosome 9p21 containing the lncRNA to CAD [120, 121, 125, 126]. Deletion of this risk locus affected the expression of 38 of the 91 known risk genes associated with CAD identified in the GWAS. The key SNPs in the locus associated with CVD risk, rs1333049, rs1075724, rs2383206 and rs2383207 are found within the region spanning ANRIL signifying its potential involvement in the risk association [127]. rs1333049 is located within the 3′ end of ANRIL and is associated with reduced ANRIL expression and increased VSMC proliferation [128]. Expression of ANRIL has been associated with risk for atherosclerosis and MI [120, 129]. The transcript is alternatively spliced into at least 20 isoforms depending on the cell type and pathology including a circular form, which can be found in serum [130]. Two ANRIL isoforms have been identified in testes, five in ECs and three in lung [131]. Several RNA and protein-­ binding partners have been identified to interact with ANRIL A cis-regulatory mechanism was described for ANRIL where it binds to the PRC2

E. J. Hennessy

component SUZ12, which also binds to the lncRNAs Braveheart, CARMEN and APOA1-AS. The interaction with SUZ12 remodels local chromatin and silences the neighboring genes CDKN2A and CDKN2BB, key players in blocking cell cycle progression [132]. circANRIL regulates ribosomal RNA maturation by binding to PES1 (pescadillo homologue 1) a 60S pre-ribosomal assembly factor. This interaction induces a stress response and activates p53 and decreases cell proliferation [122]. The authors suggest this is a cardioprotective response to atherosclerotic signals. Despite the correlation studies, the mechanism and biological function of ANRIL remains poorly understood because of the numerous variants, low expression levels and the repetitive sequences (LINE, SINE and Alu elements) found throughout making it difficult to manipulate with knockdown and overexpression constructs [125]. TRIBAL (TRIB1 associated locus) is a lncRNA found in linkage disequilibrium with TRIB1, a gene essential for regulating plasma lipid levels [133–136]. r22001844 displays a strong association with CAD and TRIB1 and TRIBAL mRNA expression are altered in its presence [133]. rs2001844 is located in the 5′ region of TRIBAL and alters its promoter activity. It is associated with plasma triglyceride and HDL levels and is expressed in both macrophages and hepatocytes. TRIBAL stabilizes TRIB1 mRNA through MAPK (mitogen activated protein kinase) signaling [137]. The lncRNA MIAT (myocardial infarction associated transcript) is located in a MI susceptibility locus on chromosome 22 [138]. Six independent SNPs in the MIAT gene were identified further conferring genetic susceptibility to MI [139]. One of these SNPs is located in the fifth exon of MIAT and alters its ability to bind to an uncharacterized nuclear protein, possibly a transcription factor. Another study demonstrated that MIAT expression is decreased in peripheral blood cells and platelets of patients who have experienced acute MI [140]. The JUPITER (Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin) clinical trial was

5  LncRNAs and Cardiovascular Disease

the largest GWAS performed examining the use of a statin on approximately 7000 participants. rs6924995 was identified in the JUPITER study and was significantly associated with plasma LDL cholesterol [141]. This SNP is 10 kilobases from MYLIP, a gene regulating the LDL receptor. The association was originally attributed to this location but there has been no evidence that this SNP affects MYLIP expression or function. rs26924995 is located within a pseudogene called RP1-13D10.2. Expression of RP1-13D10.2 was significantly different between cells isolated from high and low LDL cholesterol responders and treated with statin. A follow-up study revealed that overexpression of either allele associated with rs26924995 (A or G) increased LDL receptor expression as well as labeled LDL uptake by hepatocytes. The A allele was typically associated with greater LDL changes than the G allele [142]. The mechanism for the regulation of the LDLR by RP1-13D10.2 and the association with the SNP remains elusive but evidence from the JUPITER study and in  vitro data suggest it will have a role in controlling uptake of LDL cholesterol. LASER (Long non-coding RNA in lipid Associated Single nucleotide polymorphism gEne Region; ENSG00000237937) was identified in a search of the lncSNP database for lncRNAs located near SNPs associated with lipid traits [143]. rs486394 is located within the second intron of the LASER gene on chromosome 11 near the APO gene cluster. The authors found that LASER was highly expressed in both hepatocytes and peripheral blood mononuclear cells (PBMCs). Transfection of the HegG2 hepatocyte cell line with siRNA targeting LASER significantly decreased intracellular cholesterol levels and HNF-1a and PCSK9 mRNA and protein levels. HNF-1a is highly expressed in the liver and is an essential regulator of genes involved in bile acid and cholesterol metabolism such as PCSK9, which regulates the cell surface expression of the LDL receptor. Inhibition of PCSK9 using monoclonal antibodies is being explored as a therapeutic strategy for reducing plasma LDL cholesterol levels. The RNA-protein interaction prediction tool catRAPID predicted binding of LASER to

87

the histone demethylase LSD1 (lysine (K)-specific demethylase 1A). This interaction results in the repression of LSD1 binding and H3K4me mediated demethylation of the HNF-1a promoter resulting in HNF-1a activation and PCSK9 transcription. Because the majority of SNPs are found within non-coding regions of the genome and as evidenced by the studies described here, whole genome sequencing must be used to identify susceptible patients not simply exome sequencing which only captures protein coding genes. There needs to be a better understanding of the impact of SNPs on both the secondary structure of lncRNAs as well as their ability to interact with binding partners. Both of these factors will lead to effects on downstream signaling pathways and the ability to use lncRNAs to therapeutically control the pathologies associated with CVD.

5.5

LncRNAs as Biomarkers for CVD Associated Pathologies

Because of the diversity of cellular processes implicated in lncRNA function and the cell type specificity of their patterns of expression they provide prospective targets for therapeutic application however few have been reported thus far. Many studies identify a lncRNA in a disease scenario but do not translate expression changes into an opportunity for disease intervention through the modulation of the lncRNA. One of the most prominent therapeutic applications for lncRNAs is their use as biomarkers for disease, predicting onset or severity. lncRNAs can be detected in whole blood in PBMCs and in biofluids like saliva, plasma and urine and display dynamic changes in response to disease [144–146]. They can be encapsulated in exosomes or inside apoptotic bodies released from dying cells [147]. miRNA studies have shown that encapsulated miRNAs in the circulation can be taken up by distal cells and exert their effects on target mRNAs in the new host cell [148]. Studies like these demonstrate that non-coding RNAs can act as more than a biomarker for disease, they can be

88

encapsulated into nanoparticles and potentially used therapeutically. Several lncRNAs have been identified in different studies as potential biomarkers for severity of MI.  Plasma LIPCAR (long intergenic non-­ coding RNA predicting cardiac remodeling) expression is associated with left ventricular (LV) remodeling after MI and increased risk of developing heart failure. There is a higher abundance of LIPCAR in the plasma of patients with LV remodeling after MI compared to MI patients without LV remodeling. The magnitude of circulating LIPCAR is associated with an increased risk of cardiovascular death in chronic heart failure patients. Expression in plasma was downregulated early after MI but upregulated during later stages post-MI [149]. LIPCAR was also identified in a study examining plasma levels of a selected set of cardiovascular related lncRNAs for use as novel biomarkers of CAD in a Chinese population. LIPCAR and H19 were both significantly increased in patients with CAD [150]. H19 was also increased in plasma of patients with pulmonary arterial hypertension [151]. A study of PBMCs taken from 93 CAD patients and 48 healthy controls identified six lncRNAs, uc010yfd.1, RNA147299, ENST00000444488.1, ASO3973, ENST00000602558.1 and ENST00000561165.1 to be differentially expressed in response to MI. Using a stepwise selection model found that ENST00000444488.1 and uc010yfd.1 offered the most reliable diagnostic ability [152]. Another study examining PBMCs from patients with MI identified the differential expression of the lncRNAs uc002ddj.1, ENST00000509938.1, ENST00000581794.1 and NR_047662.2. ENST00000509938.1 and ENST00000581794.1 increased significantly and uc002ddj.1 and NR_047662.2 decreased. This study performed a comparison between level of detection of the lncRNAs measured in PBMCs and plasma and found an obvious difference in terms of quality of RNA and the abundance of RNA in PBMCs, which was higher than in plasma. Only 60 lncRNAs were found differentially expressed in both [153]. Coromarker was identified in a microarray screen comparing plasma from 20

E. J. Hennessy

CAD patients with 20 healthy controls for differentially expressed lncRNAs. Its elevated expression was validated in a second cohort of 221 CAD patients and 187 healthy control samples [154]. The lncRNA GAS5 is increased in atherosclerotic plaques [155] and was found to be released from THP-1 derived macrophages treated with oxLDL.  Exosomes isolated from THP-1 derived macrophages with GAS5 knocked down were able to reduce apoptosis of endothelial cells in vitro illustrating the ability of lncRNA containing exosomes to be taken up by other cells to affect signaling pathways [156]. This study signifies the potential of exosomes to be used as a treatment for atherosclerosis. Each of the listed lncRNAs expressed in either PBMCs or plasma isolated from patients who have experienced MI are potential biomarkers for severity of CAD but the issue of RNA quality, expression levels and whether to use PBMCs or plasma for detection remains. Circular lncRNAs like circANRIL are more stable than their linear counterparts due to their resistance to degradation by ribonucleases and have been found in circulation [122, 130]. As mentioned earlier in this review, circANRIL plays an atheroprotective role and was identified in PBMCs isolated from patients with CAD. Carriers of the CAD-protective haplotype at the 9p21 locus had increased expression of circANRIL. circANRIL has also been identified in serum collected from children suffering from Kawasaki disease, an inflammatory disease of the vasculature. CircANRIL levels were lower in patients in the acute phase of the disease compared to healthy controls and increased in patients after receiving immunoglobulin therapy to reduce artery abnormalities [157]. Studies of extracellular miRNAs have shown their association with HDL in the circulation [158–160]. Epidemiological studies have revealed that plasma HDL and cardiovascular outcomes are inversely related, implying that modulation of HDL might alter cardiovascular risk; however recent analysis of such studies has shown that the relationship may be U-shaped and too much HDL can result in morbidity [161]. It is thought that the quality and not the quantity of HDL determines its beneficial effects and that

5  LncRNAs and Cardiovascular Disease

dysfunctional HDL is associated with the acute coronary syndrome [162]. An interesting therapeutic approach against CVD could be to modulate the concentration of a specific lncRNA by encapsulating it in an HDL particle, thus increasing beneficial HDL levels in the circulation. Ultimately, the effectiveness of lncRNA-based therapeutics will ultimately depend on the strategy used to manipulate its expression and the efficiency of delivery to target cells.

5.6

 he Future of LncRNAs T in CVD

There are many obstacles in the quest for uncovering novel lncRNAs and the mechanisms they use to exert their effects. One of the primary impediments lies in their lack of conservation. Because they rapidly evolve, they have highly variable sequences between species making translational studies between mouse and human difficult. There have been arguments about why we should care about lncRNAs that are exclusively expressed in mouse. Evolution of these transcripts is a very intriguing question and one that is fundamental to understanding our genomic make up. On the other hand, lncRNAs like NEAT1, which has a role in all four cell types examined in this review is highly conserved and should be used as a model to develop techniques for the identification of novel lncRNAs. The ­primary method used to identify novel and differentially expressed lncRNAs in various environmental stimuli or disease states is RNA-­seq. Microarrays containing probes for annotated genes are also frequently used however they can introduce certain biases into the data such as skewing G-C rich or repeat regions often found in lncRNAs from PCR based amplification that RNA-seq does not require. RNA-seq offers more high-throughput and comprehensive coverage of the transcriptome, revealing splicing sites and different isoforms of a gene, but the question remains of how deep the sequencing coverage should be to identify more lowly expressed lncRNAs that could be highly differentially expressed in different tissues or disease states.

89

In order to understand the mechanism being used by lncRNAs, bioinformatic and experimental approaches to identify binding partners of lncRNAs needs to be more easily executed. Research groups need to engage in more interdisciplinary collaborations and readily integrate computational methods with in vitro and in vivo methods. We need to further develop techniques like the sequence comparison method developed to asses linear sequence relationships between lncRNAs identify common short motifs called k-mers [163]. CREPE (correlative recurrent expression of predicted elements) is a computational approach that tracks the expression, cell-­ type specificity, subcellular localization, differential expression, correlation with neighboring genes, and 3D genome interactions. CREPE calculates a rankable score for lncRNAs to predict potential functions like transcriptional regulation, enhancer associations and scaffold function. One of the greatest advancements in biology in the last decade was the discovery of CRISPR/Cas9 (clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9) and in particular CRISPRa (activation) and CRISPRi (interference). This technique will enable a better understanding of the function of lncRNAs through the control of gene transcription with very few off target effects [164]. We are only just beginning to break the surface in terms of identifying relevant lncRNAs in various disease states and the cleaner our techniques become the better the odds at defeating a disease like CVD.

References 1. Nikpay M et  al (2015) A comprehensive 1,000 Genomes-based genome-wide association meta-­ analysis of coronary artery disease. Nat Genet 47:1121–1130. https://doi.org/10.1038/ng.3396 2. Djebali S et  al (2012) Landscape of transcription in human cells. Nature 489:101–108. https://doi. org/10.1038/nature11233 3. Hangauer MJ, Vaughn IW, McManus MT (2013) Pervasive transcription of the human genome produces thousands of previously unidentified long intergenic noncoding RNAs. PLoS Genet 9:e1003569. https://doi.org/10.1371/journal.pgen.1003569

90 4. Iyer MK et al (2015) The landscape of long noncoding RNAs in the human transcriptome. Nat Genet 47:199–208. https://doi.org/10.1038/ng.3192 5. Hon CC et  al (2017) An atlas of human long non-coding RNAs with accurate 5′ ends. Nature 543:199–204. https://doi.org/10.1038/nature21374 6. Ruan X et  al (2020) In vivo functional analysis of non-conserved human lncRNAs associated with cardiometabolic traits. Nat Commun 11:45. https://doi. org/10.1038/s41467-­019-­13688-­z 7. Taft RJ, Pheasant M, Mattick JS (2007) The relationship between non-protein-coding DNA and eukaryotic complexity. BioEssays 29:288–299. https://doi. org/10.1002/bies.20544 8. Awan HM, Shah A, Rashid F, Shan G (2017) Primate-­ specific long non-coding RNAs and microRNAs. Genomics Proteomics Bioinformatics 15:187–195. https://doi.org/10.1016/j.gpb.2017.04.002 9. Ulitsky I (2016) Evolution to the rescue: using comparative genomics to understand long non-coding RNAs. Nat Rev Genet 17:601–614. https://doi. org/10.1038/nrg.2016.85 10. Yen ZC, Meyer IM, Karalic S, Brown CJ (2007) A cross-species comparison of X-chromosome inactivation in Eutheria. Genomics 90:453–463. https:// doi.org/10.1016/j.ygeno.2007.07.002 11. Hennessy EJ (2017) Cardiovascular disease and long noncoding RNAs: tools for unraveling the mystery Lnc-ing RNA and phenotype. Circ Cardiovasc Genet 10:e001556. https://doi.org/10.1161/ CIRCGENETICS.117.001556 12. Amaral PP, Mattick JS (2008) Noncoding RNA in development. Mamm Genome 19:454–492. https:// doi.org/10.1007/s00335-­008-­9136-­7 13. Wang KC, Chang HY (2011) Molecular mechanisms of long noncoding RNAs. Mol Cell 43:904– 914. https://doi.org/10.1016/j.molcel.2011.08.018 14. Mumbach MR et  al (2019) HiChIRP reveals RNA-associated chromosome conformation. Nat Methods 16:489–492. https://doi.org/10.1038/ s41592-­019-­0407-­x 15. Isoda T et  al (2017) Non-coding transcription instructs chromatin folding and compartmentalization to dictate enhancer-Ppomoter communication and T cell fate. Cell 171:103–119.e118. https://doi. org/10.1016/j.cell.2017.09.001 16. Bosson AD, Zamudio JR, Sharp PA (2014) Endogenous miRNA and target concentrations determine susceptibility to potential ceRNA competition. Mol Cell 56:347–359. https://doi.org/10.1016/j. molcel.2014.09.018 17. Salmena L, Poliseno L, Tay Y, Kats L, Pandolfi PP (2011) A ceRNA hypothesis: the Rosetta Stone of a hidden RNA language? Cell 146:353–358. https:// doi.org/10.1016/j.cell.2011.07.014 18. Dinger ME et  al (2008) Long noncoding RNAs in mouse embryonic stem cell pluripotency and differentiation. Genome Res 18:1433–1445. https://doi. org/10.1101/gr.078378.108

E. J. Hennessy 19. Jia H et al (2010) Genome-wide computational identification and manual annotation of human long noncoding RNA genes. RNA 16:1478–1487. https://doi. org/10.1261/rna.1951310 20. Ponjavic J, Oliver PL, Lunter G, Ponting CP (2009) Genomic and transcriptional co-localization of protein-­ coding and long non-coding RNA pairs in the developing brain. PLoS Genet 5:e1000617. https://doi.org/10.1371/journal.pgen.1000617 21. Ponjavic J, Ponting CP, Lunter G (2007) Functionality or transcriptional noise? Evidence for selection within long noncoding RNAs. Genome Res 17:556–565. https://doi.org/10.1101/gr.6036807 22. Sauvageau M et al (2013) Multiple knockout mouse models reveal lincRNAs are required for life and brain development. eLife 2:e01749. https://doi. org/10.7554/eLife.01749 23. Grote P, Herrmann BG (2013) The long non-coding RNA Fendrr links epigenetic control mechanisms to gene regulatory networks in mammalian embryogenesis. RNA Biol 10:1579–1585 24. Grote P et  al (2013) The tissue-specific lncRNA Fendrr is an essential regulator of heart and body wall development in the mouse. Dev Cell 24:206– 214. https://doi.org/10.1016/j.devcel.2012.12.012 25. Klattenhoff CA et  al (2013) Braveheart, a long noncoding RNA required for cardiovascular lineage commitment. Cell 152:570–583. https://doi. org/10.1016/j.cell.2013.01.003 26. Hennessy EJ et al (2019) The long noncoding RNA CHROME regulates cholesterol homeostasis in primate. Nat Metab 1:98–110. https://doi.org/10.1038/ s42255-­018-­0004-­9 27. Sallam T et al (2016) Feedback modulation of cholesterol metabolism by the lipid-responsive non-­ coding RNA LeXis. Nature 534:124–128. https:// doi.org/10.1038/nature17674 28. Sallam T et  al (2018) Transcriptional regulation of macrophage cholesterol efflux and atherogenesis by a long noncoding RNA.  Nat Med 24:304–312. https://doi.org/10.1038/nm.4479 29. Barker DJ (2000) In utero programming of cardiovascular disease. Theriogenology 53:555–574. https://doi.org/10.1016/s0093-­691x(99)00258-­7 30. Dobbing J (1993) Fetal nutrition and cardiovascular disease in adult life. Lancet 341:1421–1422 31. Wang T et  al (2019) Congenital heart disease and risk of cardiovascular disease: a meta-analysis of cohort studies. J Am Heart Assoc 8:e012030. https:// doi.org/10.1161/JAHA.119.012030 32. Niwa K (2019) Metabolic syndrome in adult congenital heart disease. Korean Circ J 49:691–708. https://doi.org/10.4070/kcj.2019.0187 33. Ulitsky I, Shkumatava A, Jan CH, Sive H, Bartel DP (2011) Conserved function of lincRNAs in vertebrate embryonic development despite rapid sequence evolution. Cell 147:1537–1550. https:// doi.org/10.1016/j.cell.2011.11.055 34. Wang KC et  al (2011) A long noncoding RNA maintains active chromatin to coordinate homeotic

5  LncRNAs and Cardiovascular Disease gene expression. Nature 472:120–124. https://doi. org/10.1038/nature09819 35. Rinn JL, Chang HY (2012) Genome regulation by long noncoding RNAs. Annu Rev Biochem 81:145–166. https://doi.org/10.1146/ annurev-­biochem-­051410-­092902 36. Munteanu MC et al (2020) Long non-coding RNA FENDRR regulates IFNgamma-induced M1 phenotype in macrophages. Sci Rep 10:13672. https://doi. org/10.1038/s41598-­020-­70633-­7 37. Xue Z et  al (2016) A G-rich motif in the lncRNA braveheart interacts with a zinc-finger transcription factor to specify the cardiovascular lineage. Mol Cell 64:37–50. https://doi.org/10.1016/j. molcel.2016.08.010 38. Kim DN et al (2020) Zinc-finger protein CNBP alters the 3-D structure of lncRNA Braveheart in solution. Nat Commun 11:148. https://doi.org/10.1038/ s41467-­019-­13942-­4 39. Ounzain S et  al (2015) CARMEN, a human super enhancer-associated long noncoding RNA controlling cardiac specification, differentiation and homeostasis. J Mol Cell Cardiol 89:98–112. https:// doi.org/10.1016/j.yjmcc.2015.09.016 40. Honold L, Nahrendorf M (2018) Resident and monocyte-­ derived macrophages in cardiovascular disease. Circ Res 122:113–127. https://doi. org/10.1161/CIRCRESAHA.117.311071 41. Zelcer N, Tontonoz P (2006) Liver X receptors as integrators of metabolic and inflammatory signaling. J Clin Invest 116:607–614. https://doi.org/10.1172/ JCI27883 42. Ignatova ID, Angdisen J, Moran E, Schulman IG (2013) Differential regulation of gene expression by LXRs in response to macrophage cholesterol loading. Mol Endocrinol 27:1036–1047. https://doi. org/10.1210/me.2013-­1051 43. Schultz JR et al (2000) Role of LXRs in control of lipogenesis. Genes Dev 14:2831–2838. https://doi. org/10.1101/gad.850400 44. Xiao X, Song BL (2013) SREBP: a novel therapeutic target. Acta Biochim Biophys Sin 45:2–10. https://doi.org/10.1093/abbs/gms112 45. Sokolov A, Radhakrishnan A (2010) Accessibility of cholesterol in endoplasmic reticulum membranes and activation of SREBP-2 switch abruptly at a common cholesterol threshold. J Biol Chem 285:29480– 29490. https://doi.org/10.1074/jbc.M110.148254 46. Radhakrishnan A, Goldstein JL, McDonald JG, Brown MS (2008) Switch-like control of SREBP-2 transport triggered by small changes in ER cholesterol: a delicate balance. Cell Metab 8:512–521. https://doi.org/10.1016/j.cmet.2008.10.008 47. Yvan-Charvet L, Wang N, Tall AR (2010) Role of HDL, ABCA1, and ABCG1 transporters in cholesterol efflux and immune responses. Arterioscler Thromb Vasc Biol 30:139–143. https://doi. org/10.1161/ATVBAHA.108.179283 48. Ouimet M et  al (2016) miRNA targeting of oxysterol-­ binding protein-like 6 regulates choles-

91 terol trafficking and efflux. Arterioscler Thromb Vasc Biol 36:942–951. https://doi.org/10.1161/ ATVBAHA.116.307282 49. Li Y, Sun T, Shen S, Wang L, Yan J (2019) LncRNA DYNLRB2-2 inhibits THP-1 macrophage foam cell formation by enhancing autophagy. Biol Chem 400(8):1047–1057. https://doi.org/10.1515/ hsz-­2018-­0461 50. Hu YW et  al (2014) A lincRNA-DYNLRB2-2/ GPR119/GLP-1R/ABCA1-dependent signal transduction pathway is essential for the regulation of cholesterol homeostasis. J Lipid Res 55:681–697. https://doi.org/10.1194/jlr.M044669 51. Yan Y, Song D, Wu J, Wang J (2020) Long non-­ coding RNAs link oxidized low-density lipoprotein with the inflammatory response of macrophages in atherogenesis. Front Immunol 11:24. https://doi. org/10.3389/fimmu.2020.00024 52. Huang-Fu N, Cheng JS, Wang Y, Li ZW, Wang SH (2018) Neat1 regulates oxidized low-density lipoprotein-induced inflammation and lipid uptake in macrophages via paraspeckle formation. Mol Med Rep 17:3092–3098. https://doi.org/10.3892/ mmr.2017.8211 53. Hirose T et al (2014) NEAT1 long noncoding RNA regulates transcription via protein sequestration within subnuclear bodies. Mol Biol Cell 25:169– 183. https://doi.org/10.1091/mbc.E13-­09-­0558 54. Prinz F, Kapeller A, Pichler M, Klec C (2019) The implications of the long non-coding RNA NEAT1 in non-cancerous diseases. Int J Mol Sci 20:627. https://doi.org/10.3390/ijms20030627 55. Hu YW et  al (2015) RP5-833A20.1/miR-382-5p/ NFIA-dependent signal transduction pathway contributes to the regulation of cholesterol homeostasis and inflammatory reaction. Arterioscler Thromb Vasc Biol 35:87–101. https://doi.org/10.1161/ ATVBAHA.114.304296 56. Reddy MA et al (2014) Regulation of inflammatory phenotype in macrophages by a diabetes-induced long noncoding RNA.  Diabetes 63:4249–4261. https://doi.org/10.2337/db14-­0298 57. Schmidt E et al (2018) LincRNA H19 protects from dietary obesity by constraining expression of monoallelic genes in brown fat. Nat Commun 9:3622. https://doi.org/10.1038/s41467-­018-­05933-­8 58. Raveh E, Matouk IJ, Gilon M, Hochberg A (2015) The H19 Long non-coding RNA in cancer initiation, progression and metastasis  – a proposed unifying theory. Mol Cancer 14:184. https://doi.org/10.1186/ s12943-­015-­0458-­2 59. Gabory A et al (2009) H19 acts as a trans regulator of the imprinted gene network controlling growth in mice. Development 136:3413–3421. https://doi. org/10.1242/dev.036061 60. Poole RL et  al (2012) Beckwith-Wiedemann syndrome caused by maternally inherited mutation of an OCT-binding motif in the IGF2/H19-imprinting control region, ICR1. Eur J Hum Genet 20:240–243. https://doi.org/10.1038/ejhg.2011.166

92 61. Wevrick R, Kerns JA, Francke U (1994) Identification of a novel paternally expressed gene in the Prader-­ Willi syndrome region. Hum Mol Genet 3:1877– 1882. https://doi.org/10.1093/hmg/3.10.1877 62. Zhang Y, Tycko B (1992) Monoallelic expression of the human H19 gene. Nat Genet 1:40–44. https://doi. org/10.1038/ng0492-­40 63. Li X et  al (2016) LncRNA H19/miR-675 axis regulates cardiomyocyte apoptosis by targeting VDAC1  in diabetic cardiomyopathy. Sci Rep 6:36340. https://doi.org/10.1038/srep36340 64. Hu Y, Li S, Zou Y (2019) Knockdown of lncRNA H19 relieves LPS-induced damage by modulating miR-130a in osteoarthritis. Yonsei Med J 60:381– 388. https://doi.org/10.3349/ymj.2019.60.4.381 65. Zheng H et al (2016) Regulation and mechanism of mouse miR-130a/b in metabolism-related inflammation. Int J Biochem Cell Biol 74:72–83. https://doi. org/10.1016/j.biocel.2016.02.021 66. Han DK, Khaing ZZ, Pollock RA, Haudenschild CC, Liau G (1996) H19, a marker of developmental transition, is reexpressed in human atherosclerotic plaques and is regulated by the insulin family of growth factors in cultured rabbit smooth muscle cells. J Clin Invest 97:1276–1285. https://doi. org/10.1172/JCI118543 67. Han Y, Ma J, Wang J, Wang L (2018) Silencing of H19 inhibits the adipogenesis and inflammation response in ox-LDL-treated Raw264.7 cells by up-­ regulating miR-130b. Mol Immunol 93:107–114. https://doi.org/10.1016/j.molimm.2017.11.017 68. Schwartz SM, Murry CE (1998) Proliferation and the monoclonal origins of atherosclerotic lesions. Annu Rev Med 49:437–460. https://doi.org/10.1146/ annurev.med.49.1.437 69. Campbell JH, Campbell GR (1994) The role of smooth muscle cells in atherosclerosis. Curr Opin Lipidol 5:323–330. https://doi. org/10.1097/00041433-­199410000-­00003 70. Sun Y et al (2019) LncRNA H19 promotes vascular inflammation and abdominal aortic aneurysm formation by functioning as a competing endogenous RNA.  J Mol Cell Cardiol 131:66–81. https://doi. org/10.1016/j.yjmcc.2019.04.004 71. Pan JX (2017) LncRNA H19 promotes atherosclerosis by regulating MAPK and NF-kB signaling pathway. Eur Rev Med Pharmacol Sci 21:322–328 72. Ahmed ASI et  al (2018) Long noncoding RNA NEAT1 (nuclear paraspeckle assembly transcript 1) is critical for phenotypic switching of vascular smooth muscle cells. Proc Natl Acad Sci U S A 115:E8660–E8667. https://doi.org/10.1073/ pnas.1803725115 73. Wu G et al (2014) LincRNA-p21 regulates neointima formation, vascular smooth muscle cell proliferation, apoptosis, and atherosclerosis by enhancing p53 activity. Circulation 130:1452–1465. https://doi. org/10.1161/CIRCULATIONAHA.114.011675 74. Guevara NV, Kim HS, Antonova EI, Chan L (1999) The absence of p53 accelerates atherosclerosis by

E. J. Hennessy increasing cell proliferation in vivo. Nat Med 5:335– 339. https://doi.org/10.1038/6585 75. Bell RD et al (2014) Identification and initial functional characterization of a human vascular cell-­ enriched long noncoding RNA. Arterioscler Thromb Vasc Biol 34:1249–1259. https://doi.org/10.1161/ ATVBAHA.114.303240 76. Ballantyne MD et al (2016) Smooth muscle enriched long noncoding RNA (SMILR) regulates cell proliferation. Circulation 133:2050–2065. https://doi. org/10.1161/CIRCULATIONAHA.115.021019 77. Reglero-Real N, Colom B, Bodkin JV, Nourshargh S (2016) Endothelial cell junctional adhesion molecules: role and regulation of expression in inflammation. Arterioscler Thromb Vasc Biol 36:2048–2057. https://doi.org/10.1161/ATVBAHA.116.307610 78. Huang Y, Wang L, Mao Y, Nan G (2019) Long noncoding RNA-H19 contributes to atherosclerosis and induces ischemic stroke via the upregulation of acid phosphatase 5. Front Neurol 10:32. https://doi. org/10.3389/fneur.2019.00032 79. Voellenkle C et  al (2016) Implication of long noncoding RNAs in the endothelial cell response to hypoxia revealed by RNA-sequencing. Sci Rep 6:24141. https://doi.org/10.1038/srep24141 80. Zhu A, Chu L, Ma Q, Li Y (2019) Long non-­coding RNA H19 down-regulates miR-181a to facilitate endothelial angiogenic function. Artif Cells Nanomed Biotechnol 47:2698–2705. https://doi.org/ 10.1080/21691401.2019.1634577 81. Hofmann P et  al (2019) Long non-coding RNA H19 regulates endothelial cell aging via inhibition of STAT3 signalling. Cardiovasc Res 115:230–242. https://doi.org/10.1093/cvr/cvy206 82. Michalik KM et  al (2014) Long noncoding RNA MALAT1 regulates endothelial cell function and vessel growth. Circ Res 114:1389–1397. https://doi. org/10.1161/CIRCRESAHA.114.303265 83. Trapani L, Segatto M, Pallottini V (2012) Regulation and deregulation of cholesterol homeostasis: the liver as a metabolic “power station”. World J Hepatol 4:184–190. https://doi.org/10.4254/wjh.v4.i6.184 84. Qin W et  al (2016) A long non-coding RNA, APOA4-AS, regulates APOA4 expression depending on HuR in mice. Nucleic Acids Res 44:6423– 6433. https://doi.org/10.1093/nar/gkw341 85. Halley P et al (2014) Regulation of the apolipoprotein gene cluster by a long noncoding RNA.  Cell Rep 6:222–230. https://doi.org/10.1016/j. celrep.2013.12.015 86. Zhang Z et  al (2020) Collaborative interactions of heterogenous ribonucleoproteins contribute to transcriptional regulation of sterol metabolism in mice. Nat Commun 11:984. https://doi.org/10.1038/ s41467-­020-­14711-­4 87. Lan X et  al (2015) Identification of differentially expressed genes related to metabolic syndrome induced with high-fat diet in E3 rats. Exp Biol Med 240:235–241. https://doi. org/10.1177/1535370214554531

5  LncRNAs and Cardiovascular Disease 88. Lan X et  al (2016) A novel long noncoding RNA Lnc-HC binds hnRNPA2B1 to regulate expressions of Cyp7a1 and Abca1  in hepatocytic cholesterol metabolism. Hepatology 64:58–72. https://doi. org/10.1002/hep.28391 89. Lan X et  al (2019) Long noncoding RNA lnc-HC regulates PPARgamma-mediated hepatic lipid metabolism through miR-130b-3p. Mol Ther Nucleic Acids 18:954–965. https://doi.org/10.1016/j. omtn.2019.10.018 90. Liu X et al (2018) Long non-coding RNA NEAT1-­ modulated abnormal lipolysis via ATGL drives hepatocellular carcinoma proliferation. Mol Cancer 17:90. https://doi.org/10.1186/s12943-­018-­0838-­5 91. Jin SS, Lin XF, Zheng JZ, Wang Q, Guan HQ (2019) LncRNA NEAT1 regulates fibrosis and inflammatory response induced by nonalcoholic fatty liver by regulating miR-506/GLI3. Eur Cytokine Netw 30:98–106. https://doi.org/10.1684/ecn.2019.0432 92. Javidan A et al (2019) miR-146a deficiency accelerates hepatic inflammation without influencing diet-­ induced obesity in mice. Sci Rep 9:12626. https:// doi.org/10.1038/s41598-­019-­49090-­4 93. Huang H et al (2018) Rho-kinase/AMPK axis regulates hepatic lipogenesis during overnutrition. J Clin Invest 128:5335–5350. https://doi.org/10.1172/ JCI63562 94. Chen X, Tan XR, Li SJ, Zhang XX (2019) LncRNA NEAT1 promotes hepatic lipid accumulation via regulating miR-146a-5p/ROCK1  in nonalcoholic fatty liver disease. Life Sci 235:116829. https://doi. org/10.1016/j.lfs.2019.116829 95. Qu L et  al (2016) Exosome-transmitted lncARSR promotes sunitinib resistance in renal cancer by acting as a competing endogenous RNA.  Cancer Cell 29:653–668. https://doi.org/10.1016/j. ccell.2016.03.004 96. Li Y, Ye Y, Feng B, Qi Y (2017) Long noncoding RNA lncARSR promotes doxorubicin resistance in hepatocellular carcinoma via modulating PTEN-­ PI3K/Akt pathway. J Cell Biochem 118:4498–4507. https://doi.org/10.1002/jcb.26107 97. Zhang M, Chi X, Qu N, Wang C (2018) Long noncoding RNA lncARSR promotes hepatic lipogenesis via Akt/SREBP-1c pathway and contributes to the pathogenesis of nonalcoholic steatohepatitis. Biochem Biophys Res Commun 499:66–70. https:// doi.org/10.1016/j.bbrc.2018.03.127 98. Chi Y, Gong Z, Xin H, Wang Z, Liu Z (2020) Long noncoding RNA lncARSR promotes nonalcoholic fatty liver disease and hepatocellular carcinoma by promoting YAP1 and activating the IRS2/AKT pathway. J Transl Med 18:126. https://doi.org/10.1186/ s12967-­020-­02225-­y 99. Jeong SH et al (2018) Hippo-mediated suppression of IRS2/AKT signaling prevents hepatic steatosis and liver cancer. J Clin Invest 128:1010–1025. https://doi.org/10.1172/JCI95802 100. Li D et  al (2017) Identification of a novel human long non-coding RNA that regulates hepatic lipid

93 metabolism by inhibiting SREBP-1c. Int J Biol Sci 13:349–357. https://doi.org/10.7150/ijbs.16635 101. Waris G, Felmlee DJ, Negro F, Siddiqui A (2007) Hepatitis C virus induces proteolytic cleavage of sterol regulatory element binding proteins and stimulates their phosphorylation via oxidative stress. J Virol 81:8122–8130. https://doi.org/10.1128/ JVI.00125-­07 102. Li D et al (2018) Long noncoding RNA HR1 participates in the expression of SREBP1c through phosphorylation of the PDK1/AKT/FoxO1 pathway. Mol Med Rep 18:2850–2856. https://doi.org/10.3892/ mmr.2018.9278 103. Puthanveetil P, Chen S, Feng B, Gautam A, Chakrabarti S (2015) Long non-coding RNA MALAT1 regulates hyperglycaemia induced inflammatory process in the endothelial cells. J Cell Mol Med 19:1418–1425. https://doi.org/10.1111/ jcmm.12576 104. Yan B et  al (2014) Aberrant expression of long noncoding RNAs in early diabetic retinopathy. Invest Ophthalmol Vis Sci 55:941–951. https://doi. org/10.1167/iovs.13-­13221 105. Watts R et al (2014) Myostatin inhibits proliferation and insulin-stimulated glucose uptake in mouse liver cells. Biochem Cell Biol 92:226–234. https://doi. org/10.1139/bcb-­2014-­0004 106. Yan C, Chen J, Chen N (2016) Long noncoding RNA MALAT1 promotes hepatic steatosis and insulin resistance by increasing nuclear SREBP-1c protein stability. Sci Rep 6:22640. https://doi.org/10.1038/ srep22640 107. Yang L et al (2016) Integrative transcriptome analyses of metabolic responses in mice define pivotal lncRNA metabolic regulators. Cell Metab 24:627– 639. https://doi.org/10.1016/j.cmet.2016.08.019 108. Li P et al (2015) A liver-enriched long non-coding RNA, lncLSTR, regulates systemic lipid metabolism in mice. Cell Metab 21:455–467. https://doi. org/10.1016/j.cmet.2015.02.004 109. Rohrig H, Schmidt J, Miklashevichs E, Schell J, John M (2002) Soybean ENOD40 encodes two peptides that bind to sucrose synthase. Proc Natl Acad Sci U S A 99:1915–1920. https://doi.org/10.1073/ pnas.022664799 110. Sousa ME, Farkas MH (2018) Micropeptide. PLoS Genet 14:e1007764. https://doi.org/10.1371/journal. pgen.1007764 111. Calviello L, Ohler U (2017) Beyond read-counts: Ribo-seq data analysis to understand the functions of the transcriptome. Trends Genet 33:728–744. https://doi.org/10.1016/j.tig.2017.08.003 112. Andrews SJ, Rothnagel JA (2014) Emerging evidence for functional peptides encoded by short open reading frames. Nat Rev Genet 15:193–204. https:// doi.org/10.1038/nrg3520 113. Ingolia NT, Brar GA, Rouskin S, McGeachy AM, Weissman JS (2012) The ribosome profiling strategy for monitoring translation in vivo by deep sequencing of ribosome-protected mRNA fragments. Nat

94 Protoc 7:1534–1550. https://doi.org/10.1038/ nprot.2012.086 114. Ingolia NT, Ghaemmaghami S, Newman JR, Weissman JS (2009) Genome-wide analysis in vivo of translation with nucleotide resolution using ribosome profiling. Science 324:218–223. https://doi. org/10.1126/science.1168978 115. Ingolia NT, Lareau LF, Weissman JS (2011) Ribosome profiling of mouse embryonic stem cells reveals the complexity and dynamics of mammalian proteomes. Cell 147:789–802. https://doi. org/10.1016/j.cell.2011.10.002 116. Samandi S et al (2017) Deep transcriptome annotation enables the discovery and functional characterization of cryptic small proteins. eLife 6:e27860. https://doi.org/10.7554/eLife.27860 117. van Heesch S et  al (2019) The translational landscape of the human heart. Cell 178:242–260.e229. https://doi.org/10.1016/j.cell.2019.05.010 118. Brunet MA et al (2019) OpenProt: a more comprehensive guide to explore eukaryotic coding potential and proteomes. Nucleic Acids Res 47:D403–D410. https://doi.org/10.1093/nar/gky936 119. Spencer HL et al (2020) The LINC00961 transcript and its encoded micropeptide SPAAR regulate endothelial cell function. Cardiovasc Res 116(12):1981– 1994. https://doi.org/10.1093/cvr/cvaa008 120. Holdt LM et al (2010) ANRIL expression is associated with atherosclerosis risk at chromosome 9p21. Arterioscler Thromb Vasc Biol 30:620–627. https:// doi.org/10.1161/ATVBAHA.109.196832 121. Holdt LM et al (2013) Alu elements in ANRIL non-­ coding RNA at chromosome 9p21 modulate atherogenic cell functions through trans-regulation of gene networks. PLoS Genet 9:e1003588. https://doi. org/10.1371/journal.pgen.1003588 122. Holdt LM et  al (2016) Circular non-coding RNA ANRIL modulates ribosomal RNA maturation and atherosclerosis in humans. Nat Commun 7:12429. https://doi.org/10.1038/ncomms12429 123. Zhang F, Lupski JR (2015) Non-coding genetic variants in human disease. Hum Mol Genet 24:R102– R110. https://doi.org/10.1093/hmg/ddv259 124. Skuratovskaia D, Vulf M, Komar A, Kirienkova E, Litvinova L (2019) Promising directions in atherosclerosis treatment based on epigenetic regulation using microRNAs and long noncoding RNAs. Biomol Ther 9:226. https://doi.org/10.3390/ biom9060226 125. Jarinova O et  al (2009) Functional analysis of the chromosome 9p21.3 coronary artery disease risk locus. Arterioscler Thromb Vasc Biol 29:1671–1677. https://doi.org/10.1161/ATVBAHA.109.189522 126. Wellcome Trust Case Control Consortium (2007) Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447:661–678. https://doi.org/10.1038/ nature05911 127. Abdullah KG et  al (2008) Four SNPS on chromosome 9p21 confer risk to premature, familial CAD

E. J. Hennessy and MI in an American Caucasian population (GeneQuest). Ann Hum Genet 72:654–657. https:// doi.org/10.1111/j.1469-­1809.2008.00454.x 128. Motterle A et al (2012) Functional analyses of coronary artery disease associated variation on chromosome 9p21  in vascular smooth muscle cells. Hum Mol Genet 21:4021–4029. https://doi.org/10.1093/ hmg/dds224 129. Cheng J et  al (2017) Variants in ANRIL gene correlated with its expression contribute to myocardial infarction risk. Oncotarget 8:12607–12619. https:// doi.org/10.18632/oncotarget.14721 130. Burd CE et al (2010) Expression of linear and novel circular forms of an INK4/ARF-associated non-­ coding RNA correlates with atherosclerosis risk. PLoS Genet 6:e1001233. https://doi.org/10.1371/ journal.pgen.1001233 131. Folkersen L et al (2009) Relationship between CAD risk genotype in the chromosome 9p21 locus and gene expression. Identification of eight new ANRIL splice variants. PLoS One 4:e7677. https://doi. org/10.1371/journal.pone.0007677 132. Gil J, Peters G (2006) Regulation of the INK4b-­ ARF-­INK4a tumour suppressor locus: all for one or one for all. Nat Rev Mol Cell Biol 7:667–677. https://doi.org/10.1038/nrm1987 133. Douvris A et  al (2014) Functional analysis of the TRIB1 associated locus linked to plasma triglycerides and coronary artery disease. J Am Heart Assoc 3:e000884. https://doi.org/10.1161/ JAHA.114.000884 134. The CARDIoGRAMplusC4D Consortium et  al (2013) Large-scale association analysis identifies new risk loci for coronary artery disease. Nat Genet 45:25–33. https://doi.org/10.1038/ng.2480 135. Bauer RC, Yenilmez BO, Rader DJ (2015) Tribbles-1: a novel regulator of hepatic lipid metabolism in humans. Biochem Soc Trans 43:1079–1084. https://doi.org/10.1042/BST20150101 136. Johnston JM et al (2019) Myeloid Tribbles 1 induces early atherosclerosis via enhanced foam cell expansion. Sci Adv 5:eaax9183. https://doi.org/10.1126/ sciadv.aax9183 137. Zhang QH, Yin RX, Chen WX, Cao XL, Wu JZ (2019) TRIB1 and TRPS1 variants, G x G and G x E interactions on serum lipid levels, the risk of coronary heart disease and ischemic stroke. Sci Rep 9:2376. https://doi.org/10.1038/s41598-­019-­38765-­7 138. Ohnishi Y et  al (2000) Identification of 187 single nucleotide polymorphisms (SNPs) among 41 candidate genes for ischemic heart disease in the Japanese population. Hum Genet 106:288–292 139. Ishii N et  al (2006) Identification of a novel non-­ coding RNA, MIAT, that confers risk of myocardial infarction. J Hum Genet 51:1087–1099. https://doi. org/10.1007/s10038-­006-­0070-­9 140. Eicher JD et al (2016) Characterization of the platelet transcriptome by RNA sequencing in patients with acute myocardial infarction. Platelets 27:230–

5  LncRNAs and Cardiovascular Disease 239. https://doi.org/10.3109/09537104.2015.10835 43 141. Mitchel K et  al (2016) RP1-13D10.2 is a novel modulator of statin-induced changes in cholesterol. Circ Cardiovasc Genet 9:223–230. https://doi. org/10.1161/CIRCGENETICS.115.001274 142. Chasman DI et  al (2012) Genetic determinants of statin-induced low-density lipoprotein cholesterol reduction: the Justification for the Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER) trial. Circ Cardiovasc Genet 5:257–264. https://doi.org/10.1161/ CIRCGENETICS.111.961144 143. Li C et  al (2019) Regulation of cholesterol homeostasis by a novel long non-coding RNA LASER.  Sci Rep 9:7693. https://doi.org/10.1038/ s41598-­019-­44195-­2 144. Martignano F et  al (2017) Urinary RNA-based biomarkers for prostate cancer detection. Clin Chim Acta 473:96–105. https://doi.org/10.1016/j. cca.2017.08.009 145. Li Q et al (2015) Plasma long noncoding RNA protected by exosomes as a potential stable biomarker for gastric cancer. Tumour Biol 36:2007–2012. https://doi.org/10.1007/s13277-­014-­2807-­y 146. Lin Y, Leng Q, Zhan M, Jiang F (2018) A plasma long noncoding RNA signature for early detection of lung cancer. Transl Oncol 11:1225–1231. https://doi. org/10.1016/j.tranon.2018.07.016 147. Sun Z et  al (2018) Emerging role of exosome-­ derived long non-coding RNAs in tumor microenvironment. Mol Cancer 17:82. https://doi.org/10.1186/ s12943-­018-­0831-­z 148. Aghabozorgi AS et al (2019) Circulating exosomal miRNAs in cardiovascular disease pathogenesis: new emerging hopes. J Cell Physiol 234:21796– 21809. https://doi.org/10.1002/jcp.28942 149. Kumarswamy R et al (2014) Circulating long noncoding RNA, LIPCAR, predicts survival in patients with heart failure. Circ Res 114:1569–1575. https:// doi.org/10.1161/CIRCRESAHA.114.303915 150. Zhang Z et  al (2017) Increased plasma levels of lncRNA H19 and LIPCAR are associated with increased risk of coronary artery disease in a Chinese population. Sci Rep 7:7491. https://doi.org/10.1038/ s41598-­017-­07611-­z 151. Omura J et  al (2020) Identification of the long non-coding RNA H19 as a new biomarker and therapeutic target in right ventricular failure in pulmonary arterial hypertension. Circulation 142(15):1464–1484. https://doi.org/10.1161/ CIRCULATIONAHA.120.047626 152. Li L et  al (2018) Characterization of lncRNA expression profile and identification of novel lncRNA biomarkers to diagnose coronary artery

95 disease. Atherosclerosis 275:359–367. https://doi. org/10.1016/j.atherosclerosis.2018.06.866 153. Song N et  al (2020) Construction and analysis for differentially expressed long non-­ coding RNAs and mRNAs in acute myocardial infarction. Sci Rep 10:6989. https://doi.org/10.1038/ s41598-­020-­63840-­9 154. Yang Y et al (2015) Plasma long non-coding RNA, CoroMarker, a novel biomarker for diagnosis of coronary artery disease. Clin Sci 129:675–685. https:// doi.org/10.1042/CS20150121 155. Chen L et al (2016) Global transcriptomic study of atherosclerosis development in rats. Gene 592:43– 48. https://doi.org/10.1016/j.gene.2016.07.023 156. Chen L et  al (2017) Exosomal lncRNA GAS5 regulates the apoptosis of macrophages and vascular endothelial cells in atherosclerosis. PLoS One 12:e0185406. https://doi.org/10.1371/journal. pone.0185406 157. Wu J et  al (2019) Aberrant expression of serum circANRIL and hsa_circ_0123996 in children with Kawasaki disease. J Clin Lab Anal 33:e22874. https://doi.org/10.1002/jcla.22874 158. Vickers KC, Palmisano BT, Shoucri BM, Shamburek RD, Remaley AT (2011) MicroRNAs are transported in plasma and delivered to recipient cells by high-­ density lipoproteins. Nat Cell Biol 13:423–433. https://doi.org/10.1038/ncb2210 159. Tabet F et  al (2016) High-density lipoprotein-­ associated miR-223 is altered after diet-induced weight loss in overweight and obese males. PLoS One 11:e0151061. https://doi.org/10.1371/journal. pone.0151061 160. Sedgeman LR et  al (2019) Beta cell secretion of miR-375 to HDL is inversely associated with insulin secretion. Sci Rep 9:3803. https://doi.org/10.1038/ s41598-­019-­40338-­7 161. Hamer M, O’Donovan G, Stamatakis E (2018) High-density lipoprotein cholesterol and mortality: too much of a good thing? Arterioscler Thromb Vasc Biol 38:669–672. https://doi.org/10.1161/ ATVBAHA.117.310587 162. Soria-Florido MT et al (2020) Dysfunctional HDLs are associated with a greater incidence of acute coronary syndrome in a population at high cardiovascular risk: a nested-case control study. Circulation 141(6):444–453. https://doi.org/10.1161/ CIRCULATIONAHA.119.041658 163. Kirk JM et  al (2018) Functional classification of long non-coding RNAs by k-mer content. Nat Genet 50:1474–1482. https://doi.org/10.1038/ s41588-­018-­0207-­8 164. Gilbert LA et  al (2014) Genome-scale CRISPR-­ mediated control of gene repression and activation. Cell 159:647–661. https://doi.org/10.1016/j. cell.2014.09.029

6

A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes Thomas C. Brodnicki

Abstract

Type 1 diabetes (T1D) is an autoimmune disease in which immune cells mediate the specific destruction of the insulin-producing β cells in the pancreatic islets. Genetic and transcriptome studies for T1D indicate that a relatively large number of long noncoding RNAs (lncRNAs), detected in both immune cells and β cells, contribute to the underlying inflammation and autoimmune pathology. Although lncRNAs do not encode proteins, their biochemical versatility as RNA molecules enables them to interact with proteins, DNA or RNA to exert regulatory effects on various cellular processes. Recent studies have begun to determine these effects for a small number of lncRNAs in modulating specific immune cell and β-cell responses to elevated glucose levels and pro-inflammatory cytokines that are present within the islets during T1D pathogenesis. These findings are reviewed here and highlight the potential for different lncRNAs to act in concert to inhibit or exacerbate inflammatory and autoimmune responses. T. C. Brodnicki (*) St Vincent’s Institute, Fitzroy, VIC, Australia University of Melbourne, Department of Microbiology & Immunology, Department of Medicine, St Vincent’s Hospital, Fitzroy, VIC, Australia e-mail: [email protected]

Despite this progress to date, additional investigations are required for a more in-depth understanding of their individual functional roles in this interplay, as well as identifying which lncRNAs are likely diagnostic biomarkers or therapeutic targets for autoimmune diseases such as T1D. Keywords

Type 1 diabetes · Pancreatic islets · β cells · Noncoding RNA · Genetic · Pathogenesis

6.1

Introduction

The islets of Langerhans, located within the pancreas, contain a collection of highly differentiated endocrine cells that are responsible for regulating the body’s metabolism by responding to changes in the levels of nutrients, hormones and neurotransmitters. In particular, β cells make up the majority of the endocrine cell population in the pancreatic islets and are the only cells in the body that synthesize and release insulin, a hormone that is critical for regulating glucose homeostasis and biochemical energy storage [1, 2]. Chronic exposure to increased levels of glucose and fatty acids, or the development of islet inflammation, can alter β-cell gene expression and lead to dysfunctional insulin production and β-cell death [1–4]. Once there is a significant loss

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_6

97

T. C. Brodnicki

98

of functional β-cell mass, blood glucose is no longer properly utilized and remains elevated above a healthy concentration. This results in the development of diabetes mellitus and the associated acute and chronic metabolic complications that arise in different tissues and organs of affected individuals, even with exogenous insulin treatment [5–7]. Transcriptome analysis of pancreatic islets, and β cells specifically, have identified differential expression of long noncoding RNAs (lncRNAs) resulting from physiological and pathological conditions, such as fluctuating glucose concentrations or islet-infiltrating immune cells secreting pro-inflammatory cytokines [8– 11]. As initially discovered and extensively reviewed by others (e.g. [12–15], Chap. 2), thousands of lncRNAs can be transcribed from the mammalian genome. Expression analyses readily detect these RNA transcripts that are broadly defined as >200 nucleotides in length and, similar to mRNA, are often spliced and polyadenlyated. In contrast to mRNA, they lack apparent protein-­ coding potential and are typically expressed at lower levels [15–17]. Most lncRNAs, however, have not been functionally characterized leaving their general classification based on proximity to protein-coding genes: intergenic (i.e. lincRNA), intronic, anti-sense, bi-directional or overlapping [18, 19]. When possible, further classification is based on their known mechanistic function via interactions with DNA, proteins or other RNA molecules. These interactions can take place in the nucleus or cytoplasm where lncRNAs typically serve as guides, scaffolds, activators or decoys that regulate various cellular processes, including gene expression, translation, protein transport and signaling pathways (reviewed in [20–23] & Chap. 3). It is also well noted that lncRNAs often have a high degree of cell-type specificity; certain lncRNAs appear to be critical, if not limited, to the development and function of highly differentiated cell types [14, 17, 24]. These combined features of lncRNAs suggest that they have evolved to exquisitely modulate specific cellular responses to unique types of stimuli, for example responses to increased levels

of glucose concentrations in the case of β cells or immune cells. Hundreds of lncRNAs have been identified in pancreatic islets with many of these being specifically expressed in β cells or, in the case of diabetes, potentially within immune cells that have infiltrated the islets [9, 10, 25–28]. A number of these lncRNAs are now known to be important for β-cell or immune cell differentiation, development or function, and these are reviewed in excellent detail elsewhere (e.g. [29–32]). The focus here is on the growing evidence implicating certain lncRNAs in immune-mediated destruction of β cells (Table  6.1). Despite the identification and functional characterization of numerous lncRNAs to date, a major question is: How do lncRNAs contribute to the initiation and progression of the inflammatory and autoimmune disease pathology that underlies the onset of type 1 diabetes (T1D)?

6.2

 ype 1 Diabetes: T Inflammation and Autoimmune-Mediated Destruction of Pancreatic β Cells

Type 1 diabetes (T1D), also known as juvenile diabetes or insulin-dependent diabetes mellitus (IDDM), is an autoimmune disease in which various immune cells, in particular autoreactive T cells, mediate specific destruction of the insulin-­ producing β cells [33, 34]. It is one of the most common chronic diseases in children, but onset also occurs in adults. Globally, it is estimated that over 20 million people suffer from T1D, with the incidence of this disease increasing or stabilizing in different regions of the world [35]. Due to the loss of insulin-producing β cells, individuals with T1D suffer from insulin deficiency and disruption of normal blood glucose homeostasis. Clinical symptoms include polyphagia, polydipsia, polyuria, weight loss, glycosuria and hyperglycemia [6, 36]. Treatment consists of regular monitoring of blood glucose levels and daily injections of exogenous insulin to maintain normal glucose homeostasis. If left untreated, poten-

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes

99

Table 6.1  lncRNAs with genetics, expression or function implicated in type 1 diabetes lncRNA AC008079.1

AK005651

Genetics/Expression rs4759229 located in T1D locus that overlaps an enhancer region predicted to interact with gene encoding AC008079.1 Idd11 haplotypes encompassing AK005651 associated with T1D; expression decreased in NOD mice

ENSG00000099869.7 (IGF2-AS)

Increased expression in macrophages from mice with induced T1D; increased expression in macrophages cultured with high glucose concentrations; overexpression in RAW264.7 cells upregulates expression of inflammatory genes; RNAi-mediated knockdown reduces expression of inflammatory genes Colocalizes with an overlapping eQTL and T1D locus

ENSG00000260302.1 (LINC01882) ENSG00000263179.1 (HNRNPCP4)

Colocalizes with an overlapping eQTL and T1D locus Colocalizes with an overlapping eQTL and T1D locus

ENSG00000274038.1 (AC007014.2)

Colocalizes with an overlapping eQTL and T1D locus

Lethe

Decreased expression in RAW264.7 cells cultured with high glucose concentrations

lnc13

rs917997 located within lnc13 associated with T1D; expression upregulated in islets cells or β-cell line when cultured with virus or double-­ stranded RNA Increased expression in MIN6 cells and mouse islets cultured with IFNγ, IL-1β, TNF-α; increased expression correlates with progression of insulits in NOD mice Increased expression in MIN6 cells and mouse islets cultured with IFNγ, IL-1β, TNF-α; increased expression correlates with progression of insulits in NOD mice Increased expression in MIN6 cells and mouse islets cultured with IFNγ, IL-1β, TNF-α; increased expression correlates with progression of insulits in NOD mice

Dnm3os

lncRNA-1 (gm5970)

lncRNA-2 (AI451557)

lncRNA-3 (BC002288)

Function/Mechanism Anti-sense lncRNA to TUBA8; function/mechanism related to T1D not determined

References [28]

lincRNA; AK005651-deficient immune cells have enhanced TLR-stimulated cytokine production; human ortholog not identified Anti-sense lncRNA to Dnm3 and encodes microRNAs; interacts with nuclear proteins; interaction with nucleolin prevents ability to alter histone modifications in promoters of inflammatory genes

[107]

Anti-sense lncRNA to IGF2; function/mechanism related to T1D not determined lincRNA; function/mechanism related to T1D not determined Processed pseudogene lncRNA; function/mechanism related to T1D not determined Intronic lncRNA in CLEC16A; function/mechanism related to T1D not determined Rps15a-ps4 pseudogene lncRNA; overexpression in macrophages reduces nuclear translocation of NK-kB and decreases NOX2 expression and ROS production lncRNA overlapping IL18RAP; interacts with PCBP2, which binds to the 3’UTR of the STAT1 transcript and leads to production of pro-inflammatory cytokines Processed pseudogene lncRNA overlapping Gm4951; overexpression induces apoptosis in MIN6 cells in response to low doses of IL-1β or TNF-α Anti-sense lncRNA to Nlrc5; overexpression induces apoptosis in MIN6 cells in response to low doses of IL-1β or TNF-α Intronic lncRNA in B2m; overexpression induces apoptosis in MIN6 cells without cytokines

[142]

[114]

[114] [114]

[114]

[147]

[105, 176]

[9]

[9]

[9]

(continued)

T. C. Brodnicki

100 Table 6.1 (continued) lncRNA lncRNA-4 (gm16675)

Malat1

Genetics/Expression Increased expression in MIN6 cells and mouse islets cultured with IFNγ, IL-1β, TNF-α Increased expression in MIN6 cells cultured with IL-1β; increased expression in islets correlates with progression of insulits in NOD mice

MEG3

SNPs associated with paternally inherited risk of T1D; decreased expression in islets correlates with progression of insulits in NOD mice; decreased expression in MIN6 β-cell line cultured with TNF-α

NONHSAG044354 (lnc-MAP377–3)

rs3757247 alters the predicted secondary structure of NONHSAG044354 and contributes to a cis-eQTL for the T1D candidate gene BACH2 Increased expression in MIN6 cells cultured with IFNγ, IL-1β, TNF-α

NONMMUT034373

PVT1

rs13447075 associated with increased risk for end-stage renal disease in patients with T1D

uc.48+

Increased expression in RAW264.7 cells cultured with high glucose and high free fatty acid concentrations

tially fatal metabolic disorders such as ketoacidosis can occur [37]. Even with regular insulin injections, individuals with T1D can still develop serious long-term complications including cardiovascular disease, diabetic retinopathy and nephropathy [7, 38–41]. Currently there is no cure for T1D or therapy for preventing this disease in at-risk individuals. The current model for the natural history of T1D proposes that an individual is born with some degree of genetic risk, and exposure to sub-

Function/Mechanism Bi-directional lncRNA to Irf2; overexpression induces apoptosis in MIN6 cells without cytokines lincRNA; overexpression reduces H3 histone acetylation of Pdx-1 promoter and decreases its expression; knockdown increases Pdx-1 expression and reverses decreased insulin secretion caused by exposure of β cells to IL-1β; functional roles in various immune cells are defined, but not determined in T1D lincRNA; rs34552516 affects binding to nuclear protein extracts of the Jurkat T-cell line; MEG3 interacts with p53 and EZH2 or acts as a ceRNA; knockdown decreases expression of MafA leading to decreased insulin production in β cells. lncRNA overlapping BACH2; function/mechanism related to T1D not determined

lncRNA overlapping Pdl1; function/ mechanism related to T1D not determined lincRNA; depletion in T cells decreases c-Myc protein, decreases cell proliferation and inhibits T-cell differentiation into Th1 effector cells; function/mechanism related to T1D not determined lincRNA; knockdown decreases expression of P2X7 receptor and production of pro-inflammatory cytokines; function/mechanism related to T1D not determined

References [9]

[178, 200–202]

[106, 128, 130]

[103]

[160]

[135, 136]

[149]

sequent environmental factors influences the onset of β-cell destruction [6, 42, 43]. The development of the initial inflammatory and progressive autoimmune responses within the pancreatic islets, along with the physiological turnover of β cells, contributes to this process [44–46]. In the asymptomatic phase, it is the initial immune dysregulation, facilitated by genetic predisposition, that results in early serological evidence of β-cell destruction and includes detection of serum autoantibodies specific for β-cell proteins such as pro-

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 101

insulin, glutamic acid decarboxylase 65 (GAD65), insulinoma-associated autoantigen 2 (IA2), islet-specific glucose-6-phosphatase catalytic subunit related protein (IGRP) and zinc transporter 8 (ZnT8) [47–51]. Although not readily detected as early as serum autoantibodies, T cells isolated from T1D patients also recognize autoantigens derived from β-cell proteins, including epitopes derived from insulin [52–57]. In most T1D patients, a decrease in insulin secretion and disruption of blood glucose regulation may take months to years after the initial detection of multiple T1D-associated autoantibodies [58, 59]. For reasons that are still unknown, some patients do not progress to overt diabetes despite the detection of multiple islet autoantibodies [58, 59]. For those patients that do progress to overt diabetes, various metabolic changes occur, including decreased serum C-peptide (an insulin-derived biomarker for measuring insulin production), increased plasma glucose fluctuations and an overall rise of serum and urinary glucose concentrations [60, 61]. Once a critical functional mass of β cells has been destroyed, exogenous insulin treatment is required [6, 36]. Studies in humans and the nonobese diabetic (NOD) mouse model have dissected different stages of the development of T1D [6, 62, 63]. The initial stage is asymptomatic and is initiated by stochastic yet unknown event(s) that trigger the invasion of the pancreatic islets by a mixed population of leukocytes. This initial inflammatory process is termed insulitis and in the first stage consists mainly of macrophages, neutrophils and dendritic cells, which can be detected in the pancreatic islets of NOD mice as early as 3–6 weeks of age. These innate immune cells are followed by the accumulation of autoreactive lymphocytes (T cells and B cells), which can be detected in the islets by 10 weeks of age [64–69]. Each of these immune cell types contribute to the inflammation found within the pancreatic islets of NOD mice, but it is the CD8+ T cells that are primarily responsible for the autoimmune destruction of the insulin-producing β cells, while other islet cells are for the most part spared. Leukocyte infiltrates are also observed in the pancreatic islets of human individuals recently diagnosed with T1D

[70–72]. The cellular composition of this islet infiltrate includes macrophages, dendritic cells, B cells and T cells [34, 70, 71, 73–76]. However, it is the presence of β-cell specific T cells that provides the primary evidence, along with HLA genetic association, for the autoimmune basis of T1D [45].

6.3

Genetic and Expression Studies Point to lncRNAs in Type 1 Diabetes

An individual’s susceptibility to developing T1D results from a complex interaction between genetic and relatively undefined environmental factors, with genetic predisposition strongly implicated in the overall risk for developing disease [42, 43, 77]. This is evident in first-degree relatives of T1D patients, who have at least a 10 times higher risk of developing the disease than the general population [78–81]. Genetic association studies have also identified numerous T1D susceptibility loci in the human genome with allelic variation for genes encoding human leukocyte antigen (HLA) class II providing the greatest genetic risk [82, 83]. Allelic variation associated with T1D was subsequently identified for the insulin gene (INS) [84, 85], as well as for genes implicated in immune responses, such as CTLA4 [86], PTPN22 [87] and IL2RA [88]. More recent genome-wide association studies (GWAS) and meta-analyses of different populations have detected additional genomic intervals that harbor T1D-associated alleles; at least 80 loci at present (e.g. [77, 89–99],). Apart from the HLA locus, these loci have small individual effects upon disease risk, but together they account for a large proportion of predicted T1D genetic heterogeneity in human populations [77, 100]. However, identification of the actual causative sequence within these T1D loci can be challenging. Associated single nucleotide polymorphisms (SNPs) localize to genomic intervals for which linkage disequilibrium can stretch to hundreds of kilobases (kb), which often contain many genes and regulatory regions, making it difficult to pinpoint the causative sequence variants [77, 100,

102

101]. Therefore, additional analyses of these intervals are still required to further dissect their genetic and functional contribution to T1D pathogenesis. As noted in Chap. 8, a majority of the disease-associated SNPs, including those associated with T1D, map to noncoding intervals that often include lncRNAs. Analysis of human GWAS for various diseases, including inflammatory and autoimmune disease, indicate that up to 90% of the associated SNPs are located in noncoding regions [102]. One bioinformatic analysis set out to identify the lncRNAs associated with T1D in two ways. In the first instance, they identified those lncRNAs that are located in and around 5  kb up/down-­ stream flanking regions of candidate protein-­ coding genes harboring T1D-associated SNPs [103]. A total of 816 lncRNA genes were identified with 381 classified as sense exonic, 53 as sense non-exonic, 188 as antisense and 225 as intergenic lncRNAs (note: some lncRNAs fulfill the criteria of more than one classification). This result suggested that lncRNAs overlapping or in close proximity to candidate protein-coding genes may also be affected by the SNPs within these T1D susceptibility loci. In the second instance, specific characterization of nominally associated T1D SNPs found that 1045 of these T1D-associated SNPs were directly located within 247 of these lncRNAs genes. Furthermore, 178 of these SNPs were predicted to significantly disrupt the putative secondary structure of these lncRNAs [103]. Another bioinformatic study evaluated lncRNA expression in peripheral blood mononuclear cells from 12 T1D and 10 control subjects [104]. 184 differentially expressed lncRNAs and two putative lncRNA-microRNA-­ mRNA regulatory axes were found to be associated with T1D; although no comparison was made to identify which of these lncRNAs harbor T1D-associated SNPs. While additional scrutiny and functional studies are required of both sets of T1D-associated lncRNAs, these initial studies have begun to provide compelling evidence that genetic variation and expression for lncRNAs is likely to affect T1D risk. Other independent genetic studies have focused on two human and one mouse lncRNA

T. C. Brodnicki

(Table  6.1). Lnc13 in humans harbors the SNP rs917997 for which the rs917997*C allele increases risk for T1D [105], and SNPs within the genomic interval encompassing the human lncRNA MEG3 have been associated with paternally inherited risk of T1D [106]. Similar to humans, genetic studies of the NOD mouse have also identified at least 40 T1D susceptibility loci, termed Idd loci [62, 63]. While these mouse genetic studies did not focus specifically on lncRNAs, at least one of these T1D susceptibility loci has been attributed to a lncRNA. Sequence analysis of a series of congenic NOD mouse strains and 25 other inbred mouse strains identified several haplotypes within the Idd11 locus and associated with varying levels of T1D susceptibility. The haplotype diversity underlying this T1D susceptibility locus was due to a recombination hotspot located within the putative lncRNA AK005651 [107]. Thus, this study was the first to show that changing the sequence variation within a lncRNA could alter T1D susceptibility in vivo; although at present it is not clear if there is human equivalent for this murine T1D-­ associated lncRNA. In all three cases, functional studies, described below, have begun to identify cellular and molecular mechanisms for these three lncRNAs and point to their role in T1D pathogenesis. T1D-associated SNPs also have the potential to affect lncRNA expression and tissue specificity due to their location within regulatory elements [98, 108]. These regulatory elements, such as transcription factor binding sites, histone modification marks and enhancers, are often detected as expression quantitative trait loci (eQTLs), in which genetic variation within these regulatory elements affects expression of flanking genes (i.e. cis-eQTLs) or expression of genes located elsewhere in the genome (i.e. trans-eQTLs) [108–110]. Previous studies using eQTL data along with chromosome conformational capture datasets have identified more than 200 protein-­ coding genes for which expression was associated with T1D-associated SNPs [98, 111–113]. Although less studied, similar analyses have begun to identify tissue-specific eQTLs for which a SNP is associated with T1D and located within

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 103

or nearby a lncRNA (Table  6.1). For example, SNP rs3757247 is located within lncRNA NONHSAG044354 and contributes to a cis-eQTL for the T1D candidate gene BACH2 [103]. A more recent study also analyzed colocalization of eQTLs associated with both lncRNA expression and GWAS-defined regions for various diseases including T1D using data from the Genotype Tissue Expression Project [114]. While T1D was not a primary focus of this study, four lncRNAs (ENSG00000099869.7, ENSG00000260302.1, ENSG00000263179.1, ENSG00000274038.1) were noted as colocalizing with overlapping eQTLs and T1D GWAS loci. Intriguingly, ENSG00000099869.7 is an antisense lncRNA to the gene IGF2 that encodes insulin-like growth factor 2 and is upregulated in β cells stimulated with high concentrations of glucose [115]. Moreover, ENSG00000274038.1 is an intronic lncRNA in CLEC16A, which is a known T1D candidate gene [116, 117]. However, neither of these four lncRNAs were identified in the previous study [103], suggesting combinations of datasets and/or complementary analysis strategies may be needed to pinpoint critical T1D-­ related lncRNAs. Chromosome conformational capture datasets have also been used to identify potential physical interactions between T1D-­ associated SNPs, regulatory elements and lncRNA genes. For example, SNP rs4759229 falls within a T1D susceptibility locus and overlaps a known enhancer region that is predicted to have a long-range interaction with the antisense lncRNA AC008079.1 [28]. How these various genetic variants and interactions impact upon the molecular and cellular effects of these lncRNAs, let alone how these affect T1D risk and pathogenesis, largely remain to be investigated.

6.4

l ncRNAs Regulate Immune Cell Responses Implicated in Type 1 Diabetes Pathogenesis

Although challenging to investigate in human T1D patients, studies in NOD mice indicate that infection and physiological β-cell death are the

two main types of events that are likely to attract macrophages and dendritic cells to the pancreatic islets [118–120]. These innate immune cells not only promote β-cell death by releasing pro-­ inflammatory cytokines, but also serve as antigen-­ presenting cells to capture and present β-cell antigens to T cells [121, 122]. There are a growing number of studies that have identified lncRNAs implicated in these immune cell functions, in particular inflammatory responses [30, 31]. These lncRNAs predominantly regulate transcription of inflammatory genes. For example, lincRNA-Cox2 interacts with hnRNP-A/B and A2/B1 to regulate the expression of pro-­ inflammatory cytokines and interferons [123, 124]; THRIL interacts with hnRNP-L to regulate the expression of TNF-α [125]; the natural anti-­ sense transcript anti-IL1-β regulates IL1-β expression by altering the promoter chromatin structure [126]; and AW112010 regulates IL-10 expression via histone demethylation, which promotes differentiation of inflammatory T cells [127]. Nonetheless, these and other lncRNAs, which have been shown to regulate immune cell function, have yet to be functionally tested and reported for their role in insulitis and β-cell destruction. However, a few studies have started to investigate immune cells for other lncRNAs that harbor T1D-associated SNPs or have altered expression under diabetic conditions. At least three lncRNAs have been identified for which genetic variation is associated with T1D and functional studies point to potential affects in immune cells (Table 6.1). One of these is MEG3, which harbors three T1D-associated SNPs (rs941576, rs34552516, rs56994090) and for which protein-binding and enhancer-activity assays were performed [106, 128]. Specifically, a nucleotide probe having the rs34552516*TC allele demonstrated binding to nuclear protein extracts of the Jurkat T-cell line, but not to extracts from the THP-1 monocyte cell line. Moreover, the rs34552516*TC allele led to increased luciferase activity in Jurkat T cells when compared to the rs34552516*T allele [128]. While these results suggest a role for MEG3  in T cells and T1D pathogenesis, it was acknowledged that this was a limited functional

104

analysis and did not investigate MEG3  in primary cells [128]. Analysis of MEG3 is complicated because its expression, which includes as many as ten isoforms, is detected in various cell types, including macrophages and β cells, in addition to T cells [129–131]. Mechanistically, MEG3 transcripts have been found to bind p53 or act as a competing endogenous RNA (ceRNA) that serves as a miRNA decoy [129, 132, 133]. Thus, despite genetic association of MEG3 with T1D, it is not yet clear which molecular mechanisms are affected and how this alters T-cell function that contributes to increased T1D susceptibility. Another lncRNA implicated in T-cell function is PVT1, although the functional and genetic link to T1D is more tenuous than MEG3. PVT1 is a well-known oncogene [134], but also found to be upregulated in activated CD4+ T cells of Sjögren’s syndrome patients compared to healthy donors [135]. It harbors a SNP (rs13447075) associated with increased risk for end-stage renal disease in patients with T1D [136]. PVT1 interacts with and stabilizes c-Myc, leading to activation of downstream pathways that if unchecked promote tumorigenesis [137]. Depletion of PVT1  in T cells (mouse CD4+ T cells or Jurkat T cells) results in decreased c-Myc protein leading to decreased glycolysis and cell proliferation, as well as inhibiting differentiation of naïve T cells into Th1 effector cells [135]. Whether PVT1 expression is altered in T cells of T1D patients and contributes to T1D pathogenesis remains to be investigated. In contrast to MEG3 and PVT1, which were linked to T1D based on human genetic analyses, lncRNA AK005651 was identified by positionally cloning a T1D susceptibility locus in the NOD mouse strain [107], as noted above. Notably, congenic NOD mouse strains demonstrated that sequence variation within AK005651 could alter its expression in various tissues and affect T1D susceptibility [107]. More recently, genetic disruption of AK005651 in C57BL/6 (B6)  mice, which are resistant to spontaneous diabetes, results in increased susceptibility to autoimmune diabetes induced by multiple low-­ dose streptozotocin. Subsequent analysis has

T. C. Brodnicki

determined that expression of this lncRNA is induced by toll-like receptor (TLR) activation and AK005651-deficient immune cells have enhanced cytokine production when stimulated by TLR ligands in vitro (unpublished, TC Brodnicki). This suggests that AK005651 is part of a negative feedback mechanism for attenuating TLR-mediated cytokine production, which is known to contribute to the induction and modulation of T1D pathogenesis in mice [138]. Although the human equivalent for AK005651 has yet to be defined, these congenic and knockout mouse results are the first to demonstrate in vivo that the development of T1D can be directly affected by changing the sequence and expression of a specific lncRNA. In particular, these studies suggest that AK005651 is important for regulating specific immune cell responses that are contributing to the initial stages of islet inflammation. As the functional β cell mass and insulin levels progressively decrease during T1D pathogenesis, blood glucose concentrations begin to fluctuate above the normal range prior to the onset of clinical symptoms [6, 36, 81]. Although not definitively shown in vivo, fluctuation of blood glucose concentrations above the normal range during the asymptomatic stage of T1D may further promote production of inflammatory responses by macrophages that reinforce and escalate the ongoing T-cell mediated destruction of the β cells during the later stages of insulitis [139–141]. Intriguingly, at least three studies have shown that increased glucose concentration can alter lncRNA expression associated with enhanced inflammatory responses in macrophages (Table 6.1). Peritoneal macrophages isolated from B6 mice with autoimmune diabetes, induced by multiple low-dose streptozotocin, have increased expression of Dnm3os, as do macrophages derived from the bone-marrow of healthy  B6 mice and  cultured in high glucose-containing media (Table 6.1) [142]. As an antisense lncRNA, Dnm3os overlaps Dnm3, but also serves as a host gene encoding microRNAs that have been implicated in cancer, cardiac and kidney disease [143– 146]. Overexpression of Dnm3os, but not its encoded microRNAs, in the mouse RAW264.7

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 105

macrophage cell line upregulated expression of a collection of inflammatory genes, and RNAi-­ mediated knockdown of Dnm3os attenuated this response [142]. Dnm3os interaction with a number of nuclear proteins was subsequently detected in macrophages. Further investigation demonstrated that interaction with nucleolin prevents the functional ability of Dnm3os to affect histone modifications in promoters of certain inflammatory genes. However, reduction of nucleolin and increased Dnm3os expression, under diabetic conditions, leads to enhanced macrophage inflammatory responses [142], suggesting a role for Dnm3os during early stages of T1D pathogenesis. In contrast to Dnm3os, the lncRNA Lethe exhibits decreased expression in the mouse RAW264.7 macrophage cell line when cultured in high glucose concentrations (Table 6.1) [147]. This was associated with increased expression of NOX2, which is activated by NF-κB, and resulted in increased production of reactive oxygen species (ROS). Notably, Lethe is induced by pro-­ inflammatory cytokines via NF-κB signaling and acts in a negative feedback loop by interacting with NF-κB to inhibit RelA DNA binding and prevent target gene activation [148]. As predicted, overexpression of Lethe in macrophages reduced nuclear translocation of NF-κB, which resulted in decreased NOX2 expression and ROS production [147]. Although this study was focused on diabetic wound healing [147], previous studies in the NOD mouse have demonstrated that macrophage production of ROS via NOX2 is critical for initiation of T-cell mediated destruction of β cells [140]. Taken together, this points to Lethe as a potential regulator of inflammation in the pancreatic islets during the initial stages of T1D pathogenesis. In yet another study of the mouse RAW264.7 macrophage cell line, a combination of high glucose and high free fatty acid concentrations upregulated expression of the lncRNA uc.48+ (Table  6.1). This was associated with increased expression of the P2X7 receptor along with increased production of IL-1β and TNF-α [149]. RNAi-mediated knockdown of uc.48+ decreased expression of P2X7 receptor and production of

these pro-inflammatory cytokines. While these findings suggest a role for uc.48+ in promoting macrophage-mediated inflammation under diabetic conditions, the underlying molecular mechanism remains to be elucidated [149]. Similarly, the studies for the other two lncRNAs, Dnm3os and Lethe, characterized their inflammatory effects in RAW264.7 cells, which are not necessarily representative of the primary macrophages contributing to insulitis and β-cell destruction [150]. Further investigation of these three lncRNAs, ideally using in vivo T1D models or human cadaveric pancreatic samples, are required to validate their functional effects on macrophage-­ mediated inflammation during T1D pathogenesis.

6.5

l ncRNAs Implicated in β-Cell Responses to Inflammation During Type 1 Diabetes Pathogenesis

Expression studies have also implicated intrinsic lncRNA effects in β cells that are responding to the initial stages of islet inflammation. During insulitis, but prior to T-cell mediated destruction, β cells become increasingly exposed to cytokines and pro-apoptotic mediators released by both the early infiltrating immune cells (e.g. macrophages) and by the islet cells themselves. It has been shown in vitro that IFNγ, IL-1β and TNF-α can affect β-cell function, leading to decreased capacity to produce and release insulin in response to glucose [44, 151, 152]. Moreover, prolonged exposure to these pro-inflammatory cytokines can promote apoptosis and initiate β-cell loss [3, 4, 44, 153, 154]. In both cases, pro-­ inflammatory cytokines modulate β-cell expression of various genes, including lncRNAs, that are implicated in key signaling pathways important for insulin secretion or apoptosis [155–158]. Transcriptome expression studies of human pancreatic islets and the mouse MIN6 β-cell line have identified >400 putative lncRNAs that are upregulated or downregulated upon exposure to pro-inflammatory cytokines in vitro [9, 159, 160]. Pancreatic islet cells, β cells and infiltrating

106

T. C. Brodnicki

lymphocytes can also release extracellular vesi- β-cell line when exposed to IFNγ, IL-1β and cles, such as exosomes, which not only contain TNF-α (Table  6.1) [9]. A role for these four β-cell-specific autoantigens, but also noncoding lncRNAs in actual islet inflammation was supRNAs [161–163]. The contents of these exo- ported by their increased expression in mouse somes have been shown to regulate β-cell func- islets cultured with these pro-inflammatory cytotion and death [163]. Intriguingly, a set of kines, but not by fatty acid palmitate or an insulin differentially expressed lncRNAs has been secretagogue. Increased expression of the four detected in exosomes derived from human pan- lncRNAs also correlated with the increase in creatic islets after exposure to pro-inflammatory insulitis in pre-diabetic NOD mice, indicating cytokines in vitro [164]. Taken together, these their potential role during early stages of T1D expression studies of pancreatic islets, β cell lines pathogenesis. Moreover, overexpression of the and exosomes indicate that lncRNAs are likely to four lncRNAs, alone or in combination with have various roles in the different stages of islet cytokines, did not affect insulin production or inflammation and autoimmune destruction of β secretion, but did affect β-cell apoptosis. In parcells initiated by cytokine release during ticular, overexpression of lncRNA-3 or insulitis. lncRNA-4 in MIN6 cells was sufficient without Bioinformatics and functional studies for a cytokines to increase β-cell apoptosis; whereas select number of these inflammation-responsive overexpression of lncRNA-1 or lncRNA-2 could lncRNAs have begun to explore their importance only induce apoptosis in MIN6 cells in combinain β-cell function and immune-mediated destruc- tion with IL-1β or TNF-α [[9]]. Further investigation during T1D pathogenesis. In the first tion found that overexpression in MIN6 cells of instance, bioinformatics analysis has provided lncRNA-1, but not the other three lncRNAs, led various degrees of informed speculation. For to the nuclear translocation of NF-κB, which example, the lncRNA NONMMUT034373 was plays a critical role in cytokine-induced apoptoup-regulated in the MIN6 β-cell line after cyto- sis [169–173]. Although additional mechanistic kine exposure (Table  6.1) [160]. Genomic studies are required, these results suggest that sequence analysis determined it to be a sense modulation of certain sets of lncRNAs may prolncRNA overlapping the gene encoding pro- mote sensitization of β cells to apoptosis, thus grammed death-1 ligand (PD-L1). PD-L1 and its exacerbating their susceptibility to immune-­ receptor PD-1 are important for autoreactive mediated destruction during both early inflamT-cell tolerance and implicated in T1D pathogen- matory and later autoimmune stages of insulitis. esis [165–168]. It was therefore postulated that Pancreatic islets and β cells can also directly NONMMUTO34373 may regulate expression of contribute to the local islet inflammatory milieu PD-L1 and contribute to the later stages of insuli- by releasing chemokines that recruit both innate tis when β-cell-specific T cells join the inflamma- and adaptive immune cells [44, 71, 174, 175]. tory milieu within the islets. Nonetheless, the Lnc13 is one such lncRNA that has recently been expression of this lncRNA, as well as the other implicated in chemokine production by β cells 722 differentially expressed lncRNAs identified (Table 6.1). The sense sequence of Lnc13 overin this initial study, still need to be confirmed and laps the protein-coding gene IL18RAP, which has functionally tested during insulitis [160]. also been proposed as a disease candidate gene. Other studies have gone on to confirm expres- Lnc13 is detected in human β cells, but its expression of specific lncRNAs in isolated pancreatic sion is not affected by pro-inflammatory cytoislet cells, as well as test their functional effects kines. Instead, viral double-stranded RNA or by overexpression in β cell lines. Motterle et al. Coxsackie Virus B5 infection upregulates Lnc13 chose to further characterize four lncRNAs expression in cultured human islets and β cells (termed lncRNA-1/gm5970, lncRNA-­ [176]. Lnc13 subsequently interacts with poly(rC) 2/AI451557, lncRNA-3/BC002288 and lncRNA-­ binding protein 2 (PCBP2), which binds to the 4/gm16675) that were upregulated in the MIN6 3’UTR of the STAT1 transcript. Overexpression

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 107

of Lnc13 in the EndoC-βH1 human β cell line led to increased STAT1 activation and production of pro-inflammatory chemokines. This suggests that Lnc13 plays a regulatory role in STAT1-driven inflammation at the pancreatic β-cell level. Notably, Lnc13 harbors the SNP rs917997 for which the rs917997*C allele increases risk for T1D [105]. This allele promotes a stronger interaction between PCBP2 and the 3’UTR of the STAT1 transcript as compared to the non-risk allele rs917997*T [176]. This enhanced stabilization of STAT1 mRNA is associated with sustained production of chemokines in virally stimulated β cells, which is predicted to promote inflammation in the pancreatic islets leading to β-cell destruction [176]. More broadly, Lnc13 adds to the accumulating evidence for the interaction between environmental factors, such as viral infections, and genetic variation that increases T1D susceptibility. The SNP rs917997 located within Lnc13 is also associated with celiac disease, a chronic immune-mediated disease that involves inflammation of the small intestine [177]. However, the allelic effect is different to that identified for T1D. While the rs917997*C allele increases risk for T1D, the rs917997*T allele increases the risk for celiac disease. In contrast to T1D, the rs917997*T allele decreases Lnc13’s affinity to bind hnRNPD and chromatin. This decreased interaction correlates with increased expression of STAT1, MYD88, IL1RA and TRAF2, along with increased expression of pro-inflammatory genes in macrophages or intestinal biopsies from active celiac disease patients [177]. While both alleles result in a similar phenotype – increased inflammatory responses  – this appears to occur via allelic effects on different molecular mechanisms depending on the cell type (i.e. macrophages, β cells or intestinal cells). Thus Lnc13’s effects are not only cell specific, but also allele specific. This observation provides additional evidence that disease-associated polymorphisms within lncRNAs are likely to alter their secondary structure. These structural alterations affect molecular interactions important for different intracellular pathways that modulate inflammatory responses in different cell types.

Inflammation can also affect insulin production in β cells that have not yet been destroyed [4, 44, 71]. At least two lncRNAs have been described as having abnormal expression in the islets of aging NOD mice as insulitis develops: Malat1 shows increased expression, whereas Meg3 shows decreased expression (Table  6.1) [130, 178]. Separate functional studies of these lncRNAs were subsequently carried out in the mouse MIN6 β-cell line to begin elucidating their roles in modulating β-cell function as a result of inflammation. In the case of Malat1, exposure of MIN6 cells to inflammation in the form of IL-1β led to increased Malat1 expression associated with decreased insulin secretion [178]. Increased expression of Malat1 was shown to reduce H3 histone acetylation of the Pdx-1 promoter leading to decreased expression of Pdx-1, a transcription factor that promotes insulin synthesis and is important for β-cell development [178–180]. In turn, RNAi-mediated knockdown of Malat1 increased Pdx-1 expression and reversed the decreased insulin secretion caused by exposure of β cells to IL-1β [178]. In the case of Meg3, exposure of MIN6 cells to inflammation, this time in the form of TNF-α, led to decreased Meg3 expression, which was also associated with reduced insulin secretion [130]. Decreased expression of Meg3, via RNAi-mediated knockdown, was shown to not only decrease insulin secretion, but also decrease expression of Pdx-1 and MafA, another transcription factor that promotes insulin expression [130, 181]. In particular, Meg3 was shown to interact with the methyltransferase EZH2, which drives trimethylation and binding of H3K27 to the promoters of Rad21, Smc3 and Sin3α. In this way, Meg3 is proposed to negatively regulate the expression of Rad21, Smc3 and Sin3α, which bind to the MafA promoter and inhibit MafA expression in MIN6 cells. Thus decreased Meg3 expression results in impaired insulin synthesis and secretion due to downregulation of MafA [131]. Taken together, these two studies show that inflammation is likely to affect a combination of lncRNAs expressed in β cells (e.g. Malat1 and Meg3) that target different molecular mechanisms regulating the production of insulin. Although further studies are

T. C. Brodnicki

108

needed, such effects would be predicted to further exacerbate the inability of the remaining β cell mass, not yet destroyed by infiltrating immune cells, to maintain glucose homeostasis.

6.6

Overlapping lncRNAs Between Type 1 Diabetes and Type 2 Diabetes

Unlike T1D, which is ultimately the result of autoimmune destruction of β cells, type 2 diabetes (T2D) results from the progressive increase in insulin resistance in muscle, liver and adipose tissue, due in part to chronic inflammation [3, 45]. This eventually leads to impaired insulin secretion by β cells and subsequent dysregulation of normal blood glucose homeostasis [5, 182]. Increased adiposity and subsequent obesity provide the greatest risk factors for developing insulin resistance and T2D, but GWAS have also identified more than 100 polymorphisms associated with T2D [183–187]. Similar to T1D, these genetic studies in combination with transcriptome studies have identified lncRNAs for which polymorphisms or differential gene expression have been associated with T2D [8, 25, 188–190]. Given the contribution of low-grade chronic inflammation to T2D [182], it is perhaps not unexpected that lncRNAs implicated in inflammation during T1D pathogenesis may also play a role in T2D.  At least four of the lncRNAs described above for T1D have been associated with T2D: Dnm3os, uc.48+, MEG3 and Malat1. All four have been found to be differentially expressed in peripheral blood monocytes or serum of T2D patients [142, 149, 188]. Moreover, Dnm3os expression is increased in bone-marrow derived  macrophages derived from db/db mice, an animal model of T2D [142]. The RAW264.7 macrophage cell line and murine bone-marrow derived macrophages also show increased expression of Dnm3os and uc.48+ respectively when cultured with high glucose concentrations [142, 149]. On one hand, these lncRNA-related effects would be predicted to contribute to adipose tissue inflammation and insulin resistance that drives

the development of T2D, rather than the islet inflammation observed for T1D [182, 191]. On the other hand, MEG3 is also found to be differentially expressed in pancreatic islets from T2D patients and db/db mice [130, 192]. Expression of both MEG3 and Malat1 can also be detected in other tissues associated with T2D, including adipose tissue and liver [193–196]. Notably, overexpression of Meg3 in mouse primary hepatocytes promoted insulin resistance in culture [197, 198]; and siRNA depletion of Malat1  in ob/ob mice, another animal model of T2D, resulted in improved insulin sensitivity and glucose tolerance [199]. This representative, albeit small, group of lncRNAs implicated in both T1D and T2D suggests there will be those lncRNAs  that have multifaceted contributions to both types of diabetes pathogenesis, with their investigation in one type of diabetes likely to be informative for the other.

6.7

Concluding Remarks

Genetic and transcriptome studies clearly point to lncRNAs that contribute to complex signaling networks that are important for regulating inflammation-­induced responses in both the islet-­ infiltrating immune cells and  the pancreatic β cells. Intriguingly, certain lncRNAs are implicated in both cell types contributing to diabetes. As noted above, inflammation-induced inhibition of Meg3 expression decreases insulin production in β cells [130, 131], while genetic variation for MEG3 is associated with increased T1D risk and altered intracellular signaling in T cells [128]. Inflammation-induced Malat1 expression also decreases insulin production in β cells with recent studies indicating Malat1’s functional role in T cells, dendritic cells and macrophages, albeit not yet linked to islet inflammation [178, 200–202]. In both cases, different molecular mechanisms are used by a given lncRNA depending on the stimulus and cell type (β cells or immune cells); thus highlighting the role of lncRNAs to act in multiple signaling pathways and often regulating more than one node of the underlying signaling network.

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 109

The increasing number of lncRNAs that have been studied to date and relevant to T1D also highlight how the interplay between genetic variation, environmental factors and the local tissue milieu might trigger and exacerbate the contributions of lncRNAs to disease pathogenesis. T1D-­ associated variants for lncRNAs can alter secondary structure that affect function (e.g. lnc13, MEG3) or regulatory/splicing elements that affect expression levels (e.g. AC008079.1, AK005651), both of which lead to enhanced inflammatory and autoimmune responses [28, 105–107, 128, 176]. From this perspective, such lncRNAs appear to participate in negative or positive feedback loops that can induce or modulate the intensity of immune responses. One facet of immune tolerance is to limit the intensity of an immune response that would otherwise lead to pathological inflammation and autoimmunity [203]. The current set of in vitro studies not only show that genetic variation for a lncRNA can affect β-cell responses to viruses [176], but that increased glucose concentrations, reflecting in vivo β-cell death or tissue insulin resistance, can alter immune cell expression of lncRNAs that are implicated in pro-inflammatory cytokine and ROS production [142, 147, 149]. It is becoming clear, based on these studies, that glucose concentration and inflammation together drive differential expression of  lncRNAs that influence the dialog between β cells and islet-infiltrating immune cells, which can either inhibit or exacerbate immune cell infiltration and subsequent β-cell destruction. Advanced sequencing technologies continue to improve and enable the update of genomic information that has increased our understanding of T1D. While protein-coding genes were more often the primary focus of past studies, new bioinformatics and statistical approaches have enabled retrospective analysis and identification of genetic and expression variation for lncRNAs that are likely to play key roles in the development or prevention of T1D [9, 98, 103, 104, 159, 160]. Nonetheless, definitively showing that a specific lncRNA has a functional effect in promoting or preventing T1D pathogenesis is the next challenging step. Even the studies described

above were only able at best to strongly implicate a given lncRNA in islet inflammation. In the first instance, and certainly not particular to the T1D field, many of the T1D-associated lncRNAs are still poorly annotated with regards to cell specificity, isoforms and regulatory elements, leaving investigators to often prioritize previously studied lncRNAs or those for which functional prediction and/or characterization are more amenable (e.g. lncRNAs that do not overlap protein-­coding genes or do not encode miRNAs or potential small bioactive peptides). Current lncRNA functional studies also predominantly rely on cell lines. Such studies are clearly an efficient and informative starting point especially when employing genetic manipulation and controlled stimulus conditions for assessing expression, binding partners and functional effects of T1D-associated SNPs in immune cell responses (reviewed in [204]). However, the challenge is to identify and confirm a causal mechanistic effect for a given lncRNA during T1D pathogenesis. One strategy unique to T1D research is to use the NOD mouse strain, which develops T1D and enables validation of genetic effects and characterization of islet inflammation at different time points during T1D pathogenesis [6, 62, 63, 107]. A caveat is that genetic alterations for lncRNAs in mice often reveal subtle or no disease phenotypes [204, 205]. In this regard, the NOD mouse provides an advantage because it has a ‘sensitized’ genetic background in which genetic alterations for lncRNAs can be investigated for preventing, delaying or accelerating the onset of T1D.  Analysis of patient peripheral blood can also provide some insights as has already been shown [104], but it is access to the pancreas during different disease stages that is ultimately needed to validate functional effects on immune-­ cell or β-cell mediated inflammation within human islets. While access to pancreatic biopsies is not possible, cadaveric pancreatic tissue and the potential to use pluripotent stem cells with specific genetic variants from T1D patients are available albeit limited options [206, 207]. The current technical challenge is how to use such pancreatic tissue and in vitro cell models for functional studies of lncRNAs. This problem is

T. C. Brodnicki

110

not exclusive to T1D as research groups in this and other disease fields are developing new or improved methods that can be used for investigating lncRNA function in these types of disease tissue studies (reviewed in [208] & Chap. 10). Accumulating genomic and functional findings by numerous groups have undeniably revealed lncRNAs to be a novel class of regulatory molecules with both cell- and stimulus-­ specific expression patterns relevant to inflammatory and autoimmune diseases. The finding that polymorphisms associated with T1D fall within genes encoding lncRNAs has created new research opportunities for better understanding genetic risk and the molecular mechanisms underlying T1D pathogenesis. Although functional studies have so far been limited to relatively few lncRNAs and cell types relevant to T1D, these have indicated that lncRNAs will play complex roles in both immune cell and β-cell responses. Moreover, expression analyses suggest a relatively large number of lncRNAs act in concert to attenuate or exacerbate inflammatory and autoimmune responses during both early and late stages of disease progression. The future challenge will be to dissect out which of these lncRNAs are critical for regulating specific molecular pathways and cellular responses. It is these lncRNAs that will likely elucidate new diagnostics and therapeutic strategies for not only T1D, but other related inflammatory and autoimmune diseases. Acknowledgments  Various colleagues and past students provided informative discussions, in particular Colleen Elso, Leanne Mackin, Edward Po-Fan Chu, Michelle Papadimitriou (née Ashton), May Abdulaziz Alsayb, Stuart Mannering, Helen Thomas, Thomas Kay, Guarang Jhala, Balasubramanian Krishnamurthy, Pablo Silveira, Ashley Mansell and Meredith O’Keeffe. Any novel insights provided here were in part stimulated by those discussions, whereas any errors or failure to cite certain work rest solely with the author.

References 1. Nolan CJ, Prentki M (2008) The islet β-cell: fuel responsive and vulnerable. Trends Endocrinol Metab 19:285–291. https://doi.org/10.1016/j. tem.2008.07.006 2. Regazzi R, Rodriguez-Trejo A, Jacovetti C (2016) Insulin secretion in health and disease: nutrients dictate the pace. Proc Nutr Soc 75:19–29. https://doi. org/10.1017/S0029665115004152 3. Eizirik DL, Pasquali L, Cnop M (2020) Pancreatic β-cells in type 1 and type 2 diabetes mellitus: different pathways to failure. Nat Rev Endocrinol 16:349– 362. https://doi.org/10.1038/s41574-­020-­0355-­7 4. Usmani-Brown S et  al (2019) β-cell responses to inflammation. Mol Metab 27S:S104–S113. https:// doi.org/10.1016/j.molmet.2019.06.013 5. DeFronzo RA et  al (2015) Type 2 diabetes mellitus. Nat Rev Dis Primers 1:15019. https://doi. org/10.1038/nrdp.2015.19 6. DiMeglio LA, Evans-Molina C, Oram RA (2018) Type 1 diabetes. Lancet 391:2449–2462. https://doi. org/10.1016/S0140-­6736(18)31320-­5 7. Cole JB, Florez JC (2020) Genetics of diabetes mellitus and diabetes complications. Nat Rev Nephrol 16:377–390. https://doi.org/10.1038/ s41581-­020-­0278-­5 8. Pullen TJ, Rutter GA (2014) Roles of lncRNAs in pancreatic β cell identity and diabetes susceptibility. Front Genet 5:193. https://doi.org/10.3389/ fgene.2014.00193 9. Motterle A, Gattesco S, Caille D, Meda P, Regazzi R (2015) Involvement of long non-coding RNAs in β cell failure at the onset of type 1 diabetes in NOD mice. Diabetologia 58:1827–1835. https://doi. org/10.1007/s00125-­015-­3641-­5 10. Motterle A et  al (2017) Identification of islet-­ enriched long non-coding RNAs contributing to β-cell failure in type 2 diabetes. Mol Metab 6:1407– 1418. https://doi.org/10.1016/j.molmet.2017.08.005 11. Singer RA, Sussel L (2018) Islet long noncoding RNAs: a playbook for discovery and characterization. Diabetes 67:1461–1470. https://doi. org/10.2337/dbi18-­0001 12. Ravasi T et al (2006) Experimental validation of the regulated expression of large numbers of non-coding RNAs from the mouse genome. Genome Res 16:11– 19. https://doi.org/10.1101/gr.4200206 13. Guttman M et al (2009) Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals. Nature 458:223–227. https:// doi.org/10.1038/nature07672 14. Iyer MK et al (2015) The landscape of long noncoding RNAs in the human transcriptome. Nat Genet 47:199–208. https://doi.org/10.1038/ng.3192 15. Ponting CP, Oliver PL, Reik W (2009) Evolution and functions of long noncoding RNAs. Cell 136:629– 641. https://doi.org/10.1016/j.cell.2009.02.006

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 111 16. Quinn JJ, Chang HY (2016) Unique features of long non-coding RNA biogenesis and function. Nat Rev Genet 17:47–62. https://doi.org/10.1038/ nrg.2015.10 17. Gloss BS, Dinger ME (2016) The specificity of long noncoding RNA expression. Biochim Biophys Acta 1859:16–22. https://doi.org/10.1016/j. bbagrm.2015.08.005 18. Mattick JS, Rinn JL (2015) Discovery and annotation of long noncoding RNAs. Nat Struct Mol Biol 22:5–7. https://doi.org/10.1038/nsmb.2942 19. Jarroux J, Morillon A, Pinskaya M (2017) History, discovery, and classification of lncRNAs. Adv Exp Med Biol 1008:1–46. https://doi. org/10.1007/978-­981-­10-­5203-­3_1 20. Guttman M, Rinn JL (2012) Modular regulatory principles of large non-coding RNAs. Nature 482:339–346. https://doi.org/10.1038/nature10887 21. Geisler S, Coller J (2013) RNA in unexpected places: long non-coding RNA functions in diverse cellular contexts. Nat Rev Mol Cell Biol 14:699–712. https:// doi.org/10.1038/nrm3679 22. Yao RW, Wang Y, Chen LL (2019) Cellular functions of long noncoding RNAs. Nat Cell Biol 21:542–551. https://doi.org/10.1038/s41556-­019-­0311-­8 23. Shields EJ, Petracovici AF, Bonasio R (2019) lncRedibly versatile: biochemical and biological functions of long noncoding RNAs. Biochem J 476:1083– 1104. https://doi.org/10.1042/BCJ20180440 24. Pervouchine DD et  al (2015) Enhanced transcriptome maps from multiple mouse tissues reveal evolutionary constraint in gene expression. Nat Commun 6:5903. https://doi.org/10.1038/ncomms6903 25. Moran I et  al (2012) Human β cell transcriptome analysis uncovers lncRNAs that are tissue-specific, dynamically regulated, and abnormally expressed in type 2 diabetes. Cell Metab 16:435–448. https://doi. org/10.1016/j.cmet.2012.08.010 26. Benner C et al (2014) The transcriptional landscape of mouse β cells compared to human β cells reveals notable species differences in long non-coding RNA and protein-coding gene expression. BMC Genomics 15:620. https://doi.org/10.1186/1471-­2164-­15-­620 27. Akerman I et  al (2017) Human pancreatic β cell lncRNAs control cell-specific regulatory networks. Cell Metab 25:400–411. https://doi.org/10.1016/j. cmet.2016.11.016 28. Mirza AH, Kaur S, Pociot F (2017) Long non-­ coding RNAs as novel players in β cell function and type 1 diabetes. Hum Genomics 11:17. https://doi. org/10.1186/s40246-­017-­0113-­7 29. Motterle A, Sanchez-Parra C, Regazzi R (2016) Role of long non-coding RNAs in the determination of β-cell identity. Diabetes Obes Metab 18 S3uppl 1:41–50. https://doi.org/10.1111/dom.12714 30. Carpenter S (2016) Long noncoding RNA: novel links between gene expression and innate immunity. Virus Res 212:137–145. https://doi.org/10.1016/j. virusres.2015.08.019

31. Atianand MK, Caffrey DR, Fitzgerald KA (2017) Immunobiology of long noncoding RNAs. Annu Rev Immunol 35:177–198. https://doi.org/10.1146/ annurev-­immunol-­041015-­055459 32. Wong WKM, Sorensen AE, Joglekar MV, Hardikar AA, Dalgaard LT (2018) Non-coding RNA in pancreas and β-cell development. Noncoding RNA 4:41. https://doi.org/10.3390/ncrna4040041 33. Bluestone JA, Herold K, Eisenbarth G (2010) Genetics, pathogenesis and clinical interventions in type 1 diabetes. Nature 464:1293–1300. https://doi. org/10.1038/nature08933 34. Coppieters KT et al (2012) Demonstration of islet-­ autoreactive CD8 T cells in insulitic lesions from recent onset and long-term type 1 diabetes patients. J Exp Med 209:51–60. https://doi.org/10.1084/ jem.20111187 35. Patterson CC et  al (2019) Worldwide estimates of incidence, prevalence and mortality of type 1 diabetes in children and adolescents: results from the International Diabetes Federation Diabetes Atlas, 9th edition. Diabetes Res Clin Pract 157:107842. https://doi.org/10.1016/j.diabres.2019.107842 36. Bizzarri C et  al (2010) Clinical presentation and autoimmune characteristics of very young children at the onset of type 1 diabetes mellitus. J Pediatr Endocrinol Metab 23:1151–1157. https://doi. org/10.1515/jpem.2010.180 37. Nyenwe EA, Kitabchi AE (2016) The evolution of diabetic ketoacidosis: an update of its etiology, pathogenesis and management. Metabolism 65:507– 521. https://doi.org/10.1016/j.metabol.2015.12.007 38. de Ferranti SD et  al (2014) Type 1 diabetes mellitus and cardiovascular disease: a scientific statement from the American Heart Association and American Diabetes Association. Circulation 130:1110–1130. https://doi.org/10.1161/CIR.0000000000000034 39. Monti MC et al (2007) Familial risk factors for microvascular complications and differential male-female risk in a large cohort of American families with type 1 diabetes. J Clin Endocrinol Metab 92:4650–4655. https://doi.org/10.1210/jc.2007-­1185 40. Fong DS, Aiello LP, Ferris FL 3rd, Klein R (2004) Diabetic retinopathy. Diabetes Care 27:2540–2553. https://doi.org/10.2337/diacare.27.10.2540 41. Finne P, Reunanen A, Stenman S, Groop PH, Gronhagen-Riska C (2005) Incidence of end-stage renal disease in patients with type 1 diabetes. JAMA 294:1782–1787. https://doi.org/10.1001/ jama.294.14.1782 42. Knip M, Simell O (2012) Environmental triggers of type 1 diabetes. Cold Spring Harb Perspect Med 2:a007690. https://doi.org/10.1101/cshperspect. a007690 43. Hamilton-Williams EE, Lorca GL, Norris JM, Dunne JL (2021) A triple threat? The role of diet, nutrition, and the microbiota in T1D pathogenesis. Front Nutr 8:600756. https://doi.org/10.3389/fnut.2021.600756 44. Eizirik DL, Colli ML, Ortis F (2009) The role of inflammation in insulitis and β-cell loss in type 1

112 diabetes. Nat Rev Endocrinol 5:219–226. https://doi. org/10.1038/nrendo.2009.21 45. Mannering SI, Pathiraja V, Kay TW (2016) The case for an autoimmune aetiology of type 1 diabetes. Clin Exp Immunol 183:8–15. https://doi.org/10.1111/ cei.12699 46. Atkinson MA et al (2011) How does type 1 diabetes develop?: the notion of homicide or β-cell suicide revisited. Diabetes 60:1370–1379. https://doi. org/10.2337/db10-­1797 47. Yang J et  al (2006) Islet-specific glucose-6-­ phosphatase catalytic subunit-related proteinreactive CD4+ T cells in human subjects. J Immunol 176:2781–2789. https://doi.org/10.4049/ jimmunol.176.5.2781 48. Palmer JP et al (1983) Insulin antibodies in insulin-­ dependent diabetics before insulin treatment. Science 222:1337–1339. https://doi.org/10.1126/ science.6362005 49. Baekkeskov S et  al (1990) Identification of the 64K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylase. Nature 347:151–156. https://doi. org/10.1038/347151a0 50. Solimena M et al (1996) ICA 512, an autoantigen of type I diabetes, is an intrinsic membrane protein of neurosecretory granules. EMBO J 15:2102–2114 51. Wenzlau JM et  al (2007) The cation efflux transporter ZnT8 (Slc30A8) is a major autoantigen in human type 1 diabetes. Proc Natl Acad Sci U S A 104:17040–17045. https://doi.org/10.1073/ pnas.0705894104 52. Ahmed S et  al (2019) Standardizing T-cell biomarkers in type 1 diabetes: challenges and recent advances. Diabetes 68:1366–1379. https://doi. org/10.2337/db19-­0119 53. Pathiraja V et  al (2015) Proinsulin-specific, HLA-DQ8, and HLA-DQ8-transdimer-restricted CD4+ T cells infiltrate islets in type 1 diabetes. Diabetes 64:172–182. https://doi.org/10.2337/ db14-­0858 54. Mannering SI et al (2005) The insulin A-chain epitope recognized by human T cells is posttranslationally modified. J Exp Med 202:1191–1197. https:// doi.org/10.1084/jem.20051251 55. Mannering SI et al (2009) The A-chain of insulin is a hot-spot for CD4+ T cell epitopes in human type 1 diabetes. Clin Exp Immunol 156:226–231. https:// doi.org/10.1111/j.1365-­2249.2009.03907.x 56. Delong T et  al (2016) Pathogenic CD4 T cells in type 1 diabetes recognize epitopes formed by peptide fusion. Science 351:711–714. https://doi. org/10.1126/science.aad2791 57. Mallone R, Brezar V, Boitard C (2011) T cell recognition of autoantigens in human type 1 diabetes: clinical perspectives. Clin Dev Immunol 2011:513210. https://doi.org/10.1155/2011/513210 58. Bonifacio E, Ziegler AG (2010) Advances in the prediction and natural history of type 1 diabetes.

T. C. Brodnicki Endocrinol Metab Clin N Am 39:513–525. https:// doi.org/10.1016/j.ecl.2010.05.007 59. So M et al (2021) Advances in Type 1 diabetes prediction using islet autoantibodies: beyond a simple count. Endocr Rev. https://doi.org/10.1210/endrev/ bnab013 60. Sosenko JM et al (2010) Trends of earlier and later responses of C-peptide to oral glucose challenges with progression to type 1 diabetes in diabetes prevention trial-type 1 participants. Diabetes Care 33:620–625. https://doi.org/10.2337/dc09-­1770 61. Sosenko JM et al (2010) Glucose excursions between states of glycemia with progression to type 1 diabetes in the diabetes prevention trial-type 1 (DPT-1). Diabetes 59:2386–2389. https://doi.org/10.2337/ db10-­0534 62. Driver JP, Serreze DV, Chen YG (2011) Mouse models for the study of autoimmune type 1 diabetes: a NOD to similarities and differences to human disease. Semin Immunopathol 33:67–87. https://doi. org/10.1007/s00281-­010-­0204-­1 63. Jayasimhan A, Mansour KP, Slattery RM (2014) Advances in our understanding of the pathophysiology of Type 1 diabetes: lessons from the NOD mouse. Clin Sci (Lond) 126:1–18. https://doi. org/10.1042/CS20120627 64. Diana J et al (2013) Crosstalk between neutrophils, B-1a cells and plasmacytoid dendritic cells initiates autoimmune diabetes. Nat Med 19:65–73. https:// doi.org/10.1038/nm.3042 65. Jansen A et al (1994) Immunohistochemical characterization of monocytes-macrophages and dendritic cells involved in the initiation of the insulitis and β-cell destruction in NOD mice. Diabetes 43:667– 675. https://doi.org/10.2337/diab.43.5.667 66. Dahlen E, Dawe K, Ohlsson L, Hedlund G (1998) Dendritic cells and macrophages are the first and major producers of TNF-α in pancreatic islets in the nonobese diabetic mouse. J Immunol 160:3585–3593 67. Serreze DV, Leiter EH (2001) Genes and cellular requirements for autoimmune diabetes susceptibility in nonobese diabetic mice. Curr Dir Autoimmun 4:31–67. https://doi.org/10.1159/000060527 68. Lehuen A, Diana J, Zaccone P, Cooke A (2010) Immune cell crosstalk in type 1 diabetes. Nat Rev Immunol 10:501–513. https://doi.org/10.1038/ nri2787 69. Unanue ER, Ferris ST, Carrero JA (2016) The role of islet antigen presenting cells and the presentation of insulin in the initiation of autoimmune diabetes in the NOD mouse. Immunol Rev 272:183–201. https://doi.org/10.1111/imr.12430 70. Bender C, Rajendran S, von Herrath MG (2020) New insights into the role of autoreactive CD8 T cells and cytokines in human type 1 diabetes. Front Endocrinol (Lausanne) 11:606434. https://doi. org/10.3389/fendo.2020.606434 71. Morgan NG, Richardson SJ (2018) Fifty years of pancreatic islet pathology in human type 1 diabetes: insights gained and progress made. Diabetologia

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 113 61:2499–2506. https://doi.org/10.1007/ s00125-­018-­4731-­y 72. Somoza N et  al (1994) Pancreas in recent onset insulin-­ dependent diabetes mellitus. Changes in HLA, adhesion molecules and autoantigens, restricted T cell receptor V beta usage, and cytokine profile. J Immunol 153:1360–1377 73. Uno S et al (2007) Macrophages and dendritic cells infiltrating islets with or without β cells produce tumour necrosis factor-α in patients with recent-­ onset type 1 diabetes. Diabetologia 50:596–601. https://doi.org/10.1007/s00125-­006-­0569-­9 74. Willcox A, Richardson SJ, Bone AJ, Foulis AK, Morgan NG (2009) Analysis of islet inflammation in human type 1 diabetes. Clin Exp Immunol 155:173–181. https://doi. org/10.1111/j.1365-­2249.2008.03860.x 75. Campbell-Thompson M et  al (2016) Insulitis and β-cell mass in the natural history of type 1 diabetes. Diabetes 65:719–731. https://doi.org/10.2337/ db15-­0779 76. Lundberg M, Seiron P, Ingvast S, Korsgren O, Skog O (2017) Insulitis in human diabetes: a histological evaluation of donor pancreases. Diabetologia 60:346–353. https://doi.org/10.1007/ s00125-­016-­4140-­z 77. Shapiro MR et al (2021) De-coding genetic risk variants in type 1 diabetes. Immunol Cell Biol 99:496– 508. https://doi.org/10.1111/imcb.12438 78. Allen C, Palta M, D'Alessio DJ (1991) Risk of diabetes in siblings and other relatives of IDDM subjects. Diabetes 40:831–836. https://doi.org/10.2337/ diab.40.7.831 79. Pociot F, Norgaard K, Hobolth N, Andersen O, Nerup J (1993) A nationwide population-based study of the familial aggregation of type 1 (insulin-dependent) diabetes mellitus in Denmark. Danish Study Group of Diabetes in Childhood. Diabetologia 36:870–875. https://doi.org/10.1007/BF00400364 80. Harjutsalo V, Podar T, Tuomilehto J (2005) Cumulative incidence of type 1 diabetes in 10,168 siblings of Finnish young-onset type 1 diabetic patients. Diabetes 54:563–569. https://doi. org/10.2337/diabetes.54.2.563 81. Ferrannini E et al (2010) Progression to diabetes in relatives of type 1 diabetic patients: mechanisms and mode of onset. Diabetes 59:679–685. https://doi. org/10.2337/db09-­1378 82. Todd JA, Bell JI, McDevitt HO (1987) HLA-DQ β gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus. Nature 329:599–604. https://doi.org/10.1038/329599a0 83. Erlich H et al (2008) HLA DR-DQ haplotypes and genotypes and type 1 diabetes risk: analysis of the type 1 diabetes genetics consortium families. Diabetes 57:1084–1092. https://doi.org/10.2337/ db07-­1331 84. Bell GI, Horita S, Karam JH (1984) A polymorphic locus near the human insulin gene is associated

with insulin-dependent diabetes mellitus. Diabetes 33:176–183. https://doi.org/10.2337/diab.33.2.176 85. Lucassen AM et  al (1993) Susceptibility to insulin dependent diabetes mellitus maps to a 4.1 kb segment of DNA spanning the insulin gene and associated VNTR.  Nat Genet 4:305–310. https://doi. org/10.1038/ng0793-­305 86. Ueda H et al (2003) Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 423:506–511. https://doi. org/10.1038/nature01621 87. Bottini N et al (2004) A functional variant of lymphoid tyrosine phosphatase is associated with type I diabetes. Nat Genet 36:337–338. https://doi. org/10.1038/ng1323 88. Vella A et al (2005) Localization of a type 1 diabetes locus in the IL2RA/CD25 region by use of tag single-nucleotide polymorphisms. Am J Hum Genet 76:773–779. https://doi.org/10.1086/429843 89. Barrett JC et  al (2009) Genome-wide association study and meta-analysis find that over 40 loci affect risk of type 1 diabetes. Nat Genet 41:703–707. https://doi.org/10.1038/ng.381 90. Bradfield JP et  al (2011) A genome-wide meta-­ analysis of six type 1 diabetes cohorts identifies multiple associated loci. PLoS Genet 7:e1002293. https://doi.org/10.1371/journal.pgen.1002293 91. Concannon P et  al (2009) Genome-wide scan for linkage to type 1 diabetes in 2,496 multiplex families from the Type 1 Diabetes Genetics Consortium. Diabetes 58:1018–1022. https://doi.org/10.2337/ db08-­1551 92. Cooper JD et al (2012) Confirmation of novel type 1 diabetes risk loci in families. Diabetologia 55:996– 1000. https://doi.org/10.1007/s00125-­012-­2450-­3 93. Hakonarson H et al (2007) A genome-wide association study identifies KIAA0350 as a type 1 diabetes gene. Nature 448:591–594. https://doi.org/10.1038/ nature06010 94. Howson JM, Walker NM, Smyth DJ, Todd JA, Type IDGC (2009) Analysis of 19 genes for association with type I diabetes in the Type I Diabetes Genetics Consortium families. Genes Immun 10 Suppl 1:S74– S84. https://doi.org/10.1038/gene.2009.96 95. Onengut-Gumuscu S et  al (2015) Fine mapping of type 1 diabetes susceptibility loci and evidence for colocalization of causal variants with lymphoid gene enhancers. Nat Genet 47:381–386. https://doi. org/10.1038/ng.3245 96. Todd JA et al (2007) Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet 39:857–864. https://doi. org/10.1038/ng2068 97. Wellcome Trust Case Control, C (2007) Genome-­ wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447:661–678. https://doi.org/10.1038/nature05911 98. Chiou J et  al (2021) Interpreting type 1 diabetes risk with genetics and single-cell epigenom-

114 ics. Nature 594:398–402. https://doi.org/10.1038/ s41586-­021-­03552-­w 99. Zhu M et al (2019) Identification of novel T1D risk loci and their association with age and islet function at diagnosis in autoantibody-positive T1D individuals: based on a two-stage genome-wide association study. Diabetes Care 42:1414–1421. https://doi. org/10.2337/dc18-­2023 100. Polychronakos C, Li Q (2011) Understanding type 1 diabetes through genetics: advances and prospects. Nat Rev Genet 12:781–792. https://doi.org/10.1038/ nrg3069 101. Pociot F et  al (2010) Genetics of type 1 diabetes: what's next? Diabetes 59:1561–1571. https://doi. org/10.2337/db10-­0076 102. Castellanos-Rubio A, Ghosh S (2019) Disease-­ associated SNPs in inflammation-related lncRNAs. Front Immunol 10:420. https://doi.org/10.3389/ fimmu.2019.00420 103. Mirza AH, Kaur S, Brorsson CA, Pociot F (2014) Effects of GWAS-associated genetic variants on lncRNAs within IBD and T1D candidate loci. PLoS One 9:e105723. https://doi.org/10.1371/journal. pone.0105723 104. Li C, Wei B, Zhao J (2021) Competing endogenous RNA network analysis explores the key lncRNAs, miRNAs, and mRNAs in type 1 diabetes. BMC Med Genet 14:35. https://doi.org/10.1186/ s12920-­021-­00877-­3 105. Smyth DJ et  al (2008) Shared and distinct genetic variants in type 1 diabetes and celiac disease. N Engl J Med 359:2767–2777. https://doi.org/10.1056/ NEJMoa0807917 106. Wallace C et al (2010) The imprinted DLK1-MEG3 gene region on chromosome 14q32.2 alters susceptibility to type 1 diabetes. Nat Genet 42:68–71. https:// doi.org/10.1038/ng.493 107. Tan IK et  al (2010) A recombination hotspot leads to sequence variability within a novel gene (AK005651) and contributes to type 1 diabetes susceptibility. Genome Res 20:1629–1638. https://doi. org/10.1101/gr.101881.109 108. Kumar V et  al (2013) Human disease-associated genetic variation impacts large intergenic non-­ coding RNA expression. PLoS Genet 9:e1003201. https://doi.org/10.1371/journal.pgen.1003201 109. Orom UA et al (2010) Long noncoding RNAs with enhancer-like function in human cells. Cell 143:46– 58. https://doi.org/10.1016/j.cell.2010.09.001 110. Guttman M et al (2011) lincRNAs act in the circuitry controlling pluripotency and differentiation. Nature 477:295–300. https://doi.org/10.1038/nature10398 111. Nyaga DM, Vickers MH, Jefferies C, Perry JK, O'Sullivan JM (2018) Type 1 diabetes mellitus-­ associated genetic variants contribute to overlapping immune regulatory networks. Front Genet 9:535. https://doi.org/10.3389/fgene.2018.00535 112. Guo H et  al (2015) Integration of disease association and eQTL data using a Bayesian colocalisation approach highlights six candidate causal genes

T. C. Brodnicki in immune-mediated diseases. Hum Mol Genet 24:3305–3313. https://doi.org/10.1093/hmg/ddv077 113. Hu X et  al (2014) Regulation of gene expression in autoimmune disease loci and the genetic basis of proliferation in CD4+ effector memory T cells. PLoS Genet 10:e1004404. https://doi.org/10.1371/ journal.pgen.1004404 114. de Goede OM et  al (2021) Population-scale tissue transcriptomics maps long non-coding RNAs to complex disease. Cell 184:2633–2648 e2619. https://doi.org/10.1016/j.cell.2021.03.050 115. Mutskov V, Felsenfeld G (2009) The human insulin gene is part of a large open chromatin domain specific for human islets. Proc Natl Acad Sci U S A 106:17419–17424. https://doi.org/10.1073/ pnas.0909288106 116. Tomlinson MJ, t. et  al (2014) Fine mapping and functional studies of risk variants for type 1 diabetes at chromosome 16p13.13. Diabetes 63:4360–4368. https://doi.org/10.2337/db13-­1785 117. Nieves-Bonilla JM, Kiaf B, Schuster C, Kissler S (2020) The type 1 diabetes candidate gene Dexi does not affect disease risk in the nonobese diabetic mouse model. Genes Immun 21:71–77. https://doi. org/10.1038/s41435-­019-­0083-­y 118. Trudeau JD et al (2000) Neonatal β-cell apoptosis: a trigger for autoimmune diabetes? Diabetes 49:1–7. https://doi.org/10.2337/diabetes.49.1.1 119. Turley S, Poirot L, Hattori M, Benoist C, Mathis D (2003) Physiological β cell death triggers priming of self-reactive T cells by dendritic cells in a type-1 diabetes model. J Exp Med 198:1527–1537. https:// doi.org/10.1084/jem.20030966 120. Christen U, Bender C, von Herrath MG (2012) Infection as a cause of type 1 diabetes? Curr Opin Rheumatol 24:417–423. https://doi.org/10.1097/ BOR.0b013e3283533719 121. Hoglund P et  al (1999) Initiation of autoimmune diabetes by developmentally regulated presentation of islet cell antigens in the pancreatic lymph nodes. J Exp Med 189:331–339. https://doi.org/10.1084/ jem.189.2.331 122. Gagnerault MC, Luan JJ, Lotton C, Lepault F (2002) Pancreatic lymph nodes are required for priming of β cell reactive T cells in NOD mice. J Exp Med 196:369–377. https://doi.org/10.1084/jem.20011353 123. Elling R et al (2018) Genetic models reveal cis and trans immune-regulatory activities for lincRNA-­ Cox2. Cell Rep 25:1511–1524 e1516. https://doi. org/10.1016/j.celrep.2018.10.027 124. Carpenter S et  al (2013) A long noncoding RNA mediates both activation and repression of immune response genes. Science 341:789–792. https://doi. org/10.1126/science.1240925 125. Li Z et al (2014) The long noncoding RNA THRIL regulates TNFα expression through its interaction with hnRNPL. Proc Natl Acad Sci U S A 111:1002– 1007. https://doi.org/10.1073/pnas.1313768111 126. Chan J et  al (2015) Cutting edge: a natural antisense transcript, AS-IL1α, controls inducible tran-

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 115 scription of the proinflammatory cytokine IL-1α. J Immunol 195:1359–1363. https://doi.org/10.4049/ jimmunol.1500264 127. Yang X, Bam M, Becker W, Nagarkatti PS, Nagarkatti M (2020) Long noncoding RNA AW112010 promotes the differentiation of inflammatory T cells by suppressing IL-10 expression through histone demethylation. J Immunol 205:987–993. https://doi. org/10.4049/jimmunol.2000330 128. Westra HJ et  al (2018) Fine-mapping and functional studies highlight potential causal variants for rheumatoid arthritis and type 1 diabetes. Nat Genet 50:1366–1374. https://doi.org/10.1038/ s41588-­018-­0216-­7 129. Li R et al (2018) MEG3–4 is a miRNA decoy that regulates IL-1β abundance to initiate and then limit inflammation to prevent sepsis during lung infection. Sci Signal 11:eaao2387. https://doi.org/10.1126/scisignal.aao2387 130. You L et al (2016) Downregulation of long noncoding RNA Meg3 affects insulin synthesis and secretion in mouse pancreatic β cells. J Cell Physiol 231:852–862. https://doi.org/10.1002/jcp.25175 131. Wang N et  al (2018) Long noncoding RNA Meg3 regulates Mafa expression in mouse β cells by inactivating Rad21, Smc3 or Sin3α. Cell Physiol Biochem 45:2031–2043. https://doi.org/10.1159/000487983 132. Zhou Y et  al (2007) Activation of p53 by MEG3 non-coding RNA.  J Biol Chem 282:24731–24742. https://doi.org/10.1074/jbc.M702029200 133. Lu KH et al (2013) Long non-coding RNA MEG3 inhibits NSCLC cells proliferation and induces apoptosis by affecting p53 expression. BMC Cancer 13:461. https://doi.org/10.1186/1471-­2407-­13-­461 134. Colombo T, Farina L, Macino G, Paci P (2015) PVT1: a rising star among oncogenic long noncoding RNAs. Biomed Res Int 2015:304208. https://doi. org/10.1155/2015/304208 135. Fu J et al (2020) LncRNA PVT1 links Myc to glycolytic metabolism upon CD4(+) T cell activation and Sjogren's syndrome-like autoimmune response. J Autoimmun 107:102358. https://doi.org/10.1016/j. jaut.2019.102358 136. Millis MP, Bowen D, Kingsley C, Watanabe RM, Wolford JK (2007) Variants in the plasmacytoma variant translocation gene (PVT1) are associated with end-stage renal disease attributed to type 1 diabetes. Diabetes 56:3027–3032. https://doi. org/10.2337/db07-­0675 137. Jin K et  al (2019) Long non-coding RNA PVT1 interacts with MYC and its downstream molecules to synergistically promote tumorigenesis. Cell Mol Life Sci 76:4275–4289. https://doi.org/10.1007/ s00018-­019-­03222-­1 138. Tai N, Wong FS, Wen L (2016) The role of the innate immune system in destruction of pancreatic β cells in NOD mice and humans with type I diabetes. J Autoimmun 71:26–34. https://doi.org/10.1016/j. jaut.2016.03.006

139. Stunault MI, Bories G, Guinamard RR, Ivanov S (2018) Metabolism plays a key role during macrophage activation. Mediat Inflamm 2018:2426138. https://doi.org/10.1155/2018/2426138 140. Thayer TC et  al (2011) Superoxide production by macrophages and T cells is critical for the induction of autoreactivity and type 1 diabetes. Diabetes 60:2144–2151. https://doi.org/10.2337/db10-­1222 141. Shanmugam N, Reddy MA, Guha M, Natarajan R (2003) High glucose-induced expression of proinflammatory cytokine and chemokine genes in monocytic cells. Diabetes 52:1256–1264. https://doi. org/10.2337/diabetes.52.5.1256 142. Das S et  al (2018) Diabetes mellitus-induced long noncoding RNA Dnm3os regulates macrophage functions and inflammation via nuclear mechanisms. Arterioscler Thromb Vasc Biol 38:1806–1820. https://doi.org/10.1161/ATVBAHA.117.310663 143. Zhang H et  al (2019) Cancer-associated fibroblast-­ promoted lncRNA DNM3OS confers radioresistance by regulating DNA damage response in esophageal squamous cell carcinoma. Clin Cancer Res 25:1989–2000. https://doi.org/10.1158/1078­­0432.CCR-­18-­0773 144. Wang P et  al (2016) miR-214/199a/199a* cluster levels predict poor survival in hepatocellular carcinoma through interference with cell-cycle regulators. Oncotarget 7:929–945. https://doi.org/10.18632/ oncotarget.6137 145. el Azzouzi H et  al (2013) The hypoxia-inducible microRNA cluster miR-199a approximately 214 targets myocardial PPARδ and impairs mitochondrial fatty acid oxidation. Cell Metab 18:341–354. https:// doi.org/10.1016/j.cmet.2013.08.009 146. Lakhia R et  al (2020) Interstitial microRNA miR-­ 214 attenuates inflammation and polycystic kidney disease progression. JCI Insight 5:e133785. https:// doi.org/10.1172/jci.insight.133785 147. Zgheib C, Hodges MM, Hu J, Liechty KW, Xu J (2017) Long non-coding RNA Lethe regulates hyperglycemia-induced reactive oxygen species production in macrophages. PLoS One 12:e0177453. https://doi.org/10.1371/journal.pone.0177453 148. Rapicavoli NA et  al (2013) A mammalian pseudogene lncRNA at the interface of inflammation and anti-inflammatory therapeutics. elife 2:e00762. https://doi.org/10.7554/eLife.00762 149. Wu H, Wen F, Jiang M, Liu Q, Nie Y (2018) LncRNA uc.48+ is involved in the diabetic immune and inflammatory responses mediated by P2X7 receptor in RAW264.7 macrophages. Int J Mol Med 42:1152– 1160. https://doi.org/10.3892/ijmm.2018.3661 150. Berghaus LJ et al (2010) Innate immune responses of primary murine macrophage-lineage cells and RAW 264.7 cells to ligands of Toll-like receptors 2, 3, and 4. Comp Immunol Microbiol Infect Dis 33:443–454. https://doi.org/10.1016/j.cimid.2009.07.001 151. Donath MY, Storling J, Berchtold LA, Billestrup N, Mandrup-Poulsen T (2008) Cytokines and β-cell biology: from concept to clinical translation.

116 Endocr Rev 29:334–350. https://doi.org/10.1210/ er.2007-­0033 152. Singh B, Nikoopour E, Huszarik K, Elliott JF, Jevnikar AM (2011) Immunomodulation and regeneration of islet β cells by cytokines in autoimmune type 1 diabetes. J Interf Cytokine Res 31:711–719. https://doi.org/10.1089/jir.2011.0025 153. Kaminitz A, Stein J, Yaniv I, Askenasy N (2007) The vicious cycle of apoptotic β-cell death in type 1 diabetes. Immunol Cell Biol 85:582–589. https://doi. org/10.1038/sj.icb.7100093 154. Kim KA, Lee MS (2009) Recent progress in research on β-cell apoptosis by cytokines. Front Biosci (Landmark Ed) 14:657–664. https://doi. org/10.2741/3271 155. Ramos-Rodriguez M et  al (2019) The impact of proinflammatory cytokines on the β-cell regulatory landscape provides insights into the genetics of type 1 diabetes. Nat Genet 51:1588–1595. https://doi. org/10.1038/s41588-­019-­0524-­6 156. Kutlu B et al (2003) Discovery of gene networks regulating cytokine-induced dysfunction and apoptosis in insulin-producing INS-1 cells. Diabetes 52:2701– 2719. https://doi.org/10.2337/diabetes.52.11.2701 157. Ortis F et  al (2010) Cytokines interleukin-1β and tumor necrosis factor-α regulate different transcriptional and alternative splicing networks in primary β-cells. Diabetes 59:358–374. https://doi. org/10.2337/db09-­1159 158. Eizirik DL et al (2012) The human pancreatic islet transcriptome: expression of candidate genes for type 1 diabetes and the impact of pro-inflammatory cytokines. PLoS Genet 8:e1002552. https://doi. org/10.1371/journal.pgen.1002552 159. Li B et  al (2014) RNA-seq methods for identifying differentially expressed gene in human pancreatic islet cells treated with pro-inflammatory cytokines. Mol Biol Rep 41:1917–1925. https://doi. org/10.1007/s11033-­013-­3016-­2 160. Sun C et  al (2016) Insights from lncRNAs profiling of MIN6 β cells undergoing inflammation. Mediat Inflamm 2016:9275106. https://doi. org/10.1155/2016/9275106 161. Cianciaruso C et  al (2017) Primary human and rat β-cells release the intracellular autoantigens GAD65, IA-2, and proinsulin in exosomes together with cytokine-induced enhancers of immunity. Diabetes 66:460–473. https://doi.org/10.2337/db16-­0671 162. Guay C et al (2019) Lymphocyte-derived exosomal microRNAs promote pancreatic β cell death and may contribute to type 1 diabetes development. Cell Metab 29:348–361 e346. https://doi.org/10.1016/j. cmet.2018.09.011 163. Ribeiro D et  al (2017) Extracellular vesicles from human pancreatic islets suppress human islet amyloid polypeptide amyloid formation. Proc Natl Acad Sci U S A 114:11127–11132. https://doi. org/10.1073/pnas.1711389114 164. Krishnan P, Syed F, Jiyun Kang N, Mirmira RG, Evans-Molina C (2019) Profiling of RNAs from

T. C. Brodnicki human islet-derived exosomes in a model of type 1 diabetes. Int J Mol Sci 20:5903. https://doi. org/10.3390/ijms20235903 165. Ansari MJ et  al (2003) The programmed death-1 (PD-1) pathway regulates autoimmune diabetes in nonobese diabetic (NOD) mice. J Exp Med 198:63– 69. https://doi.org/10.1084/jem.20022125 166. Fife BT et  al (2006) Insulin-induced remission in new-onset NOD mice is maintained by the PD-1-­ PD-L1 pathway. J Exp Med 203:2737–2747. https:// doi.org/10.1084/jem.20061577 167. Guleria I et  al (2007) Mechanisms of PDL1-­ mediated regulation of autoimmune diabetes. Clin Immunol 125:16–25. https://doi.org/10.1016/j. clim.2007.05.013 168. Osum KC et  al (2018) Interferon-γ drives programmed death-ligand 1 expression on islet β cells to limit T cell function during autoimmune diabetes. Sci Rep 8:8295. https://doi.org/10.1038/ s41598-­018-­26471-­9 169. Giannoukakis N, Rudert WA, Trucco M, Robbins PD (2000) Protection of human islets from the effects of interleukin-1β by adenoviral gene transfer of an Iκ B repressor. J Biol Chem 275:36509–36513. https:// doi.org/10.1074/jbc.M005943200 170. Heimberg H et  al (2001) Inhibition of cytokine-­ induced NF-κB activation by adenovirus-mediated expression of a NF-κB super-repressor prevents β-cell apoptosis. Diabetes 50:2219–2224. https:// doi.org/10.2337/diabetes.50.10.2219 171. Eldor R et al (2006) Conditional and specific NF-κB blockade protects pancreatic β cells from diabetogenic agents. Proc Natl Acad Sci U S A 103:5072– 5077. https://doi.org/10.1073/pnas.0508166103 172. Ortis F et al (2006) Cytokine-induced proapoptotic gene expression in insulin-producing cells is related to rapid, sustained, and nonoscillatory nuclear factor-κB activation. Mol Endocrinol 20:1867–1879. https://doi.org/10.1210/me.2005-­0268 173. Ortis F et  al (2008) Induction of nuclear factor-κB and its downstream genes by TNF-α and IL-1β has a pro-apoptotic role in pancreatic β cells. Diabetologia 51:1213–1225. https://doi.org/10.1007/ s00125-­008-­0999-­7 174. Shan Z, Xu B, Mikulowska-Mennis A, Michie SA (2014) CCR7 directs the recruitment of T cells into inflamed pancreatic islets of nonobese diabetic (NOD) mice. Immunol Res 58:351–357. https://doi. org/10.1007/s12026-­014-­8500-­9 175. Christen U, Kimmel R (2020) Chemokines as drivers of the autoimmune destruction in type 1 diabetes: opportunity for therapeutic intervention in consideration of an optimal treatment schedule. Front Endocrinol (Lausanne) 11:591083. https://doi. org/10.3389/fendo.2020.591083 176. Gonzalez-Moro I et  al (2020) The T1D-associated lncRNA Lnc13 modulates human pancreatic β cell inflammation by allele-specific stabilization of STAT1 mRNA. Proc Natl Acad Sci U S A 117:9022– 9031. https://doi.org/10.1073/pnas.1914353117

6  A Role for lncRNAs in Regulating Inflammatory and Autoimmune Responses Underlying Type 1 Diabetes 117 177. Castellanos-Rubio A et al (2016) A long noncoding RNA associated with susceptibility to celiac disease. Science 352:91–95. https://doi.org/10.1126/science. aad0467 178. Ding H et al (2020) LncRNA MALAT1 induces the dysfunction of β cells via reducing the histone acetylation of the PDX-1 promoter in type 1 diabetes. Exp Mol Pathol 114:104432. https://doi.org/10.1016/j. yexmp.2020.104432 179. Macfarlane WM et al (2000) Glucose modulation of insulin mRNA levels is dependent on transcription factor PDX-1 and occurs independently of changes in intracellular Ca2+. Diabetes 49:418–423. https:// doi.org/10.2337/diabetes.49.3.418 180. Melloul D (2004) Transcription factors in islet development and physiology: role of PDX-1  in β-cell function. Ann N Y Acad Sci 1014:28–37. https://doi. org/10.1196/annals.1294.003 181. Artner I, Hang Y, Guo M, Gu G, Stein R (2008) MafA is a dedicated activator of the insulin gene in  vivo. J Endocrinol 198:271–279. https://doi. org/10.1677/JOE-­08-­0063 182. Zatterale F et  al (2019) Chronic adipose tissue inflammation linking obesity to insulin resistance and type 2 diabetes. Front Physiol 10:1607. https:// doi.org/10.3389/fphys.2019.01607 183. Sladek R et  al (2007) A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature 445:881–885. https://doi.org/10.1038/ nature05616 184. Diabetes Genetics Initiative of Broad Institute of, H et al (2007) Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science 316:1331–1336. https://doi.org/10.1126/ science.1142358 185. Voight BF et al (2010) Twelve type 2 diabetes susceptibility loci identified through large-scale association analysis. Nat Genet 42:579–589. https://doi. org/10.1038/ng.609 186. Morris AP et al (2012) Large-scale association analysis provides insights into the genetic architecture and pathophysiology of type 2 diabetes. Nat Genet 44:981–990. https://doi.org/10.1038/ng.2383 187. Pasquali L et  al (2014) Pancreatic islet enhancer clusters enriched in type 2 diabetes risk-­associated variants. Nat Genet 46:136–143. https://doi. org/10.1038/ng.2870 188. Sathishkumar C, Prabu P, Mohan V, Balasubramanyam M (2018) Linking a role of lncRNAs (long non-coding RNAs) with insulin resistance, accelerated senescence, and inflammation in patients with type 2 diabetes. Hum Genomics 12:41. https://doi.org/10.1186/s40246-­018-­0173-­3 189. Dieter C, Lemos NE, Correa NRF, Assmann TS, Crispim D (2021) The impact of lncRNAs in diabetes mellitus: a systematic review and in silico analyses. Front Endocrinol (Lausanne) 12:602597. https:// doi.org/10.3389/fendo.2021.602597 190. Reddy MA et  al (2021) lncRNA DRAIR is downregulated in diabetic monocytes and modulates the

inflammatory phenotype via epigenetic mechanisms. JCI Insight 6:143289. https://doi.org/10.1172/jci. insight.143289 191. Marchetti P et  al (2020) A direct look at the dysfunction and pathology of the β cells in human type 2 diabetes. Semin Cell Dev Biol 103:83–93. https:// doi.org/10.1016/j.semcdb.2020.04.005 192. Kameswaran V et al (2014) Epigenetic regulation of the DLK1-MEG3 microRNA cluster in human type 2 diabetic islets. Cell Metab 19:135–145. https://doi. org/10.1016/j.cmet.2013.11.016 193. Carter S, Miard S, Boivin L, Salle-Lefort S, Picard F (2018) Loss of Malat1 does not modify age- or diet-­ induced adipose tissue accretion and insulin resistance in mice. PLoS One 13:e0196603. https://doi. org/10.1371/journal.pone.0196603 194. Zhang Y, Fu Y, Zheng Y, Wen Z, Wang C (2020) Identification of differentially expressed mRNA and the Hub mRNAs modulated by lncRNA Meg3 as a competing endogenous RNA in brown adipose tissue of mice on a high-fat diet. Adipocytes 9:346– 358. https://doi.org/10.1080/21623945.2020.178928 3 195. White RR et al (2015) Comprehensive transcriptional landscape of aging mouse liver. BMC Genomics 16:899. https://doi.org/10.1186/s12864-­015-­2061-­8 196. Mahpour A, Mullen AC (2021) Our emerging understanding of the roles of long non-coding RNAs in normal liver function, disease, and malignancy. JHEP Rep 3:100177. https://doi.org/10.1016/j. jhepr.2020.100177 197. Zhu X, Wu YB, Zhou J, Kang DM (2016) Upregulation of lncRNA MEG3 promotes hepatic insulin resistance via increasing FoxO1 expression. Biochem Biophys Res Commun 469:319–325. https://doi.org/10.1016/j.bbrc.2015.11.048 198. Zhu X et al (2019) lncRNA MEG3 promotes hepatic insulin resistance by serving as a competing endogenous RNA of miR-214 to regulate ATF4 expression. Int J Mol Med 43:345–357. https://doi.org/10.3892/ ijmm.2018.3975 199. Yan C, Chen J, Chen N (2016) Long noncoding RNA MALAT1 promotes hepatic steatosis and insulin resistance by increasing nuclear SREBP-1c protein stability. Sci Rep 6:22640. https://doi.org/10.1038/ srep22640 200. Ma Z et al (2019) LncRNA expression profile during autophagy and Malat1 function in macrophages. PLoS One 14:e0221104. https://doi.org/10.1371/ journal.pone.0221104 201. Wu J et  al (2018) The long noncoding RNA MALAT1 induces tolerogenic dendritic cells and regulatory T cells via miR155/dendritic cell-specific intercellular adhesion molecule-3 grabbing nonintegrin/IL10 axis. Front Immunol 9:1847. https://doi. org/10.3389/fimmu.2018.01847 202. Hewitson JP et al (2020) Malat1 suppresses immunity to infection through promoting expression of Maf and IL-10  in Th cells. J Immunol 204:2949– 2960. https://doi.org/10.4049/jimmunol.1900940

118 203. Murray PJ, Smale ST (2012) Restraint of inflammatory signaling by interdependent strata of negative regulatory pathways. Nat Immunol 13:916–924. https://doi.org/10.1038/ni.2391 204. Robinson EK, Covarrubias S, Carpenter S (2020) The how and why of lncRNA function: an innate immune perspective. Biochim Biophys Acta Gene Regul Mech 1863:194419. https://doi.org/10.1016/j. bbagrm.2019.194419 205. Li L, Chang HY (2014) Physiological roles of long noncoding RNAs: insight from knockout mice. Trends Cell Biol 24:594–602. https://doi. org/10.1016/j.tcb.2014.06.003

T. C. Brodnicki 206. Kaddis JS, Pugliese A, Atkinson MA (2015) A run on the biobank: what have we learned about type 1 diabetes from the nPOD tissue repository? Curr Opin Endocrinol Diabetes Obes 22:290–295. https:// doi.org/10.1097/MED.0000000000000171 207. Joshi K et al (2021) Modeling type 1 diabetes using pluripotent stem cell technology. Front Endocrinol (Lausanne) 12:635662. https://doi.org/10.3389/ fendo.2021.635662 208. Borkiewicz L, Kalafut J, Dudziak K, Przybyszewska-­ Podstawka A, Telejko I (2021) Decoding lncRNAs. Cancers (Basel) 13. https://doi.org/10.3390/ cancers13112643

Part IV Long Noncoding RNAs as Therapeutic Targets

7

LncRNA Biomarkers of Inflammation and Cancer Roman E. Reggiardo, Sreelakshmi Velandi Maroli, and Daniel H. Kim

Abstract

Keywords

Long noncoding RNAs (lncRNAs) are promising candidates as biomarkers of inflammation and cancer. LncRNAs have several properties that make them well-suited as molecular markers of disease: (1) many lncRNAs are expressed in a tissue-specific manner, (2) distinct lncRNAs are upregulated based on different inflammatory or oncogenic stimuli, (3) lncRNAs released from cells are packaged and protected in extracellular vesicles, and (4) circulating lncRNAs in the blood are detectable using various RNA sequencing approaches. Here we focus on the potential for lncRNA biomarkers to detect inflammation and cancer, highlighting key biological, technological, and analytical considerations that will help advance the development of lncRNA-­based liquid biopsies.

Biomarkers · Therapeutics · lncRNA · Transposable elements

Roman E.  Reggiardo and Sreelakshmi Velandi Maroli contributed equally with all other contributors.

R. E. Reggiardo Department of Biomolecular Engineering, University of California Santa Cruz, Santa Cruz, CA, USA S. V. Maroli Department of Molecular, Cell and Developmental Biology, University of California Santa Cruz, Santa Cruz, CA, USA

7.1

Long Noncoding RNAs

Long noncoding RNAs (lncRNAs) are greater than 200 nucleotides in length and are not translated into proteins [1–3]. Their length serves to distinguish them from small noncoding RNAs, such as microRNAs (miRNAs) [4]. Many lncRNAs have features similar to protein-coding genes: they often have more than one exon, and about 60% of lncRNAs are polyadenylated [5]. LncRNAs are a heterogeneous group of noncoding RNAs whose expression patterns depend largely on tissue and cellular contexts D. H. Kim (*) Department of Biomolecular Engineering, University of California Santa Cruz, Santa Cruz, CA, USA Institute for the Biology of Stem Cells, University of California Santa Cruz, Santa Cruz, CA, USA Genomics Institute, University of California Santa Cruz, Santa Cruz, CA, USA Center for Molecular Biology of RNA, University of California Santa Cruz, Santa Cruz, CA, USA Canary Center at Stanford for Cancer Early Detection, Stanford University School of Medicine, Stanford, CA, USA e-mail: [email protected]

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_7

121

122

[6–9]. Additionally, their secondary structures enable lncRNAs to interact with proteins [10]. Some of the known mechanisms of action of lncRNAs are as: (i) scaffolds for protein complexes, (ii) guides for protein targeting, (iii) transcriptional enhancers, (iv) decoys to release proteins from chromatin, and (v) antagonists for other regulatory noncoding RNAs, such as miRNAs [11–13]. Based on their distance and relative orientation to nearby coding genes, lncRNAs are classified as sense, antisense, bidirectional, intronic, intergenic, promoter-associated, or untranslated region-associated lncRNAs [5]. Genomic technologies such as RNA sequencing (RNA-seq) have enabled robust characterization of lncRNA expression and distribution throughout the human genome, but classification of these diverse transcripts remains a challenge [5, 14].

7.1.1 Transposable Element Sequences in lncRNAs Many lncRNA gene bodies are initiated in repetitive sequence elements called transposable elements (TEs), with 75% containing at least one exon that overlaps a TE insertion into the human genome [15]. In addition to seeding lncRNA evolution, these TE-derived sequences may contribute to lncRNA function [16, 17].

7.1.1.1 Transposable Elements TEs are mobile DNA elements that can replicate and re-insert themselves into different locations within the host genome. This retrotransposition has resulted in TEs accounting for more than 50% of the human genome [15, 18, 19]. Much like lncRNAs, TE insertions were once considered genomic ‘junk’. In mammals, TEs are divided into four classes: long interspersed elements (LINEs), short interspersed elements (SINEs), long terminal repeat (LTR) retrotransposons, and DNA transposons. LINE, SINE, and LTR retrotransposons transpose via RNA intermediates using a “copy and paste mechanism”, while DNA transposons transpose directly as DNA.

R. E. Reggiardo et al.

In humans, LINEs are about 6  kb long and have an internal RNA polymerase II promoter and two open reading frames (ORFs). There are three distantly related LINE families found in the human genome: LINE1, LINE2, and LINE3, with some LINE1 elements still active. SINEs, including Alu and SVA elements, are short (100– 400  bp) sequences, incapable of autonomous ­retrotransposition, that harbor an internal RNA polymerase III promoter and are polyadenylated. LINE1-encoded proteins recognize and bind to non-autonomous SINE sequences in trans to mediate their mobilization [20]. Together, LINEs and SINEs comprise ~33% of the human genome. LTR retrotransposons are flanked by direct repeats that contain transcriptional regulatory elements, and LTR retrotransposons contain gag and pol genes, which encode a protease, reverse transcriptase, RNase-H, and an integrase. These retrotransposons comprise 8% of the human genome. In particular, LTR retrotransposons called human endogenous retroviruses (HERVs) constitute about 1% of the human genome. TEs act as a source of cis elements regulating adjacent host genes, such as promoters [21], transcription factor binding sites [22, 23], enhancers [24, 25], or insulators [26, 27]. TEs can also be ‘exonized’ into novel coding and noncoding exons [28]. As a source of noncoding exons, TEs have been shown to contribute substantially to untranslated regions [29, 30] and to alternatively spliced exons of protein-coding genes [31].

7.1.1.2 Transposable Element-Derived lncRNAs TEs are enriched within lncRNA exons relative to protein-coding gene exons, and it is estimated that ~40% of lncRNA nucleotides are TE derived, with the majority of lncRNAs containing at least one TE element (Fig.  7.1) [32]. Numerous studies have elucidated TE-associated lncRNA functions. For example, Alu elements in lncRNAs play a significant role in STAU1mediated mRNA decay by binding to complementary Alu elements in the 3’ UTRs of mRNAs [33]. Tandem inverted Alu elements give rise to double stranded RNA (dsRNA), and Alu dsRNA can induce a type-I interferon response by acti-

7  LncRNA Biomarkers of Inflammation and Cancer

123

Fig. 7.1  Transposable element-derived lncRNAs. SINE, LINE, and LTR sequences can comprise both exonic and intronic regions of lncRNAs. (Created with BioRender.com)

vation of a cytosolic nucleic acid sensing pathway when there is a loss of adenosine-to-inosine (A-to-I) editing. This RNA editing is catalyzed by the ADAR family of adenosine deaminases that act on dsRNA templates. An essential role of A-to-I editing is the inhibition of the immune response against self dsRNA by editing such templates [16]. TE-derived RNAs are related to inflammation in the context of neurological disorders and cancer. A mutation in the LINE1-derived sequence of a lncRNA is associated with infantile encephalopathy [34]. Furthermore, TE sequences are present in exonic regions of lncRNAs such as UCA1, HULC and XIST, which are known to be tissue specific and linked with tissue specific cancers (discussed more in Sect. 7.3). Recent studies have shown that lncRNAs containing HERVH elements are expressed specifically in human embryonic stem cells (hESCs) and are required for maintaining hESC identity. The LTR region of HERVH elements acts as an enhancer and induces stem cell-specific expression of surrounding genes. Moreover, transcribed HERVH lncRNAs interact with pluripotency-related factors such as OCT4, suggesting that they act as a scaffold, recruiting those factors to HERVH LTR regions in the genome [35, 36].

7.1.2 Tissue-Specific lncRNAs in Embryonic and Adult Tissues LncRNAs are linked to numerous processes, including mammalian development [12]. During embryonic development, lncRNAs are expressed in a lineage-specific manner in epidermal [37], neural [38], hematopoietic [39], cardiac [40] and mammary gland development [41, 42]. Functional roles for these lncRNAs during cell differentiation have been uncovered by discovering their association with epigenetic modifiers. Polycomb group (PcG) proteins are epigenetic modifiers involved in regulating developmental genes in multiple cell types and tissue contexts, including embryonic and adult stem cells, and they are essential for cell fate transitions and normal development. These PcG proteins interact with lncRNAs such as Xist [10] and HOTAIR [43]. One study performed a RIP-PCR assay and revealed that ~20% of lncRNAs expressed in various tissues physically interact with several members of PcG protein complexes [44]. RNA interference-based depletion of various Polycomb repressive complex 2 (PRC2)-associated lncRNAs resulted in the activation of genes known to be repressed by PRC2.

R. E. Reggiardo et al.

124

LncRNAs are differentially expressed across a number of adult tissue types, distinct cell types, and within specific cellular compartments [45, 46], suggesting specific functional roles in these locations. For example, a liver-specific lncRNA, LncLSTR [47], is involved in the clearance of triglycerides and helps maintain systemic lipid homeostasis. OLMALINC [48] is a brain-­specific lncRNA essential for the regulation of genes responsible for human oligodendrocyte maturation. Heart-specific lncRNA Fendrr [49] is essential for development of the mouse heart wall and has an orthologous transcript in human. Linc-MD1 [50] is a skeletal muscle-specific lncRNA that aids in the regulation of muscle terminal differentiation by acting as a competing endogenous RNA for the binding of two miRNAs, miR-133 and miR-135. During retinal cell differentiation, lncRNA Six3OS [51] acts as a molecular scaffold and recruits histone modifying enzymes. Additionally, an epigenetic footprinting study across 111 reference epigenomes revealed tissue-specific epigenetic regulation of 3753 lncRNAs (67% of the examined lncRNAs), with 54% active in one of the 14 cell/tissue types and an additional 15% in two or three cell/tissue types [52].

7.2

LncRNAs in Inflammation

Inflammation is associated with an increased risk for many diseases, including cancer, and lncRNAs are involved in inflammatory signaling pathways such as NF-κB signaling [53] and in inflammatory diseases such as atherosclerosis [54].

7.2.1 L  ncRNAs Regulated by NF-κB Signaling NF-κB regulates the expression of over 300 noncoding RNAs in mouse fibroblasts treated with the pro-inflammatory cytokine TNFα, including the lncRNA Lethe and 165 other lncRNAs [55]. IL-1β, another pro-inflammatory cytokine that initiates NF-κB signaling, upregulates only some of these same lncRNAs, indicating that different

inflammatory stimuli lead to the upregulation of distinct lncRNAs. Inflammatory stimulation also leads to strong upregulation of a subset of lncRNAs in mouse macrophages [56]. In particular, lincRNA-Cox2 is induced several ­ hundred-­fold and is dependent on NF-κB. TNFα also induces the NKILA lncRNA in a breast cancer cell line [57], and both NKILA and Lethe are involved in the negative regulation of NF-κB in breast cells and fibroblasts, respectively. Additionally, NF-κB signaling also activates the expression of lincRNA-Tnfaip3 in mouse macrophages [58]. Given the specificity of mammalian lncRNAs that respond to inflammatory stimuli, lncRNAs are potentially promising biomarkers of inflammation.

7.2.2 LncRNAs in Inflammatory Diseases LncRNAs are implicated in inflammatory diseases such as atherosclerosis, where a lncRNA called VINAS is involved in the regulation of NF-κB and MAPK signaling [54]. Knockdown of VINAS led to a decrease in inflammatory gene expression and revealed a role for this lncRNA in the regulation of vascular inflammation. Additionally, the lncRNA MALAT1 is involved in plaque inflammation and atherosclerosis, exhibiting an anti-inflammatory role in the context of this disease [59]. In systemic lupus erythematous (SLE), circulating lncRNAs PVT1 and FAS-AS1 can act as biomarkers of this chronic autoimmune disorder [60]. LncRNAs are also potential biomarkers for diagnosing the autoimmune disease rheumatoid arthritis, where circulating lncRNAs such as HOTAIR can serve as a potential biomarker for this inflammatory disease [61].

7.3

LncRNAs in Cancer

Large-scale RNA-seq efforts have revealed that lncRNAs are dysregulated in many different cancers, often in a cancer-specific manner [62]. Notably, some cancers may arise through dedif-

7  LncRNA Biomarkers of Inflammation and Cancer

125

Fig. 7.2  Transposable element-derived lncRNAs in various cancers. (Created with BioRender.com)

ferentiation toward a more stem cell-like state [63, 64]. One group found that lncRNAs such as AFAP1-AS1 and HOTAIR, which are overexpressed in several cancers, are also overexpressed during early human embryonic developmental stages, but not in adult tissues [65]. LncRNAs such as H19, MALAT1 and PCA3 were linked to cancer due to their high expression in tumor tissues and were identified before the advent of next generation sequencing (Fig. 7.2) [66–68]. Dysregulated expression of lncRNAs can be linked to the development of different types of tumors and can be detected in patient body fluids for many cancer types [69]. LncRNA PCA3 has been approved as a urine biomarker for prostate cancer by the US Food and Drug Administration [70]. This lncRNA exhibits better sensitivity and specificity when compared to the widely used PSA blood test, due to its higher expression in prostate cancer patients [71]. HOTAIR is another lncRNA which is highly expressed in saliva samples of oral squamous cell carcinoma (OSCC) patients. Additionally, higher expression of HOTAIR lncRNA in salivary samples is a strong indicator of a metastatic oral cancer [72]. A potential biomarker for liver cancer is the lncRNA HULC, which is detected in plasma and peripheral blood cells [73, 74]. Similarly, the MALAT1 lncRNA is highly expressed in lung cancer tissues and in whole blood of metastatic lung cancer patients, but MALAT1 is not suitable as an independent biomarker to diagnose lung cancer, since it is also detectable in multiple cancers. However, MALAT1 can be used as a com-

plementary biomarker alongside another tissue-specific biomarker [75]. Lastly, the lncRNA SAMMSON was detected in more than 90% of human melanomas [76], and HERV lncRNAs have also been implicated in certain cancers [77], highlighting the potential for lncRNAs as biomarkers of various cancers.

7.3.1 L  ncRNAs Regulated by RAS Signaling RAS signaling regulates cellular growth and proliferation and is often perturbed in many different cancers. During epigenetic reprogramming of mouse fibroblasts to a dedifferentiated, stem cell-­ like state, RAS signaling genes are significantly upregulated in individual cells and are coordinately regulated with lncRNAs, representing the first example of RAS signaling-regulated lncRNAs [78]. In human lung airway epithelial cells, oncogenic RAS signaling regulates the expression of TE-derived noncoding RNAs, including lncRNAs. Importantly, many of these TE-derived noncoding RNAs are released in extracellular vesicles, providing proof-of-­concept that TE-derived lncRNAs and other noncoding RNAs can serve as potential biomarkers of oncogenic RAS signaling [79]. Given the importance of oncogenic RAS signaling in pancreatic, lung, colorectal, and other cancers [80], identifying RAS-regulated lncRNAs will likely yield promising biomarker candidates for cancers driven by mutations in the RAS signaling pathway.

R. E. Reggiardo et al.

126

7.4

LncRNAs in Extracellular Vesicles

LncRNAs are found in various body fluids, and the detection of circulating lncRNAs is a promising, non-invasive method for liquid biopsies, where diseases can be diagnosed using blood or other biofluids. Circulating lncRNAs would potentially be useful for: (i) distinguishing tumor patients from healthy individuals with high sensitivity and specificity, (ii) predicting the prognosis of tumor patients, and (iii) predicting tumor metastasis and recurrence after treatment. LncRNAs found in the circulation are packaged in extracellular vesicles (EVs), a heterogeneous group of cell-derived membranous structures released by cells [81]. The secretion of EVs was initially described as a means of selective elimination of cellular components such as proteins, lipids and RNA from cells [82]. However, they are now thought to play a role in intercellular communication [40, 83].

7.4.1 Types of Extracellular Vesicles Based on their biogenesis, EVs are broadly divided into different sized groups, including exosomes and microvesicles. Microvesicles range in size from 200 to 1000  nm in diameter and are generated by the outward budding and fission of the plasma membrane [84]. Microvesicles have a role in cell-cell communication in various cell types, including cancer cells [85]. Exosomes are EVs that are 30–150  nm in diameter, and being among the smallest EVs, they can pass through small interstitial spaces. Exosomes are generated by the exocytosis pathway and are eventually released in bodily fluids like blood and saliva and can be internalized by neighboring or recipient cells [86]. They are known to harbor various types of molecules such as proteins, lipids, DNA, mRNAs, and noncoding RNAs, facilitating the transfer of genetic and epigenetic information between distant cells and potentially allowing epigenetic synchronization of cells from different tissues [87].

Each cell type tunes EV biogenesis depending on its physiological state and releases EVs with different compositions [83]. Most of the published reports of EVs have focused on their potential role in cellular communication rather than their origins, and it is still unclear which EV subtypes are responsible for which effects, given their overlapping size ranges, similar morphologies, and variable compositions. Exosomes are frequently associated with cancer [88], and many studies have shown that exosomes decrease the immune surveillance of tumors, resulting in their growth, progression, and dissemination [89–94]. Several mechanisms have been shown to inhibit immunity, including increases in suppressive immune cells, decreases in the proliferation and cytotoxicity of NK cells and T cells, and decreases in the number and function of antigen presenting cells [89, 95]. Exosomes also have a role in cancer tissues via angiogenesis. It was reported that exosomes secreted by glioblastoma cells are taken up by vascular endothelial cells and promote angiogenesis [96]. It was also shown that cell cycle-related mRNAs in exosomes from colon cancer cells promote proliferation of vascular endothelial cells [97]. Moreover, noncoding RNAs such as miRNAs are known to be carried by exosomes to promote angiogenesis, cell migration, cell invasion, and formation of a metastatic niche in different types of cancers [98–100]. The broad and increasing interest in exosomes has opened up the possibility of leveraging them as biomarkers for various pathological states, such as inflammation, cancer progression and metastasis. Biomolecules such as proteins, DNA, and RNA are potentially optimal candidates as biomarkers and can be isolated from various biological samples. In the last few years, research efforts have focused on the detection of such biomarkers in bodily fluids. Among these exosomal cargos, lncRNAs are highly promising biomarker candidates. As opposed to circulating DNA, circulating RNA biomarkers provide dynamic insights into cell regulation and disease states. Various isolation and sequencing technologies are being developed to enrich for disease-­

7  LncRNA Biomarkers of Inflammation and Cancer

associated exosomes to define their lncRNA cargos.

7.4.2 Exosomal lncRNAs LncRNAs are present inside exosomes, and a number of cancer-specific exosomal lncRNAs have been isolated from blood serum. MALAT1 is a lncRNA that is abundant in the serum of non-­ small cell lung cancer (NSCLC) patients. Additionally, in  vitro studies demonstrated that high levels of MALAT1 promote cancer cell migration and cell proliferation [101]. Zinc finger antisense 1 (ZFAS1) is a circulating exosomal lncRNA that is upregulated in the blood serum of gastric cancer [102]. Moreover, HOTAIR is one of the most frequently reported lncRNAs involved in cancer development, and high levels of exosomal HOTAIR also correlate with the clinical stage of laryngeal squamous cell carcinoma (LSCC) [103]. However, potential functions of exosomal lncRNAs are not well-characterized. One study showed that lncRNA HOTAIR promotes exosome secretion in hepatocellular carcinoma cells by inducing multivesicular body transport to the plasma membrane, regulating and controlling the docking process of exosomes and also facilitating the final step of fusion and co-localization [104]. Recent studies have shown that some exosomal lncRNAs are released by tumor cells and taken up by other neighboring cells, inducing physiological changes such as angiogenesis. For example, the lncRNA POU class 3 homeobox 3 (POU3F3) is overexpressed in glioma tissue compared to the adjacent normal tissue. Exosomes from glioma cell lines were isolated and co-cultured with human brain microvascular endothelial cells. These exosomes were internalized and led to an increase in cell proliferation, migration, tube formation, and angiogenesis in vivo. Importantly, lncRNAs exported in exosomes are protected from degradation and can reflect specific disease states, making them promising candidates for biomarker discovery. Moreover, the relative abundance of exosomes containing

127

disease-specific lncRNAs enables their robust isolation and the detection of diagnostic signals. Looking forward, biomarker platforms will likely utilize circulating lncRNAs to enable highly sensitive diagnostics [105].

7.5

LncRNAs as Biomarkers

Designing and executing biomarker discovery presents a set of challenges that should carefully be accounted for. In this section, we review current state of the art approaches to RNA liquid biopsies and discuss the technologies that may enable the future success of lncRNA biomarkers.

7.5.1 Features of a Successful Biomarker A biomarker is defined by the U.S.  National Institutes of Health (NIH) as “a defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes or responses to an exposure or intervention” [106]. We focus on the biological potential of lncRNAs to serve as the “defined characteristic” that can indicate progress and/or prognosis of inflammation and cancer. In this section, we will briefly lay out the fundamental classes of biomarkers, what can make them successful, how they can be used by the clinical community, and what technologies can enable their widespread use.

7.5.1.1 Classes of Biomarkers Biomarkers have multiple uses and can be directly associated with patient diagnoses and patient outcomes. The following is a brief overview of the classes defined by the NIH Biomarkers, EndpointS, and other Tools (BEST) resource [106]. Diagnostic “Used to detect or confirm presence of a disease or condition of interest or to identify individuals with a subtype of the disease.” [106]

R. E. Reggiardo et al.

128

These markers help clinicians assign diagnoses to patients in the clinic [107]. For example, the PCA3 lncRNA biomarker is considered diagnostic due to its ability to classify patients into healthy or prostate cancer subgroups and essentially predict the results of a prostate tissue biopsy [108]. Furthermore, the efforts of Yu et al. and Yan et al. characterized the diagnostic potential of biomarkers in pancreatic cancer and Alzheimer’s disease, respectively, by identifying RNA markers with significant changes in expression in individuals with and without the diseases [109, 110]. Owing to the fact that most biomedical data sets will not have robust outcome and drug treatment information, it is reasonable to assume that the majority of lncRNA biomarkers characterized in the near future will be of a diagnostic nature. Predictive “Used to identify individuals who are more likely than similar individuals without the biomarker to experience a favorable or unfavorable effect from exposure to a medical product or an environmental agent.” [106]

These biomarkers are derived from patient populations responding to treatment with significant changes in outcome. As such, they can be used both to advise patient intervention and also enrich clinical trial populations for individuals who are likely responders to the treatment regimen [111]. Since PCA3 is assayed without connection to a particular treatment, it is not considered predictive [108]. No current RNA studies have data including treatment and survival/outcome information. Prognostic “Used to identify likelihood of a clinical event, disease recurrence, or progression in patients who have the disease or medical condition of interest.” [106]

Whereas predictive biomarkers are associated with outcome driven by a specific treatment, prognostic biomarkers associate with future clinical events regardless of treatment [112]. Establishing prognostic biomarkers requires detailed information on patient outcome, which may not be available in many studies.

Monitoring “Measured serially for assessing status of a disease or medical condition or for evidence of ­exposure to (or effect of) a medical product or an environmental agent.” [106]

This category can encompass the previously identified biomarker classes, as long as they are measured serially. For example, PCA3 could be used to monitor occurrence of prostate cancer through regularly repeated measurements, though attempts at deploying it for this purpose have seen mixed results [113]. Additionally, Ibarra et  al. captured potential RNA biomarkers for screening response to bone marrow stimulation [114].

7.5.1.2 Assessing Biomarker Performance There are common statistical assessments applied to biomarker studies in order to evaluate their performance in a standardized manner. Most common amongst these are sensitivity, the fraction of individuals with the disease who test positive, and specificity, the fraction of individuals without the disease who test negative [106]. These quantities can be represented as follows:









TPR  True Positive Rate  P Y  1 | D  1  Sensitivity



FPR  False Positive Rate  P Y  1 | D  0  1  Specificity

where Y is the observation of the biomarker and D is the true disease state of the patient. An ideal biomarker would have a sensitivity of 100%, meaning all individuals with the disease are correctly identified, and a specificity of 100%, meaning all control individuals test negative. In reality, there is a range of acceptable values for biomarker diagnostics that have been detailed elsewhere [115]. Sensitivity and specificity are useful for determining the adequacy of a biomarker in reference to the known labels of the study samples, not its performance on a case by case basis in the clinic. In order to assess the actual probability of an individual patient being correctly diagnosed, this would require values that account for the preva-



7  LncRNA Biomarkers of Inflammation and Cancer

129

7.5.2 S  ample Collection and Study Design

7.5.2.1 Sample Collection and Handling Blood and its derivatives, plasma and serum, are commonly used as biospecimens in clinical and biomedical research [118]. The liquid biopsy field benefits from robust practices and resource guidelines established for the various analytes available in blood plasma [119]. In particular, RNA molecules have demonstrable stability in blood plasma that has been collected in EDTA tubes and frozen at −80 C within a few hours of collection and centrifugation [120]. These requirements are different from those necessary to retain DNA stability and should be taken into consideration when conducting plasma collection. While there is literature investigating the stability of small RNAs in plasma [121], less is known about the relative stability of circulating lncRNAs, which have been detected in plasma [122, 123]. While the number of patients sampled influences statistical analysis and hypothesis testing, the amount of plasma sampled affects the likelihood of detecting circulating lncRNAs that may be of lower abundance. Based on the current literature, it may be ideal to collect a few milliliters of plasma from each patient; even the low-input SILVER-seq method required 0.5  mL of input plasma compared to its much lower microliter input from serum [109, 122]. At this volume, successful library preparations are designed to capture nanogram quantities of RNA, generally the minimum input amount needed for most existing library preparation kits [109, 110, 122–124]. Additionally, a potential concern for small plasma sample volumes is sample loss after collection due to handling and filtration. Upon processing, plasma is separated from white blood cells and platelets by a buffy coat [118], and buffy coat contaminants present in the isolated plasma can potentially interfere with downstream exosome isolation and also result in sample volume being lost during filtration.

An initial challenge for identifying lncRNA biomarkers is the proper collecting, processing, and storing of the biospecimen targeted for analysis.

7.5.2.2 Study Design and Planning Biomarker studies require clinical collaboration during study design, whether via existing clinical

lence of the disease in the population. For example, an extremely rare disease is less likely to be correctly identified by even the most sensitive biomarkers [116]. These quantities, referred to as Predictive Values, are calculated as follows: PPV  Positive PredictiveValue 

NPV  Negative PredictiveValue 

pTPR





pTRP  1  p FPR



1  p 1  FPR  1  p 1  FPR   p 1  TPR 

where p is the prevalence of the condition in the population of interest (TPR and FPR are defined in the previous formula above). PPV is the probability of the individual actually having the relevant condition if the biomarker test is positive, and NPV is the probability that the individual is actually negative, given a negative biomarker test. These measurements assist decision-making in the clinic, where positive diagnoses can mean invasive follow-up procedures [115, 116]. To see how sensitivity, specificity, and PPV behave in practice, we briefly review the performance of the FDA-approved lncRNA PCA3 for the diagnosis of prostate cancer. In a cohort of 721 previously screened men, 122 of whom had prostate cancer, the sensitivity and specificity of PCA3 were anticorrelated over a range of score thresholds, and the authors determined that a sensitivity of 68% would achieve a specificity of 55.7% at a PCA3 score threshold of ≥35. This threshold produced a PPV of 23.9%, capturing 83 out of the 122 cancers in the cohort. The authors indicated that they outperformed other screening technologies for prostate cancer, suggesting its potential utility in identifying more serious cases of prostate cancer [108, 117].

Study Ngo & Mougarrej, et al. Hulstaert, et al. Zhou, et al. Everaert, et al. Yu, et al. Ibarra, et al. Yan, et al.

Year 2018 2019 2019 2019 2019 2020 2020

N (control) 57 (44) 24 (12) 128 (32) 3 (3) 501 (117) 40 (24), 12 (3) 44 (9)

Collection tube EDTA-coated vacutainer tubes NA NA K2-EDTA tube EDTA-coated vacutainer tubes K2-EDTA tube EDTA-coated vacutainer tubes

Table 7.1  Basic study and collection parameters of selected circulating RNA biomarker/characterization studies Exosome isolation N N N Y Y N N

Input volume 0.75–3 mL 200uL 3–7uL Up to 200 uL 1 mL 0.5–1 mL 0.5 mL

130 R. E. Reggiardo et al.

7  LncRNA Biomarkers of Inflammation and Cancer

trials [109, 122] or for sample collection [110, 114, 123–125] (Table 7.1). Some studies seek to characterize performance and overall signal from plasma RNA, while others are searching for putative biomarkers of a particular condition. If the goal is the latter, the onus is on the study to ensure cases have matched controls that do not introduce confounding features, such as mismatched age and sex, which introduce specific transcriptional signals [122, 126]. Study design that does not emphasize well balanced cohorts can introduce bias into the analysis of a biomarker signal [127]. Sample size has important ramifications for the likelihood of identifying relevant lncRNA biomarkers, often referred to as statistical power and reviewed in detail elsewhere [128]. Depending on the hypotheses to be tested, the feasible range of collection, and the desired sensitivity and specificity, statistical power can be estimated a priori via simulation of the proposed statistical model and the distribution of the input data [129, 130]. This challenge is further complicated by the high-dimensionality of modern sequencing data sets: the number of features (genes, transcripts) measured per patient is in the tens of thousands and must be constrained to avoid overly optimistic models [131]. Studies highlighted in Table 7.1 provide insight into this process and reflect the flexibility in sample size for respective goals, which can be influenced by cost and clinical access, as well as study design.

7.5.3 Isolating Exosomes from Blood Plasma Exosome isolation protocols can potentially influence the abundance and diversity of lncRNAs available for detection. Exosome isolation is motivated by their encapsulation of RNA within a lipid bilayer that protects the labile RNA molecules from degradation in biofluids [132, 133]. In this section, we will examine not only the efficacy of different approaches to exosome isolation but also their applicability to biomarker discovery.

7.5.3.1 Differential Ultracentrifugation Ultracentrifugation (UC) has been the primary method to isolate and study exosomes [134]. As

131

the name suggests, UC involves multiple rounds of high-speed spins that span many hours. This procedure is followed by washing and additional spins to purify and concentrate the exosome population [135]. Commonly considered the ‘gold standard’ approach to concentrating and isolating exosomes of high purity, this requires a significant amount of operator time and requires that particular attention be paid to the configuration of the expensive equipment it employs [135]. It may come as little surprise then that this particular platform has not been widely adopted by biomarker studies with clinical aspirations. For example, the SILVER-seq platform avoids any exosome isolation, and studies from Everaert et al. and Yu et al. use size exclusion chromatography alongside density gradient centrifugation and a commercial affinity column approach, respectively [109, 110, 122, 123].

7.5.3.2 Density Gradient Centrifugation Density gradient centrifugation (DG) seeks to achieve high purity and specificity of exosome isolation by improving upon the UC approach. Similar in principle to UC, DG relies on centrifugation at speeds of 100,000  g. However, the approach employs an iodixanol density gradient, first used to isolate RNA viruses, in order to selectively isolate a pure population of exosomes [136, 137]. While the isolated exosome population is more optimal, the equipment and protocols requiring a continuous 18-hour centrifugation may hinder the broad adoption and scalability of this approach in diverse biomedical and clinical settings. 7.5.3.3 Affinity Column An example of the affinity column approach is illustrated by Yu et al., where their exoRNeasy platform simplifies and standardizes EV isolation to make the process more amenable to clinical labs [138]. Implemented in a spin column compatible with benchtop centrifuges, the exoRNeasy protocol can be completed in a few hours. While the relative ease of use may facilitate wider adoption, there may be more size heterogeneity with this type of affinity-based protocol.

R. E. Reggiardo et al.

132

7.5.3.4 Other Exosome Isolation Technologies To enable isolation of EVs based on size, the Exosome Total Isolation Chip (ExoTIC) platform uses a syringe pump and a series of successively smaller filters to isolate EVs within a defined size range, including exosomes [139]. ExoTIC isolated significantly more exosomes than UC within specific size ranges in approximately 3 hours and at the cost of roughly one US dollar for the device itself. The advantages of this platform are the modular, simple components that allow for precise selection of exosomes and its standardized implementation. The ease of use and size specificity of ExoTIC are promising features for future lncRNA biomarker discovery. 7.5.3.5 Assessing Exosome Quality and Identity Exosome isolation methods should be verified to ensure the EV origin of any RNA biomarker signal and to facilitate reproducibility. Common approaches for assessing EVs are nanoparticle tracking analysis (NTA) and transmission electron microscopy (TEM) [135, 136, 138, 139]. NTA is an appealing technology for microvesicle and exosome research due to its sensitivity down to a 50  nm particle size, a significant improvement over flow cytometry approaches that only reach ~300 nm [140]. TEM also boasts nanometer resolution and is widely used to image exosomes [141]. Both TEM and NTA require specific instruments and software for analysis and are commonly used as orthogonal measurements to validate vesicle identity [142, 143]. Additional information on the adoption and deployment of various exosome isolation technologies has been categorized and surveyed in detail [144].

7.6

Transcriptomic Analysis of lncRNA Biomarkers

PCA3 was originally identified using differential display analysis and northern blotting [145, 146]. Since then, RNA biomarker discovery has undergone a significant shift, first through the application of microarrays that allowed highly

multiplexed assessment of gene expression, and now more dramatically through ultra-high-­ throughput RNA-seq approaches [147–150]. While microarrays remain valid tools for translating gene expression into clinical predictions, RNA-seq is not limited by prior knowledge of sequence identity, can better identify isoforms and splice junctions, and is more sensitive to RNA abundance [147, 151]. These technological advantages are essential to the assessment of lncRNA biomarkers, particularly those previously unidentified, poorly characterized, and/or modestly expressed transcripts that are preferentially exported to biofluids via exosomes [105, 149, 150]. Early attempts at applying RNA-seq to bottom-up transcriptomic RNA biomarker discovery focused on small RNAs, due in part to their increased stability relative to longer RNA molecules [152, 153]. These efforts progressed to combined microarray and RNAseq analysis of the RNA transcriptome, with RNA-seq as the consensus choice to develop RNA biomarkers from blood plasma and other biofluids [109, 110, 114, 122–125, 154]. We highlight RNA-seq technologies from two major platforms: Illumina sequencing-by-synthesis and Oxford Nanopore Technologies (ONT) single-molecule nanopore sequencing. This section will review their individual strengths for detecting and characterizing circulating lncRNA biomarkers.

7.6.1 Illumina The Illumina platform is the current gold standard approach to detecting, quantifying, and characterizing lncRNA molecules and has enabled the detection of the vast majority of known lncRNA genes [46, 155]. Advances in library preparation protocols have enabled the sequencing of increasingly lower amounts of RNA, allowing for the characterization of lncRNAs in single cells [78]. In particular, picogram scale library preparation kits have been widely adopted in circulating RNA biomarker studies, reflecting their utility for RNA quantification of circulating RNA biomarkers

7  LncRNA Biomarkers of Inflammation and Cancer Table 7.2 RNA sequencing parameters of selected studies

Study Ngo & Mougarrej, et al.

Hulstaert, et al. Zhou, et al. Everaert, et al.

Yu, et al.

Ibarra, et al. Yan, et al.

Read length Library prep kit/ (millions of reads) approach 2 × 75 SMARTer Stranded Total (≥10) RNAseq Pico Input Mammalian Kit 2 × 75 TruSeq RNA Exome Library (≥11) Prep Kit SILVER-seq NA (14) 2 × 75 SMARTer Stranded Total (15.3) RNA-Seq Kit v2 – Pico Input Mammalian 2 × 150 SMARTer Stranded Total (20) RNA-Seq Kit–Pico Input Mammalian NA Random (5–30) hexamer priming SILVER-seq 2 × 75 (19.2)

Approx. number of genes detected (subset) NA

5440

41,000 1598

15,000

7–13,000 (protein coding) 1514 (brain)

133

7.6.2 Nanopore Nanopore sequencing technology has continued to improve in sequencing accuracy, throughput, and portability [158–160]. It is thus a promising platform for liquid biopsy work for multiple reasons: (1) the length of the reads may improve confident mapping of repetitive TE-derived lncRNA transcripts, (2) the portability of the platform (particularly the ONT MinION) could enable point-of-care biomarker characterization across the globe, and (3) capturing individual molecules would allow for identification of isoforms, fragments, and full-length RNA sequences specific to exosomes and biofluids. In addition to cDNA, the ONT platform can also sequence native RNA molecules, enabling detection of modifications to the individual bases, which may serve as clinically relevant markers [161, 162]. A potential challenge for the adoption of Nanopore technology for circulating lncRNA quantification is the relatively large input RNA requirements, which may be harder to obtain from biofluid samples (Table  7.3). Other long-read sequencing platforms such as PacBio may also be useful for RNA liquid biopsies.

Approximate number of reads reported when available in the publication. Any biological subset specified in the paper is identified in parentheses

7.6.3 RNA Selection

[109, 110, 114, 123–125, 154] (Table 7.2). The performance of these protocols appears to depend strongly on user implementation, with transcriptome coverage varying widely across studies, even though most are employing the same or similar kits (Table 7.2). It is also worth noting that longer Illumina reads, like the paired-end 150 base pair protocol used by Yan et al., could substantially improve mapping of ambiguous sequences and splice junctions [156, 157]. While Illumina-based sequencing approaches are currently the dominant technology in the market for lncRNA biomarker discovery, there remains a clear need to assess the impact of read length, sequencing depth, and library preparation on lncRNA detection in biofluids and exosomes.

RNA-seq approaches can capture subsets of RNAs that are selected for during library preparation: polyadenylated RNAs, RNAs that contain a specific sequence, or total RNA [163–165]. Avoidance and removal of abundant ribosomal RNAs (rRNAs) is crucially important for generating informative libraries, even more so in circulating RNA preparations where rRNA seems to be a common contaminant [123]. Circulating RNA studies have focused on total RNA ­preparations (Table  7.2) that utilize proprietary technology to remove rRNA-derived cDNA before sequencing [109, 110, 114, 122–125]. The main advantage of this approach is that it accounts for any degradation or loss of quality of the RNA molecules exported in biofluids [166]. As the field isolates more specific populations of EVs that harbor intact RNA species, this advantage

R. E. Reggiardo et al.

134

Table 7.3  Performance metrics of RNA sequencing platforms for circulating RNA quantification Platform Illumina (NextSeq) Nanopore: cDNA Nanopore: nvRNA

Length 75–150 bp (paired-end) Median = 780 bpa Max = 9969a Median = 771 bpa Max = 21,000 bpa

Median per-base accuracy 99.9%

Minimal RNA input 250 pg

Throughput (millions of reads) 130–400

85a–95%b

1 ng

1.2 average per MinIONb

86%a

500 ng

0.05–.831 per MinIONa

Values with no citation provided were drawn from the product description documents available on the manufacturer website a Workman & Tang, et al. b Volden, et al.

may not outweigh the performance losses suffered in the absence of polyA selection, especially since accurate quantification is important for biomarker discovery [167]. The alternative to these approaches would be to target a specific sequence, or set of sequences, from candidate lncRNA biomarkers.

Picard implementations of PCR duplicate removal after alignment (Table 7.4) [172]. This approach may be introducing bias into the quantifications, and could limit the detection of repetitive lncRNA sequences that may align at the same coordinates [156, 170, 171]. In order to avoid alignment-­ based PCR-duplicate removal, the SILVER-seq platform used a unique molecular identifier (UMI) for each read, 7.6.4 RNA-seq Analysis which can be used to filter out duplicated reads [122, 171]. It is still unclear how much of a negRNA-seq analytical rigor applies to all RNA-seq-­ ative or positive impact PCR duplicate removal based experiments and has a large impact on the will have on the varying library preparations results and conclusions generated from the data. used for circulating RNA-seq studies, but it is The broad principles and best practices for RNA-­ likely that an approach such as UMI integration seq analysis are reviewed elsewhere [168, 169]. will offer some advantages. This section will focus on a few of the analytical steps critical to analysis of lncRNA biomarkers 7.6.4.2 Alignment and Quantification and high dimensional patient data and review the RNA-seq alignment is the most consistent step current analytical paradigms deployed to study identified in current studies, with STAR being the circulating RNA biomarkers (Fig. 7.3). most widely used (Table 7.4). STAR is among the most popular aligners for RNA-seq, and its splice-aware algorithm should capture the poten7.6.4.1 Quality Control When performing RNA-seq on biofluids for tially fragmented and/or spliced structure of cirlncRNA biomarker discovery, precautions culating lncRNAs well [173]. Quantification of should be taken to ensure that the reads used in those alignments, on the other hand, is fragthe analysis are of the highest quality [109, 110, mented across multiple approaches (Table  7.4). 114, 122–125]. The overall reduced complexity With the potential for DNA contamination in of the libraries, a byproduct of the more limited ­circulating RNA-seq libraries, it seems prudent breadth of circulating RNAs in biofluids, and to explore transcript pseudo-aligners like Salmon the number of cycles used to amplify low-input and Kallisto that map reads directly to annotated amounts, have the potential to lead to increased transcripts [174–176]. Limiting the alignment representation of contaminants and PCR- scope to only exonic regions of annotated tranduplicated reads [170, 171]. Almost all of the scripts might help clarify which reads are likely current literature applies either Samtools or derived from RNA.

7  LncRNA Biomarkers of Inflammation and Cancer

135

Fig. 7.3  RNA-seq workflow for circulating lncRNA biomarker discovery from plasma. (Created with BioRender.com)

Multi-mapping reads Aligning TE-derived lncRNA reads introduces multi-mapping reads due to the shared sequence content of TE insertions [156]. This ambiguity is largely ignored by circulating RNA-seq studies that filter to only uniquely mapped reads, but the evidence provided by Yan et al. of the value of TE expression in circulating RNA biomarker discovery encourages further exploration of their mapping [109]. Statistical models have been introduced to reassign ambiguous reads in order to generate accurate reassignment of ambiguous alignments. Most general Expectation Maximization (EM) models have been used extensively in approaches like Kallisto and Salmon and perform well, due to the precision of the initial expectations assigned to the latent variables: the correct generating transcripts [174, 175]. In multi-mapping reads, EM models are currently employed for short read-based locus-­ level quantification [177–180]. These approaches have made significant progress towards correcting ambiguous alignments, each posting simulated accuracies of 90+% for homologous LINE, Alu and HERV transcripts. This accuracy rapidly deteriorates in younger, polymorphic insertions, resulting in accuracies between 50% and 70%. It is likely that long-read RNA-seq technology will help alleviate ambiguous mapping of repetitive lncRNA biomarker reads [181].

7.6.4.3 Count Normalization Normalization of RNA-seq counts helps reduce non-biological variability that makes it difficult to accurately compare abundances between samples [182]. In biomarker discovery, there

will often be many samples, not all of which are sequenced together or under the same conditions, to compare and evaluate in aggregate. Many of the studies featured here transform their raw read data into transcripts-per-million (TPM), a unitless, normalized value scaled to transcript length and used in modern transcript quantification approaches [174, 175, 183]. While TPM is an improvement on raw counts, it cannot be used to accurately compare between samples unless the total RNA abundance and the distribution of those abundances are similar [183]. It is worthwhile to explore the normalization approaches taken in large dataset pipelines and those recommended through simulations such as DESeq2 and trimmed means of M (TMM) (edgeR) [184, 185]. Another approach to accurate normalization between samples is addition of externally standardized RNA molecules as used by Everaert et al. and reviewed in more detail elsewhere [123, 185]. These molecules should have consistent abundances across all samples and can be used to derive standard curves that, with the help of the correct models and assumptions, aid in the quantification of all transcripts [186].

7.6.4.4 Statistical Modeling The final and often most impactful stage of circulating RNA biomarker discovery is identifying the subset of genes that are able to classify samples as control or disease most accurately [110, 122, 125]. There are a few common approaches to sample classification implemented in these studies that will be reviewed and contextualized here (Table 7.4).

Custom deduplication (UMI)

Yan, et al.

Yu, et al. Ibarra, et al. STAR

HISAT2 STAR

STAR

STAR STAR

FastQC, Picard Unique alignment

Clumpify, cutadapt, subsampling FastQC Rmdup

Aligner STAR

Quality control Unique alignment, Picard

Everaert, et al.

Study Ngo & Mougarrej, et al. Hulstaert, et al. Zhou, et al.

featureCounts

NA RSEM

Kallisto

HTSeq-count HTSeq-count

Quantification HTSeq-count

TPM RSEM-estimated counts, TPM TPM

Exogenous standard

TPM, DESeq2 log2(TPM + 1)

Normalization Multiple of the Median

Table 7.4  Bioinformatic tools and approaches for the highlighted studies at sequential phases of data analysis

LASSO, OOB Negative Matrix Factori-zation NA

Feature selection Recursive Feature Elimination NA Manual curation/ filtering NA

NA

SVM NA

NA Random Forest, SVM NA

Classifier Random Forest

136 R. E. Reggiardo et al.

7  LncRNA Biomarkers of Inflammation and Cancer

Feature selection and dimensionality reduction The primary concern with modeling or predicting from RNA-seq datasets is the commonly referred to ‘curse of high dimensionality’: with so many features to describe each patient, models are likely to be ‘over-fit’ and generalize very poorly to other data sets, thus significantly hampering the applicability of the biomarkers [187]. Selecting highly informative genes from the many thousands by RNA-seq can be accomplished with Recursive Feature Elimination (RFE) or penalized regression Least Absolute Shrinkage and Selection Operator (LASSO), among other approaches [188, 189]. These tools each handle features differently, with colinear and uninformative features, and will ultimately attempt to produce the simplest model with the lowest number of features possible. Both Ngo et al. and Yu et al. used RFE and LASSO, respectively, to filter down their feature set to 7-8 genes that go on to inform promising, simple classifiers that have the potential to serve as biomarker panels [110, 125]. Dimensionality reduction, like feature selection, attempts to minimize the number of features describing the samples, while retaining the structure and variability present in the full sparse dataset, usually through unsupervised clustering. These reduced dimensions, sometimes referred to as ‘components’, contain groups of features (genes) that best describe the structure of the dataset. The matrix factorization family of algorithms include Principal Component Analysis (PCA) and Non-negative Matrix Factorization (NMF) that are the primarily tools for interpretable dimensionality reduction and are reviewed in detail elsewhere [190]. Ibarra et al. applied NMF to identify subsets of genes that had similar expression patterns across samples, estimating the fraction of circulating RNA derived from different hematopoietic cell types [114]. Zhou et al. applied PCA to cluster their samples in two dimensional space, and the structure of the components was such that the samples separated cleanly into clusters of ‘Cancer’ and ‘Normal’ groups, with the caveat that very little of the variance in the original

137

dimensions was retained in this components reduction [122]. Classification and model assessment Once proper feature selection has been performed, classification procedures can be implemented to predict label (in this case, disease or healthy) from the selected circulating RNA [122, 125, 191]. The models available for classification are numerous, and most common in biomarker modeling are Logistic Regression (logit), Random Forests (RF), Support Vector Machines (SVM), K Nearest Neighbors (KNN), and other tree-based methods [192]. Whichever model is chosen, correct validation of the prediction result is crucial to its interpretation and validity. One approach to validating predictive accuracy is the split-sample methodology that separates the data into distinct training and testing subsets, usually containing 75% and 25% of the samples, respectively [192]. Yu et al. used the split-sample to first train an SVM to classify pancreatic cancer patients and healthy samples on a 188-member training split. This model was then validated on a separate 135-member internal validation split that the model had never seen before, eventually achieving a sensitivity of 93.49% and a specificity of 85.07% [110]. Part of the reason this approach was successful for Yu et  al. was their robust sample size; with fewer samples, the size of the test split falls below a reasonable size. V-fold Cross Validation (CV) splits the entire dataset into V balanced partitions and uses a resampling approach to iteratively train on all but one of these partitions, which serves as the test set. The most extreme form of CV, Leave one out CV (LOOCV), which isolates each sample as a test set over many iterations, has been shown to perform well on small, high-dimensional data sets [191].

7.7

Conclusions and Future Perspectives

Companies are currently developing circulating RNAs as potential biomarkers of disease, including Molecular Stethoscope [114], Biogazelle

138

[123], and Genemo [122]. Isolating, sequencing, analyzing, and predicting based on circulating lncRNAs would enable lncRNA biomarkers to be used for screening at-risk patients for inflammation, cancer, and other diseases. Notably, clinical screening will have different considerations when compared to lncRNA biomarker discovery: a simple assay, like quantitative polymerase chain reaction (qPCR), that can quickly and inexpensively measure lncRNA biomarkers, may potentially be more appropriate than high-throughput RNA-seq in the short term. Once a simple and accurate set of lncRNA biomarkers has been identified, this lncRNA signature could be assayed without sequencing the entire circulating transcriptome. However, a pressing issue currently facing the biomarker field is the lack of robust, randomized clinical trials validating discovered lncRNA biomarkers and signatures. The need for lncRNA biomarker validation has been evident for several years now and will require well-designed, clinically actionable biomarker studies to become standard practice in the future [193–195]. Moving forward, the continued discovery and validation of lncRNA biomarkers will provide new opportunities for the early detection of cancer and other diseases with strong links to inflammation.

References 1. Rinn JL, Chang HY (2020) Long noncoding RNAs: molecular modalities to organismal functions. Annu Rev Biochem 89:283–308. https://doi.org/10.1146/ annurev-­biochem-­062917-­012708 2. Kopp F, Mendell JT (2018) Functional classification and experimental dissection of long noncoding RNAs. Cell 172:393–407. https://doi.org/10.1016/j. cell.2018.01.011 3. Lee JT (2012) Epigenetic regulation by long noncoding RNAs. Science 338:1435–1439. https://doi. org/10.1126/science.1231776 4. Bartel DP (2004) MicroRNAs. Cell 116:281–297. https://doi.org/10.1016/S0092-­8674(04)00045-­5 5. Derrien T, Johnson R, Bussotti G et  al (2012) The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Res 22:1775–1789. https:// doi.org/10.1101/gr.132159.111 6. Harrow J, Frankish A, Gonzalez JM et  al (2012) GENCODE: the reference human genome annotation

R. E. Reggiardo et al. for The ENCODE Project. Genome Res 22:1760– 1774. https://doi.org/10.1101/gr.135350.111 7. Alam T, Medvedeva YA, Jia H et al (2014) Promoter analysis reveals globally differential regulation of human long non-coding RNA and protein-­ coding genes. PLoS One 9:e109443. https://doi. org/10.1371/journal.pone.0109443 8. Wapinski O, Chang HY (2011) Long noncoding RNAs and human disease. Trends Cell Biol 21:354– 361. https://doi.org/10.1016/j.tcb.2011.04.001 9. Brosnan CA, Voinnet O (2009) The long and the short of noncoding RNAs. Curr Opin Cell Biol 21:416– 425. https://doi.org/10.1016/j.ceb.2009.04.001 10. Zhao J, Sun BK, Erwin JA et  al (2008) Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome. Science 322:750–756. https://doi. org/10.1126/science.1163045 11. Bolha L, Ravnik-Glavač M, Glavač D (2017) Long noncoding RNAs as biomarkers in cancer. Dis Markers 2017:1–14. https://doi. org/10.1155/2017/7243968 12. Hu W, Alvarez-Dominguez JR, Lodish HF (2012) Regulation of mammalian cell differentiation by long non-coding RNAs. EMBO Rep 13:971–983. https://doi.org/10.1038/embor.2012.145 13. Flynn RA, Chang HY (2014) Long noncoding RNAs in cell-fate programming and reprogramming. Cell Stem Cell 14:752–761. https://doi.org/10.1016/j. stem.2014.05.014 14. St Laurent G, Wahlestedt C, Kapranov P (2015) The landscape of long noncoding RNA classification. Trends Genet 31:239–251. https://doi.org/10.1016/j. tig.2015.03.007 15. International Human Genome Sequencing Consortium (2001) Initial sequencing and analysis of the human genome. Nature 409:860–921. https:// doi.org/10.1038/35057062 16. Kapusta A, Kronenberg Z, Lynch VJ et  al (2013) Transposable elements are major contributors to the origin, diversification, and regulation of vertebrate long noncoding RNAs. PLoS Genet 9:e1003470. https://doi.org/10.1371/journal.pgen.1003470 17. Johnson R, Guigo R (2014) The RIDL hypothesis: transposable elements as functional domains of long noncoding RNAs. RNA 20:959–976. https://doi. org/10.1261/rna.044560.114 18. Smit AF (1999) Interspersed repeats and other mementos of transposable elements in mammalian genomes. Curr Opin Genet Dev 9:657–663. https:// doi.org/10.1016/S0959-­437X(99)00031-­3 19. Luning Prak ET, Kazazian HH (2000) Mobile elements and the human genome. Nat Rev Genet 1:134–144. https://doi.org/10.1038/35038572 20. Wei W, Gilbert N, Ooi SL et al (2001) Human L1 retrotransposition: cisPreference versus trans complementation. Mol Cell Biol 21:1429–1439. https://doi. org/10.1128/MCB.21.4.1429-­1439.2001 21. Jordan IK, Rogozin IB, Glazko GV, Koonin EV (2003) Origin of a substantial fraction of human regulatory sequences from transposable elements.

7  LncRNA Biomarkers of Inflammation and Cancer Trends Genet 19:68–72. https://doi.org/10.1016/ S0168-­9525(02)00006-­9 22. Wang T, Zeng J, Lowe CB et  al (2007) Species-­ specific endogenous retroviruses shape the transcriptional network of the human tumor suppressor protein p53. Proc Natl Acad Sci 104:18613–18618. https://doi.org/10.1073/pnas.0703637104 23. Bourque G, Leong B, Vega VB et  al (2008) Evolution of the mammalian transcription factor binding repertoire via transposable elements. Genome Res 18:1752–1762. https://doi.org/10.1101/ gr.080663.108 24. Bejerano G, Lowe CB, Ahituv N et al (2006) A distal enhancer and an ultraconserved exon are derived from a novel retroposon. Nature 441:87–90. https:// doi.org/10.1038/nature04696 25. Lynch VJ, May G, Wagner GP (2011) Regulatory evolution through divergence of a phosphoswitch in the transcription factor CEBPB. Nature 480:383– 386. https://doi.org/10.1038/nature10595 26. Schmidt D, Schwalie PC, Wilson MD et  al (2012) Waves of retrotransposon expansion remodel genome organization and CTCF binding in multiple mammalian lineages. Cell 148:335–348. https://doi. org/10.1016/j.cell.2011.11.058 27. Wang J, Lunyak VV, Jordan IK (2012) Genome-­ wide prediction and analysis of human chromatin boundary elements. Nucleic Acids Res 40:511–529. https://doi.org/10.1093/nar/gkr750 28. Feschotte C (2008) Transposable elements and the evolution of regulatory networks. Nat Rev Genet 9:397–405. https://doi.org/10.1038/nrg2337 29. Nekrutenko A, Li W-H (2001) Transposable elements are found in a large number of human protein-­ coding genes. Trends Genet 17:619–621. https://doi. org/10.1016/S0168-­9525(01)02445-­3 30. van de Lagemaat LN, Landry J-R, Mager DL, Medstrand P (2003) Transposable elements in mammals promote regulatory variation and diversification of genes with specialized functions. Trends Genet 19:530–536. https://doi.org/10.1016/j. tig.2003.08.004 31. Zhang XH-F, Chasin LA (2006) Comparison of multiple vertebrate genomes reveals the birth and evolution of human exons. Proc Natl Acad Sci 103:13427–13432. https://doi.org/10.1073/ pnas.0603042103 32. Kelley D, Rinn J (2012) Transposable elements reveal a stem cell-specific class of long noncoding RNAs. Genome Biol 13:R107. https://doi. org/10.1186/gb-­2012-­13-­11-­r107 33. Gong C, Maquat LE (2011) lncRNAs transactivate STAU1-mediated mRNA decay by duplexing with 3′ UTRs via Alu elements. Nature 470:284–288. https://doi.org/10.1038/nature09701 34. Cartault F, Munier P, Benko E et al (2012) Mutation in a primate-conserved retrotransposon reveals a noncoding RNA as a mediator of infantile encephalopathy. Proc Natl Acad Sci 109:4980–4985. https:// doi.org/10.1073/pnas.1111596109

139 35. Santoni FA, Guerra J, Luban J (2012) HERV-H RNA is abundant in human embryonic stem cells and a precise marker for pluripotency. Retrovirology 9:111. https://doi.org/10.1186/1742-­4690-­9-­111 36. Lu X, Sachs F, Ramsay L et  al (2014) The retrovirus HERVH is a long noncoding RNA required for human embryonic stem cell identity. Nat Struct Mol Biol 21:423–425. https://doi.org/10.1038/ nsmb.2799 37. Kretz M, Webster DE, Flockhart RJ et  al (2012) Suppression of progenitor differentiation requires the long noncoding RNA ANCR. Genes Dev 26:338– 343. https://doi.org/10.1101/gad.182121.111 38. Ramos AD, Diaz A, Nellore A et  al (2013) Integration of genome-wide approaches identifies lncRNAs of adult neural stem cells and their progeny in vivo. Cell Stem Cell 12:616–628. https://doi. org/10.1016/j.stem.2013.03.003 39. Hu G, Tang Q, Sharma S et  al (2013) Expression and regulation of intergenic long noncoding RNAs during T cell development and differentiation. Nat Immunol 14:1190–1198. https://doi.org/10.1038/ ni.2712 40. Klattenhoff CA, Scheuermann JC, Surface LE et  al (2013) Braveheart, a long noncoding RNA required for cardiovascular lineage commitment. Cell 152:570–583. https://doi.org/10.1016/j. cell.2013.01.003 41. Shore AN, Kabotyanski EB, Roarty K et al (2012) Pregnancy-induced noncoding RNA (PINC) associates with polycomb repressive complex 2 and regulates mammary epithelial differentiation. PLoS Genet 8:e1002840. https://doi.org/10.1371/journal. pgen.1002840 42. Ginger MR, Shore AN, Contreras A et al (2006) A noncoding RNA is a potential marker of cell fate during mammary gland development. Proc Natl Acad Sci 103:5781–5786. https://doi.org/10.1073/ pnas.0600745103 43. Rinn JL, Kertesz M, Wang JK et al (2007) Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 129:1311–1323. https://doi.org/10.1016/j. cell.2007.05.022 44. Khalil AM, Guttman M, Huarte M et al (2009) Many human large intergenic noncoding RNAs associate with chromatin-modifying complexes and affect gene expression. Proc Natl Acad Sci 106:11667– 11672. https://doi.org/10.1073/pnas.0904715106 45. Djebali S, Davis CA, Merkel A et  al (2012) Landscape of transcription in human cells. Nature 489:101–108. https://doi.org/10.1038/nature11233 46. Cabili MN, Trapnell C, Goff L et  al (2011) Integrative annotation of human large intergenic noncoding RNAs reveals global properties and specific subclasses. Genes Dev 25:1915–1927. https:// doi.org/10.1101/gad.17446611 47. Atanasovska B, Rensen SS, van der Sijde MR et al (2017) A liver-specific long noncoding RNA with a role in cell viability is elevated in human nonal-

140 coholic steatohepatitis. Hepatology 66:794–808. https://doi.org/10.1002/hep.29034 48. Mills JD, Kavanagh T, Kim WS et  al (2015) High expression of long intervening non-coding RNA OLMALINC in the human cortical white matter is associated with regulation of oligodendrocyte maturation. Mol Brain 8:2. https://doi.org/10.1186/ s13041-­014-­0091-­9 49. Grote P, Wittler L, Hendrix D et al (2013) The tissue-­ specific lncRNA Fendrr is an essential regulator of heart and body wall development in the mouse. Dev Cell 24:206–214. https://doi.org/10.1016/j. devcel.2012.12.012 50. Cesana M, Cacchiarelli D, Legnini I et  al (2011) A long noncoding RNA controls muscle differentiation by functioning as a competing endogenous RNA.  Cell 147:358–369. https://doi.org/10.1016/j. cell.2011.09.028 51. Rapicavoli NA, Poth EM, Zhu H, Blackshaw S (2011) The long noncoding RNA Six3OS acts in trans to regulate retinal development by modulating Six3 activity. Neural Dev 6:32. https://doi. org/10.1186/1749-­8104-­6-­32 52. Amin V, Harris RA, Onuchic V et  al (2015) Epigenomic footprints across 111 reference epigenomes reveal tissue-specific epigenetic regulation of lincRNAs. Nat Commun 6:6370. https://doi. org/10.1038/ncomms7370 53. Taniguchi K, Karin M (2018) NF-kappa B, inflammation, immunity and cancer: coming of age. Nat Rev Immunol 18:309–324. https://doi.org/10.1038/ nri.2017.142 54. Simion V, Zhou H, Pierce JB et al (2020) LncRNA VINAS regulates atherosclerosis by modulating NF-κB and MAPK signaling. JCI Insight 5. https:// doi.org/10.1172/jci.insight.140627 55. Rapicavoli NA, Qu K, Zhang J et al (2013) A mammalian pseudogene lncRNA at the interface of inflammation and anti-inflammatory therapeutics. eLife 2:e00762. https://doi.org/10.7554/eLife.00762 56. Carpenter S, Aiello D, Atianand MK et  al (2013) A long noncoding RNA mediates both activation and repression of immune response genes. Science 341:789–792. https://doi.org/10.1126/ science.1240925 57. Liu B, Sun L, Liu Q et  al (2015) A cytoplasmic NF-κB interacting long noncoding RNA blocks IκB phosphorylation and suppresses breast cancer metastasis. Cancer Cell 27:370–381. https://doi. org/10.1016/j.ccell.2015.02.004 58. Ma S, Ming Z, Gong A-Y et al (2017) A long noncoding RNA, lincRNA-Tnfaip3, acts as a coregulator of NF-κB to modulate inflammatory gene transcription in mouse macrophages. FASEB J 31:1215–1225. https://doi.org/10.1096/fj.201601056R 59. Cremer S, Michalik KM, Fischer A et  al (2019) Hematopoietic deficiency of the long noncoding RNA MALAT1 promotes atherosclerosis and plaque inflammation. Circulation 139:1320–1334. https:// doi.org/10.1161/CIRCULATIONAHA.117.029015

R. E. Reggiardo et al. 60. Ali MA, Shaker OG, Khalefa AA et al (2020) Serum long noncoding RNAs FAS-AS1 & PVT1 are novel biomarkers for systemic lupus erythematous. Br J Biomed Sci 77:208–212. https://doi.org/10.1080/09 674845.2020.1765459 61. Li Z, Li X, Jiang C et  al (2018) Long non-coding RNAs in rheumatoid arthritis. Cell Prolif 51. https:// doi.org/10.1111/cpr.12404 62. Iyer MK, Niknafs YS, Malik R et  al (2015) The landscape of long noncoding RNAs in the human transcriptome. Nat Genet 47:199–208. https://doi. org/10.1038/ng.3192 63. Schoenhals M, Kassambara A, Vos JD et al (2009) Embryonic stem cell markers expression in cancers. Biochem Biophys Res Commun 383:157–162. https://doi.org/10.1016/j.bbrc.2009.02.156 64. Ben-Porath I, Thomson MW, Carey VJ et  al (2008) An embryonic stem cell–like gene expression signature in poorly differentiated aggressive human tumors. Nat Genet 40:499–507. https://doi. org/10.1038/ng.127 65. Bouckenheimer J, Assou S, Riquier S et  al (2016) Long non-coding RNAs in human early embryonic development and their potential in ART. Hum Reprod Update 23:19–40. https://doi.org/10.1093/ humupd/dmw035 66. Zhang Y, Shields T, Crenshaw T et  al (1993) Imprinting of human H19: Allele-specific CpG methylation, loss of the active Allele in Wilms tumor, and potential for somatic Allele switching. Am J Human Genet 53:113 67. de Kok JB, Verhaegh GW, Roelofs RW et al (2002) DD3PCA3, a very sensitive and specific marker to detect prostate tumors. Cancer Res 62:2695–2698 68. Luo J-H, Ren B, Keryanov S et  al (2006) Transcriptomic and genomic analysis of human hepatocellular carcinomas and hepatoblastomas. Hepatology 44:1012–1024. https://doi.org/10.1002/ hep.21328 69. Silva A, Bullock M, Calin G (2015) The clinical relevance of long non-coding RNAs in cancer. Cancers 7:2169–2182. https://doi.org/10.3390/ cancers7040884 70. Hessels D, Klein Gunnewiek JMT, van Oort I et  al (2003) DD3PCA3-based molecular urine analysis for the diagnosis of prostate cancer. Eur Urol 44:8–16. https://doi.org/10.1016/ S0302-­2838(03)00201-­X 71. Lee GL, Dobi A, Srivastava S (2011) Diagnostic performance of the PCA3 urine test. Nat Rev Urol 8:123–124. https://doi.org/10.1038/nrurol.2011.10 72. Tang H, Wu Z, Zhang J, Su B (2013) Salivary lncRNA as a potential marker for oral squamous cell carcinoma diagnosis. Mol Med Rep 7:761–766. https://doi.org/10.3892/mmr.2012.1254 73. Panzitt K, Tschernatsch MMO, Guelly C et al (2007) Characterization of HULC, a novel gene with striking up-regulation in hepatocellular carcinoma, as noncoding RNA.  Gastroenterology 132:330–342. https://doi.org/10.1053/j.gastro.2006.08.026

7  LncRNA Biomarkers of Inflammation and Cancer 74. Xie H, Ma H, Zhou D (2013) Plasma HULC as a promising novel biomarker for the detection of hepatocellular carcinoma. Biomed Res Int 2013:1–5. https://doi.org/10.1155/2013/136106 75. Weber DG, Johnen G, Casjens S et  al (2013) Evaluation of long noncoding RNA MALAT1 as a candidate blood-based biomarker for the diagnosis of non-small cell lung cancer. BMC Res Notes 6:518. https://doi.org/10.1186/1756-­0500-­6-­518 76. Leucci E, Vendramin R, Spinazzi M et  al (2016) Melanoma addiction to the long non-coding RNA SAMMSON.  Nature 531:518–522. https://doi. org/10.1038/nature17161 77. Bannert N, Hofmann H, Block A, Hohn O (2018) HERVs new role in cancer: from accused perpetrators to cheerful protectors. Front Microbiol 9:178. https://doi.org/10.3389/fmicb.2018.00178 78. Kim DH, Marinov GK, Pepke S et al (2015) Single-­ cell transcriptome analysis reveals dynamic changes in lncRNA expression during reprogramming. Cell Stem Cell 16:88–101. https://doi.org/10.1016/j. stem.2014.11.005 79. Reggiardo RE, Maroli SV, Halasz H et  al (2020) Epigenomic reprogramming of repetitive noncoding RNAs and IFN-stimulated genes by mutant KRAS.  Genomics. https://doi. org/10.1101/2020.11.04.367771 80. Simanshu DK, Nissley DV, McCormick F (2017) RAS proteins and their regulators in human disease. Cell 170:17–33. https://doi.org/10.1016/j. cell.2017.06.009 81. Wang J, Ma P, Kim DH et  al (2021) Towards microfluidic-­based exosome isolation and detection for tumor therapy. Nano Today 37:101066. https:// doi.org/10.1016/j.nantod.2020.101066 82. Johnstone RM, Adam M, Hammond JR, Orr L, Turbide C (1987) Vesicle formation during reticulocyte maturation. Association of plasma membrane activities with released vesicles (exosomes). J Biol Chem 262(19):9412–9420 83. Colombo M, Raposo G, Théry C (2014) Biogenesis, secretion, and intercellular interactions of exosomes and other extracellular vesicles. Annu Rev Cell Dev Biol 30:255–289. https://doi.org/10.1146/ annurev-­cellbio-­101512-­122326 84. Tricarico C, Clancy J, D’Souza-Schorey C (2017) Biology and biogenesis of shed microvesicles. Small GTPases 8:220–232. https://doi.org/10.1080/21541 248.2016.1215283 85. Al-Nedawi K, Meehan B, Micallef J et  al (2008) Intercellular transfer of the oncogenic receptor EGFRvIII by microvesicles derived from tumour cells. Nat Cell Biol 10:619–624. https://doi. org/10.1038/ncb1725 86. Xu R, Rai A, Chen M et al (2018) Extracellular vesicles in cancer  — implications for future improvements in cancer care. Nat Rev Clin Oncol 15:617–638. https://doi.org/10.1038/s41571-­018-­0036-­9 87. Mittelbrunn M, Sánchez-Madrid F (2012) Intercellular communication: diverse structures for

141 exchange of genetic information. Nat Rev Mol Cell Biol 13:328–335. https://doi.org/10.1038/nrm3335 88. Zhang H-G, Grizzle WE (2014) Exosomes. Am J Pathol 184:28–41. https://doi.org/10.1016/j. ajpath.2013.09.027 89. Zhang H-G, Grizzle WE (2011) Exosomes and cancer: a newly described pathway of immune suppression. Clin Cancer Res 17:959–964. https://doi. org/10.1158/1078-­0432.CCR-­10-­1489 90. Espinoza JL, Takami A, Yoshioka K et  al (2012) Human microRNA-1245 down-regulates the NKG2D receptor in natural killer cells and impairs NKG2D-­ mediated functions. Haematologica 97: 1295–1303. https://doi.org/10.3324/haematol.2011. 058529 91. Clayton A, Mitchell JP, Court J et al (2008) Human tumor-derived exosomes down-modulate NKG2D expression. J Immunol 180:7249–7258. https://doi. org/10.4049/jimmunol.180.11.7249 92. Valenti R, Huber V, Filipazzi P et al (2006) Human tumor-released microvesicles promote the differentiation of myeloid cells with transforming growth factor-β–mediated suppressive activity on T lymphocytes. Cancer Res 66:9290–9298. https://doi. org/10.1158/0008-­5472.CAN-­06-­1819 93. Yu S, Liu C, Su K et  al (2007) Tumor exosomes inhibit differentiation of bone marrow dendritic cells. J Immunol 178:6867–6875. https://doi.org/10.4049/ jimmunol.178.11.6867 94. Iero M, Valenti R, Huber V et  al (2008) Tumour-­ released exosomes and their implications in cancer immunity. Cell Death Differ 15:80–88. https://doi. org/10.1038/sj.cdd.4402237 95. Zhang H-G, Zhuang X, Sun D et al (2012) Exosomes and immune surveillance of neoplastic lesions: a review. Biotech Histochem 87:161–168. https://doi. org/10.3109/10520291003659042 96. Skog J, Würdinger T, van Rijn S et  al (2008) Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide diagnostic biomarkers. Nat Cell Biol 10:1470–1476. https://doi.org/10.1038/ncb1800 97. Hong B, Cho J-H, Kim H et  al (2009) Colorectal cancer cell-derived microvesicles are enriched in cell cycle-related mRNAs that promote proliferation of endothelial cells. BMC Genomics 10:556. https:// doi.org/10.1186/1471-­2164-­10-­556 98. Sakha S, Muramatsu T, Ueda K, Inazawa J (2016) Exosomal microRNA miR-1246 induces cell motility and invasion through the regulation of DENND2D in oral squamous cell carcinoma. Sci Rep 6:38750. https://doi.org/10.1038/srep38750 99. Kosaka N, Iguchi H, Hagiwara K et  al (2013) Neutral sphingomyelinase 2 (nSMase2)-dependent exosomal transfer of angiogenic MicroRNAs regulate cancer cell metastasis. J Biol Chem 288:10849– 10859. https://doi.org/10.1074/jbc.M112.446831 100. Zhou M, Chen J, Zhou L et  al (2014) Pancreatic cancer derived exosomes regulate the expression of TLR4  in dendritic cells via miR-203. Cell

142 Immunol 292:65–69. https://doi.org/10.1016/j. cellimm.2014.09.004 101. Zhang R, Xia Y, Wang Z et  al (2017) Serum long non coding RNA MALAT-1 protected by exosomes is up-regulated and promotes cell proliferation and migration in non-small cell lung cancer. Biochem Biophys Res Commun 490:406–414. https://doi. org/10.1016/j.bbrc.2017.06.055 102. Pan L, Liang W, Fu M et  al (2017) Exosomes-­ mediated transfer of long noncoding RNA ZFAS1 promotes gastric cancer progression. J Cancer Res Clin Oncol 143:991–1004. https://doi.org/10.1007/ s00432-­017-­2361-­2 103. Wang J, Zhou Y, Lu J et al (2014) Combined detection of serum exosomal miR-21 and HOTAIR as diagnostic and prognostic biomarkers for laryngeal squamous cell carcinoma. Med Oncol 31:148. https://doi.org/10.1007/s12032-­014-­0148-­8 104. Yang L, Peng X, Li Y et al (2019) Long non-coding RNA HOTAIR promotes exosome secretion by regulating RAB35 and SNAP23 in hepatocellular carcinoma. Mol Cancer 18:78. https://doi.org/10.1186/ s12943-­019-­0990-­6 105. Anfossi S, Babayan A, Pantel K, Calin GA (2018) Clinical utility of circulating non-coding RNAs  — an update. Nat Rev Clin Oncol 15:541–563. https:// doi.org/10.1038/s41571-­018-­0035-­x 106. FDA-NIH Biomarker Working Group (2016) BEST (Biomarkers, EndpointS, and other Tools) resource. Food and Drug Administration (US), Silver Spring 107. Lin J-A, He P (2015) Reinventing clinical trials: a review of innovative biomarker trial designs in cancer therapies. Br Med Bull 114:17–27. https://doi. org/10.1093/bmb/ldv011 108. Loeb S, Partin AW (2010) PCA3 urinary biomarker for prostate cancer. Rev Urol 12:e205–e206 109. Yan Z, Zhou Z, Wu Q et al (2020) Presymptomatic increase of an extracellular RNA in blood plasma associates with the development of Alzheimer’s disease. Curr Biol 30:1771–1782.e3. https://doi. org/10.1016/j.cub.2020.02.084 110. Yu S, Li Y, Liao Z et  al (2019) Plasma extracellular vesicle long RNA profiling identifies a diagnostic signature for the detection of pancreatic ductal adenocarcinoma. Gut:gutjnl-2019-318860. https:// doi.org/10.1136/gutjnl-­2019-­318860 111. Mandrekar SJ, Sargent DJ (2011) Design of clinical trials for biomarker research in oncology. Clin Investig 1:1627–1636. https://doi.org/10.4155/ cli.11.152 112. Hu C, Dignam JJ (2019) Biomarker-driven oncology clinical trials: key design elements, types, features, and practical considerations. JCO Precis Oncol 1:1– 12. https://doi.org/10.1200/PO.19.00086 113. Cary KC, Cooperberg MR (2013) Biomarkers in prostate cancer surveillance and screening: past, present, and future. Ther Adv Urol 5:318–329. https://doi.org/10.1177/1756287213495915

R. E. Reggiardo et al. 114. Ibarra A, Zhuang J, Zhao Y et  al (2020) Non-­ invasive characterization of human bone marrow stimulation and reconstitution by cell-free messenger RNA sequencing. Nat Commun 11. https://doi. org/10.1038/s41467-­019-­14253-­4 115. Ray P, Manach YL, Riou B, Houle TT (2010) Statistical evaluation of a biomarker. Anesthesiology 112:1023–1040. https://doi.org/10.1097/ ALN.0b013e3181d47604 116. Trevethan R (2017) Sensitivity, specificity, and predictive values: foundations, pliabilities, and pitfalls in research and practice. Front Public Health 5. https://doi.org/10.3389/fpubh.2017.00307 117. Roobol MJ, Schröder FH, van Leeuwen P et  al (2010) Performance of the prostate cancer antigen 3 (PCA3) gene and prostate-specific antigen in prescreened men: exploring the value of PCA3 for a first-line diagnostic test. Eur Urol 58:475–481. https://doi.org/10.1016/j.eururo.2010.06.039 118. Perry JN, Jasim A, Hojat A, Yong WH (2019) Procurement, storage, and use of blood in biobanks. In: Yong WH (ed) Biobanking. Springer, New York, pp 89–97 119. Carrick D, Black A, Gohagan J et  al (2015) The PLCO biorepository: creating, maintaining, and administering a unique biospecimen resource. Rev Recent Clin Trials 10:212–222. https://doi.org/10.2 174/1574887110666150730121429 120. Sorber L, Zwaenepoel K, Jacobs J et  al (2019) Circulating cell-free DNA and RNA analysis as liquid biopsy: optimal centrifugation protocol. Cancers 11:458. https://doi.org/10.3390/ cancers11040458 121. Huang L-H, Lin P-H, Tsai K-W et  al (2017) The effects of storage temperature and duration of blood samples on DNA and RNA qualities. PLoS One 12:e0184692. https://doi.org/10.1371/journal. pone.0184692 122. Zhou Z, Wu Q, Yan Z et al (2019) Extracellular RNA in a single droplet of human serum reflects physiologic and disease states. Proc Natl Acad Sci 116:19200– 19208. https://doi.org/10.1073/pnas.1908252116 123. Everaert C, Helsmoortel H, Decock A et  al (2019) Performance assessment of total RNA sequencing of human biofluids and extracellular vesicles. Sci Rep 9:1–16. https://doi.org/10.1038/ s41598-­019-­53892-­x 124. Hulstaert E, Morlion A, Cobos FA et  al (2020) Charting extracellular transcriptomes in The Human Biofluid RNA Atlas. Genomics 33:108552 125. Ngo TTM, Moufarrej MN, Rasmussen M-LH et al (2018) Noninvasive blood tests for fetal development predict gestational age and preterm delivery. Science 360:1133–1136. https://doi.org/10.1126/ science.aar3819 126. Diamandis EP (2010) Cancer biomarkers: can we turn recent failures into success? JNCI J Natl Cancer Inst 102:1462–1467. https://doi.org/10.1093/jnci/ djq306

7  LncRNA Biomarkers of Inflammation and Cancer 127. Zheng Y (2018) Study design considerations for cancer biomarker discoveries. J Appl Lab Med 3:282–289. https://doi.org/10.1373/ jalm.2017.025809 128. Maxwell SE, Kelley K, Rausch JR (2008) Sample size planning for statistical power and accuracy in parameter estimation. Annu Rev Psychol 59:537–563. https:// doi.org/10.1146/annurev.psych.59.103006.093735 129. Olvera Astivia OL, Gadermann A, Guhn M (2019) The relationship between statistical power and predictor distribution in multilevel logistic regression: a simulation-based approach. BMC Med Res Methodol 19. https://doi.org/10.1186/s12874-­019-­0742-­8 130. Guo Y, Graber A, McBurney RN, Balasubramanian R (2010) Sample size and statistical power considerations in high-dimensionality data settings: a comparative study of classification algorithms. BMC Bioinform 11:447. https://doi. org/10.1186/1471-­2105-­11-­447 131. Subramanian J, Simon R (2013) Overfitting in prediction models – is it a problem only in high dimensions? Contemp Clin Trials 36:636–641. https://doi. org/10.1016/j.cct.2013.06.011 132. Balaj L, Lessard R, Dai L et  al (2011) Tumour microvesicles contain retrotransposon elements and amplified oncogene sequences. Nat Commun 2:1–9. https://doi.org/10.1038/ncomms1180 133. Huang X, Yuan T, Tschannen M et  al (2013) Characterization of human plasma-derived exosomal RNAs by deep sequencing. BMC Genomics 14:319. https://doi.org/10.1186/1471-­2164-­14-­319 134. Raposo G, Nijman HW, Stoorvogel W et al (1996) B lymphocytes secrete antigen-presenting vesicles. J Exp Med 183:1161–1172. https://doi.org/10.1084/ jem.183.3.1161 135. Théry C, Amigorena S, Raposo G, Clayton A (2006) Isolation and characterization of exosomes from cell culture supernatants and biological fluids. Curr Protoc Cell Biol 30:3.22.1–3.22.29. https://doi. org/10.1002/0471143030.cb0322s30 136. Van Deun J, Mestdagh P, Sormunen R et al (2014) The impact of disparate isolation methods for extracellular vesicles on downstream RNA profiling. J Extracell Vesicles 3:24858. https://doi.org/10.3402/ jev.v3.24858 137. Dettenhofer M, Yu X-F (1999) Highly purified human immunodeficiency virus type 1 reveals a virtual absence of Vif in virions. J Virol 73:1460–1467. https://doi.org/10.1128/JVI.73.2.1460-­1467.1999 138. Enderle D, Spiel A, Coticchia CM et  al (2015) Characterization of RNA from exosomes and other extracellular vesicles isolated by a novel spin column-­ based method. PLoS One 10:e0136133. https://doi.org/10.1371/journal.pone.0136133 139. Liu F, Vermesh O, Mani V et al (2017) The exosome total isolation chip. ACS Nano 11:10712–10723. https://doi.org/10.1021/acsnano.7b04878 140. Dragovic RA, Gardiner C, Brooks AS et  al (2011) Sizing and phenotyping of cellular vesicles using nanoparticle tracking analysis. Nanomed

143 Nanotechnol Biol Med 7:780–788. https://doi. org/10.1016/j.nano.2011.04.003 141. Chuo ST-Y, Chien JC-Y, Lai CP-K (2018) Imaging extracellular vesicles: current and emerging methods. J Biomed Sci 25:1–10. https://doi.org/10.1186/ s12929-­018-­0494-­5 142. Rikkert LG, Nieuwland R, Terstappen LWMM, Coumans FAW (2019) Quality of extracellular vesicle images by transmission electron microscopy is operator and protocol dependent. J Extracell Vesicles 8:1555419. https://doi.org/10.1080/200130 78.2018.1555419 143. Vestad B, Llorente A, Neurauter A et al (2017) Size and concentration analyses of extracellular vesicles by nanoparticle tracking analysis: a variation study. J Extracell Vesicles 6:1344087. https://doi.org/10.10 80/20013078.2017.1344087 144. Royo F, Théry C, Falcón-Pérez JM et  al (2020) Methods for separation and characterization of extracellular vesicles: results of a worldwide survey performed by the ISEV rigor and standardization subcommittee. Cell 9:1955. https://doi.org/10.3390/ cells9091955 145. Bussemakers MJ, van Bokhoven A, Verhaegh GW et  al (1999) DD3: a new prostate-specific gene, highly overexpressed in prostate cancer. Cancer Res 59:5975–5979 146. Marks LS, Fradet Y, Lim Deras I et al (2007) PCA3 molecular urine assay for prostate cancer in men undergoing repeat biopsy. Urology 69:532–535. https://doi.org/10.1016/j.urology.2006.12.014 147. Marioni JC, Mason CE, Mane SM et al (2008) RNA-­ seq: an assessment of technical reproducibility and comparison with gene expression arrays. Genome Res 18:1509–1517. https://doi.org/10.1101/ gr.079558.108 148. Shendure J (2008) The beginning of the end for microarrays? Nat Methods 5:585–587. https://doi. org/10.1038/nmeth0708-­585 149. Cloonan N, Forrest ARR, Kolle G et  al (2008) Stem cell transcriptome profiling via massive-scale mRNA sequencing. Nat Methods 5:613–619. https:// doi.org/10.1038/nmeth.1223 150. Mortazavi A, Williams BA, McCue K et  al (2008) Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods 5:621–628. https://doi.org/10.1038/nmeth.1226 151. Zhang W, Yu Y, Hertwig F et al (2015) Comparison of RNA-seq and microarray-based models for clinical endpoint prediction. Genome Biol 16. https://doi. org/10.1186/s13059-­015-­0694-­1 152. Hu Z, Dong J, Wang L-E et  al (2012) Serum microRNA profiling and breast cancer risk: the use of miR-484/191 as endogenous controls. Carcinogenesis 33:828–834. https://doi.org/10.1093/ carcin/bgs030 153. Brase JC, Johannes M, Schlomm T et  al (2011) Circulating miRNAs are correlated with tumor progression in prostate cancer. Int J Cancer 128:608– 616. https://doi.org/10.1002/ijc.25376

144 154. Koh W, Pan W, Gawad C et al (2014) Noninvasive in  vivo monitoring of tissue-specific global gene expression in humans. Proc Natl Acad Sci 111:7361– 7366. https://doi.org/10.1073/pnas.1405528111 155. Guttman M, Garber M, Levin JZ et  al (2010) Ab initio reconstruction of cell type–specific transcriptomes in mouse reveals the conserved multi-exonic structure of lincRNAs. Nat Biotechnol 28:503–510. https://doi.org/10.1038/nbt.1633 156. Sexton CE, Han MV (2019) Paired-end mappability of transposable elements in the human genome. Mob DNA 10. https://doi.org/10.1186/ s13100-­019-­0172-­5 157. Chhangawala S, Rudy G, Mason CE, Rosenfeld JA (2015) The impact of read length on quantification of differentially expressed genes and splice junction detection. Genome Biol 16. https://doi.org/10.1186/ s13059-­015-­0697-­y 158. Volden R, Palmer T, Byrne A et  al (2018) Improving nanopore read accuracy with the R2C2 method enables the sequencing of highly multiplexed full-length single-cell cDNA.  Proc Natl Acad Sci 115:9726–9731. https://doi.org/10.1073/ pnas.1806447115 159. Sessegolo C, Cruaud C, Da Silva C et  al (2019) Transcriptome profiling of mouse samples using nanopore sequencing of cDNA and RNA molecules. Sci Rep 9. https://doi.org/10.1038/ s41598-­019-­51470-­9 160. Leggett RM, Alcon-Giner C, Heavens D et  al (2020) Rapid MinION profiling of preterm microbiota and antimicrobial-resistant pathogens. Nat Microbiol 5:430–442. https://doi.org/10.1038/ s41564-­019-­0626-­z 161. Workman RE, Tang AD, Tang PS et  al (2019) Nanopore native RNA sequencing of a human poly(A) transcriptome. Nat Methods 16:1297–1305. https://doi.org/10.1038/s41592-­019-­0617-­2 162. Liu H, Begik O, Lucas MC et  al (2019) Accurate detection of m6A RNA modifications in native RNA sequences. Nat Commun 10. https://doi.org/10.1038/ s41467-­019-­11713-­9 163. Levin JZ, Yassour M, Adiconis X et  al (2010) Comprehensive comparative analysis of strand-­ specific RNA sequencing methods. Nat Methods 7:709–715. https://doi.org/10.1038/nmeth.1491 164. Vecera M, Sana J, Oppelt J et  al (2019) Testing of library preparation methods for transcriptome sequencing of real life glioblastoma and brain tissue specimens: a comparative study with special focus on long non-coding RNAs. PLoS One 14:e0211978. https://doi.org/10.1371/journal.pone.0211978 165. Hrdlickova R, Toloue M, Tian B (2017) RNA-Seq methods for transcriptome analysis: RNA-Seq. Wiley Interdiscip Rev RNA 8:e1364. https://doi. org/10.1002/wrna.1364 166. Schuierer S, Carbone W, Knehr J et al (2017) A comprehensive assessment of RNA-seq protocols for degraded and low-quantity samples. BMC Genomics 18. https://doi.org/10.1186/s12864-­017-­3827-­y

R. E. Reggiardo et al. 167. Zhao S, Zhang Y, Gamini R et al (2018) Evaluation of two main RNA-seq approaches for gene quantification in clinical RNA sequencing: polyA+ selection versus rRNA depletion. Sci Rep 8:4781. https://doi. org/10.1038/s41598-­018-­23226-­4 168. Geraci F, Saha I, Bianchini M (2020) Editorial: RNA-seq analysis: methods, applications and challenges. Front Genet 11. https://doi.org/10.3389/ fgene.2020.00220 169. Conesa A, Madrigal P, Tarazona S et  al (2016) A survey of best practices for RNA-seq data analysis. Genome Biol 17:13. https://doi.org/10.1186/ s13059-­016-­0881-­8 170. Parekh S, Ziegenhain C, Vieth B et  al (2016) The impact of amplification on differential expression analyses by RNA-seq. Sci Rep 6. https://doi. org/10.1038/srep25533 171. Fu Y, Wu P-H, Beane T et  al (2018) Elimination of PCR duplicates in RNA-seq and small RNA-seq using unique molecular identifiers. BMC Genomics 19. https://doi.org/10.1186/ s12864-­018-­4933-­1 172. Li H, Handsaker B, Wysoker A et  al (2009) The sequence alignment/map format and SAMtools. Bioinformatics 25:2078–2079. https://doi. org/10.1093/bioinformatics/btp352 173. Dobin A, Davis CA, Schlesinger F et  al (2013) STAR: ultrafast universal RNA-seq aligner. Bioinform Oxf Engl 29:15–21. https://doi. org/10.1093/bioinformatics/bts635 174. Patro R, Duggal G, Love MI et  al (2017) Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods 14:417–419. https:// doi.org/10.1038/nmeth.4197 175. Bray NL, Pimentel H, Melsted P, Pachter L (2016) Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol 34:525–527. https://doi.org/10.1038/ nbt.3519 176. Verwilt J, Trypsteen W, Van Paemel R et al (2020) When DNA gets in the way: a cautionary note for DNA contamination in extracellular RNA-seq studies. Proc Natl Acad Sci 117:18934–18936. https:// doi.org/10.1073/pnas.2001675117 177. Bendall ML, de Mulder M, Iñiguez LP et al (2019) Telescope: characterization of the retrotranscriptome by accurate estimation of transposable element expression. PLoS Comput Biol 15:e1006453. https://doi.org/10.1371/journal.pcbi.1006453 178. Yang WR, Ardeljan D, Pacyna CN et  al (2019) SQuIRE reveals locus-specific regulation of interspersed repeat expression. Nucleic Acids Res 47:e27–e27. https://doi.org/10.1093/nar/gky1301 179. Jin Y, Tam OH, Paniagua E, Hammell M (2015) TEtranscripts: a package for including transposable elements in differential expression analysis of RNA-­ ­ seq datasets. Bioinformatics 31:3593–3599. https://doi.org/10.1093/bioinformatics/btv422 180. Francis OE, Bendall M, Manimaran S et  al (2013) Pathoscope: species identification and strain attribution with unassembled sequencing data. Genome

7  LncRNA Biomarkers of Inflammation and Cancer Res 23:1721–1729. https://doi.org/10.1101/ gr.150151.112 181. Lanciano S, Cristofari G (2020) Measuring and interpreting transposable element expression. Nat Rev Genet. https://doi.org/10.1038/s41576-­020-­0251-­y 182. Abrams ZB, Johnson TS, Huang K et  al (2019) A protocol to evaluate RNA sequencing normalization methods. BMC Bioinform 20. https://doi. org/10.1186/s12859-­019-­3247-­x 183. Zhao S, Ye Z, Stanton R (2020) Misuse of RPKM or TPM normalization when comparing across samples and sequencing protocols. RNA 26:903–909. https:// doi.org/10.1261/rna.074922.120 184. Vivian J, Rao AA, Nothaft FA et  al (2017) Toil enables reproducible, open source, big biomedical data analyses. Nat Biotechnol. https://www.nature. com/articles/nbt.3772. Accessed 3 Apr 2018 185. Evans C, Hardin J, Stoebel DM (2018) Selecting between-sample RNA-Seq normalization methods from the perspective of their assumptions. Brief Bioinform 19:776–792. https://doi.org/10.1093/bib/ bbx008 186. Jiang L, Schlesinger F, Davis CA et  al (2011) Synthetic spike-in standards for RNA-seq experiments. Genome Res 21:1543–1551. https://doi. org/10.1101/gr.121095.111 187. Clarke R, Ressom HW, Wang A et  al (2008) The properties of high-dimensional data spaces: implications for exploring gene and protein expression data. Nat Rev Cancer 8:37–49. https://doi.org/10.1038/ nrc2294

145 188. Frost HR, Amos CI (2017) Gene set selection via LASSO penalized regression (SLPR). Nucleic Acids Res 45:e114–e114. https://doi.org/10.1093/nar/ gkx291 189. Ghosh D, Chinnaiyan AM (2005) Classification and selection of biomarkers in genomic data using LASSO.  J Biomed Biotechnol 2005:147–154. https://doi.org/10.1155/JBB.2005.147 190. Stein-O’Brien GL, Arora R, Culhane AC et al (2018) Enter the matrix: factorization uncovers knowledge from omics. Trends Genet 34:790–805. https://doi. org/10.1016/j.tig.2018.07.003 191. Molinaro AM, Simon R, Pfeiffer RM (2005) Prediction error estimation: a comparison of resampling methods. Bioinformatics 21:3301–3307. https://doi.org/10.1093/bioinformatics/bti499 192. Simon R (2008) Development and validation of biomarker classifiers for treatment selection. J Statist Plann Inference 138:308–320. https://doi. org/10.1016/j.jspi.2007.06.010 193. de Bono JS, Ashworth A (2010) Translating cancer research into targeted therapeutics. Nature 467:543– 549. https://doi.org/10.1038/nature09339 194. Sawyers CL (2008) The cancer biomarker problem. Nature 452:548–552. https://doi.org/10.1038/ nature06913 195. Subramanian J, Simon R (2010) Gene expression-­ based prognostic signatures in lung cancer: ready for clinical use? JNCI J Natl Cancer Inst 102:464–474. https://doi.org/10.1093/jnci/djq025

8

Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs Ainara Castellanos-Rubio and Sankar Ghosh

Abstract

Genome wide association studies (GWAS) have identified many loci contributing to genetic variation of complex traits. Immune mediated disorders are complex diseases for which hundreds of risk alleles have been identified by GWAS. However, the intergenic location of most of the signals has make it difficult to decipher their implication in disease pathogenesis. A significant number of immune disease-­ associated SNPs are located within long noncoding RNAs (lncRNAs). LncRNAs have gained importance due to their involvement in the regulation of a wide range of biological processes, including immune responses. GWAS SNPs located within lncRNAs can affect their regulatory capacity by modifying their secondary structure, altering their expression levels or regulating the transcription of different isoforms. In this A. Castellanos-Rubio Department of Genetics, Physical Anthropology and Animal Physiology, University of the Basque Country, Leioa, Spain Ikerbasque, Basque Foundation for Science, Bilbao, Spain S. Ghosh (*) Department of Microbiology & Immunology, Vagelos College of Physicians & Surgeons, Columbia University, New York, NY, USA e-mail: [email protected]

review we discuss the functional implications of immune-related lncRNAs harboring disease associated SNPs on various disease conditions. Keywords

GWAS · Long noncoding RNA · Lnc13 · Autoimmunity · Inflammation

8.1

Introduction

Genome wide association studies (GWAS) have been used for the past couple of decades to detect associations between genetic variants and complex traits. The main aim of GWAS has been to better understand the biology of disease [92]. GWAS have demonstrated that, in general, many loci contribute to the genetic variation of complex traits, thus polymorphisms in several genes are associated with disease susceptibility in the population, and the risk associated with individual variants is generally very small [103]. Immune mediated disorders, such as celiac disease (CeD), inflammatory bowel disease (IBD), rheumatoid arthritis (RA), type 1 diabetes (T1D) or multiple sclerosis (MS) among others, are clinically heterogeneous, complex disorders that share common pathogenic mechanisms. The general belief is that they develop due to an imbalance in the interaction between genetic and

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_8

147

148

environmental factors, but the underlying mechanisms leading to their development are not fully understood [16, 104]. GWAS have identified hundreds of risk alleles, both, common and disease specific, significantly increasing our knowledge about the genetic variants underlying immune diseases [19]. Additionally, 200,000 polymorphisms in 186 immune disease related regions have been analyzed using the so called Immunochip platform, which allowed for the identification of additional immune disease associated variants thereby revealing common susceptibility loci for several diseases [2, 5, 73, 81]. Moreover, SNPs that have opposing risk alleles in different diseases, or disease-specific associated SNPs at the same loci have also been detected suggesting extensive pleiotropy [92, 104]. Although GWAS and follow up studies have identified hundreds of genetic variants in immune diseases, the link between SNPs and disease biology is not straightforward, mainly because nearly 90% of the disease associated SNPs are located in noncoding regions of the genome. Hence our understanding of the role of different disease associated SNPs in the etiology of disease remains limited [22, 38, 89]. Recent advances in transcriptome sequencing and annotation of the human genome have revealed that many noncoding regions of the genome encode long non-coding RNAs (lncRNAs). LncRNAs are defined as RNA molecules longer than 200 bp in length with no or very little protein-coding potential. The mechanisms of action of lncRNAs are heterogeneous and not well defined, and only a small number of them have been functionally characterized. It is known that they are expressed at very low levels and that their expression is generally tissue specific. LncRNAs have been classified as fundamental regulators of transcription due to their ability to control different levels of the gene expression program. LncRNAs have been implicated in posttranscriptional regulation via protein synthesis, RNA maturation, and transport, and also in transcriptional gene silencing through regulating the chromatin state [11, 65, 87, 99]. A significant number of immune disease-­ associated SNPs are located within lncRNA

A. Castellanos-Rubio and S. Ghosh

sequences suggesting that these lncRNAs might be involved in the development of these disorders [36]. Expression profile analyses of autoimmune disease-associated regions showed that lncRNAs are enriched in these loci, and alterations in the structure and function of lncRNAs have been associated with several immune-mediated diseases [40, 53]. However, in the majority of cases the mechanism by which lncRNA variants contribute to disease pathogenesis remains unknown. SNPs in lncRNAs have been generally studied using approaches that had been previously applied to protein coding genes. Several disease-­ associated SNPs located within or near lncRNAs, and previously thought to regulate adjacent coding genes, have been described as cis-eQTLs (expression Quantitiative Trait Loci) for the lncRNAs. It has been hypothesized that the genomic variants that influence the regulation of lncRNAs may play a critical role in disease pathogenesis. These SNPs do not directly regulate the expression of coding genes, as it was previously suggested, but alter lncRNA levels which affect the neighboring protein-coding genes. Thereby, these lncRNAs act as the link between the SNPs and the coding genes that are altered in disease pathogenesis [35, 50, 60, 77, 90]. An alternative mechanism by which disease associated SNPs affect lncRNAs is through exon-­ skipping. The presence of different alleles in the exon-intron junctions of lncRNAs generate different isoforms with differential ability to regulate downstream events. For example, SNPs in the ANRIL lncRNA have been suggested to be involved in alternative splicing events that modify its structure and thereby the regulation of downstream inflammatory genes [1, 12]. Last but not least, disease associated SNPs have also been proposed to directly affect lncRNA structure. Recent studies have shown that secondary structure of lncRNAs play an important role in the genetic and epigenetic processes regulated by the lncRNA and it has been shown that the function of some lncRNAs depends on the formation of the normal structure. Computational analysis have predicted secondary structure changes induced by SNP alleles and these conformational changes alter the affinity of the lncRNAs to bind

8  Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs

A

B

eQTL

Allele A

Allele B

Altered lncRNA expression

C

Exon skipping

Allele A

Allele B

Different isoforms

149

Secondary structure

Allele A

Allele B

Differential binding affinity

Fig. 8.1  Different effects of immune-disease associated SNP on lncRNA regulation and function. Red dots represent disease associated SNPs. (a) A SNP located near the transcription start site of a lncRNA can regulate the amount of transcribed lncRNA (cis-eQTL) by altering transcription factor binding sites (TFBS), chromatin accessibility or enhancer regulation, in addition to other

mechanisms. (b) An intronic SNP in a lncRNA can result in exon-skipping thereby altering splicing and generating different lncRNA isoforms with different functions. (c) SNPs located within lncRNA exons can influence the RNA secondary structure thus changing the affinity of the lncRNA to its binding partners. Allele A and allele B represent the two possible alleles of the associated SNP

to protein interaction partners [29, 34, 44, 59, 80]. Although it seems clear that lncRNA secondary structure plays an important role in their biological roles, it is still a relatively unexplored field (Fig. 8.1). Although genome browsers as UCSC [46] or Ensembl [106] can be used to find SNPs located within specific lncRNA sequences, some effort has been invested into the development of tools that find relationships between disease associated SNPs and lncRNAs. Among them, lncRNASNP [64] and LincSNP [68] databases are two of the most complete, as you can search using your lncRNA or SNP of interest and get information about their type of expected interaction. Overall, the identification of intergenic SNPs in lncRNAs through GWAS has opened the door to analyzing long noncoding RNA transcripts in the context of complex disease pathogenesis. Although the effect of each associated SNP is small, many such SNPs are in linkage disequilibrium (LD) which means that particular alleles at nearby sites are transmitted in clusters (known as haplotype block) more often than is expected by chance [95] and studying their role in lncRNA function will contribute to the understanding of

disease pathogenesis and help identify new potential therapeutic targets. In this article, we review the involvement of five long noncoding RNAs harboring GWAS SNPs that have been associated with immune disorders and discuss the functional characterization of the lncRNAs in the context of the SNP alleles (Fig. 8.2).

8.2

L nc13: Regulator of Inflammatory Responses in Celiac Disease and Type 1 Diabetes

Celiac disease (CeD) and type 1 diabetes (T1D) are complex, chronic, autoimmune diseases that develop in genetically susceptible individuals. The strongest genetic association for both disorders maps to the human leukocyte antigen (HLA) region in chromosome 6p21 [15, 21, 33, 43]. GWAS, together with the Immunochip project, have identified several SNPs in non-HLA loci associated with the genetic risk of CeD and T1D. Some SNPs are associated to both diseases, while other SNPs are only associated to one of

A. Castellanos-Rubio and S. Ghosh

Macrophages, chromatin PCBP2 lnc13

STAT1

Pancreatic beta cells, cytoplasm

B ATF1

Raly

CCR5AS

CCR5

Viral entry

CXCL10 CXCL9 CCL5 CXCL1

miRNAs

T cell activation Inflammatory genes

T cells

E

miRNAs

Apoptosis

NFkB

Monocytes, chondrocytes

C

nuclear proteins

MEG3

LINC00305

Inflammatory genes

T-bet

bowel disease

T1D

Lipopolysaccharide

and atherosclerosis

Inflammatory genes STAT1, TRAF2, IL1RN, IL2RA

rs2850711: rheumatoid arthritis

lnc13

A

rs7134599: inflammatory

HDAC1 HNRNPD

rs34552516 : type 1 diabetes and rheumatoid arthritis

rs917997: celiac disease and type1 diabetes

CeD

CD4+ T cells, PBMCs

rs1015164: HIV viral load

LncRNAs harboring immune disease GWAS SNP

150

WDR5

D

IFNG-AS1

H3K4me3

IFNG T cells, NK cells

Fig. 8.2  Schematic representation of the mechanisms of lncRNAs harboring immune-disease GWAS SNP. (a) Lnc13 harbors SNP rs917997 which is associated to CeD (risk allele *T) and to T1D (risk allele *C). This SNP changes the secondary structure of the lncRNA modifying its binding affinity to the proteins hnRNPD and HDAC1 in the nucleus of monocytes and with PCBP2  in the cytoplasm of pancreatic beta-cells. The altered binding increases the levels of inflammatory cytokines augmenting disease predisposition. (b) The SNP rs1015164 associated with viral load of HIV-1 is located upstream of CCR5AS. An SNP that is in LD with rs1015164 alters the binding of the transcription factor ATF1 which affects lncRNA expression levels. Additionally, higher levels of CCR5AS sequester the Raly protein thereby lowering the degradation of CCR5 mRNA, facilitating viral entry and increased viral load. (c) The SNP rs34552516 associated with type 1 diabetes and rheumatoid arthritis is located

within a MEG3 intron. The binding to nuclear proteins has been shown to be altered in the presence of the different alleles and it appears that the SNP can regulate lncRNA expression in T cells. The capacity of MEG3 to regulate T cell function or induce inflammatory genes has been attributed to the ability of this lncRNA to act as a sponge for different miRNAs. (d) IFNG-AS1 is located near the IFNG gene and harbors the IBD associated SNP rs7134599, located in the lncRNA enhancer. Activation of its transcription by T-bet induces the levels of IFNG by WDR5 mediated H3K4me3 methylation. (e) The SNP rs2850711 is located in an intron of LINC00305 and is associated with rheumatoid arthritis and atherosclerosis. This lncRNA is increased in patients and is induced by inflammatory signals, as lipopolysaccharide, in monocytes and chondrocytes. LINC00305 can also act as a sponge for different miRNAs regulating both the activation of NFkB and downstream inflammatory gene expression and apoptotic processes

the diseases. One of the SNPs associated with both diseases is rs917997, located in the region 2q12, 1.5 kb away from the coding gene IL18RAP, that has been proposed, but never confirmed, as the functional candidate gene in the region [37, 67, 75, 91]. Interestingly, while the risk signal in celiac disease corresponds to the T allele, the opposite allele (C) is the risk allele in T1D. rs917997 was found to be located in a lncRNA, named lnc13. Although the GWAS disease association has generally been attributed to the SNP

rs917997 [42], linkage analysis of the lnc13 region revealed that a total of 6 SNPs in absolute linkage disequilibrium are located within the lncRNA sequence. These six nucleotide changes in lnc13 cause a disruption its secondary structure which affects its function [8, 31]. Lnc13 is expressed in different human cells and tissues, including mononuclear cells in the lamina propria, and human pancreatic β-cells. Functional studies using cell lines, and human samples representing both diseases, revealed that

8  Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs

this lncRNA has a very cell specific mechanism for the regulation of inflammatory genes that contribute to the development of both diseases [8, 31]. Lnc13 quantification in small intestinal biopsy samples from celiac patients and controls showed markedly lower levels of this lncRNA in CeD samples, contrary to the expression of the neighboring coding mRNA, IL18RAP [75]. The characterization of the regulation, function and mechanisms of action of lnc13 in mononuclear cells revealed that under basal conditions lnc13 represses the expression of a subset of CeD related genes (STAT1, MYD88, IL1RA and TRAF2) via its interaction with hnRNPD (Heterogeneous Nuclear Ribonucleoprotein D), a nuclear AU1 rich RNA binding protein, and HDAC1 (Histone Deacetylase 1), a histone deacetylase which negatively regulates transcription, in a chromatin-bound protein-RNA complex. In response to inflammatory stimuli, lnc13 is degraded by Decapping enzyme 2 (DCP2), thereby releasing the protein complex from chromatin and allowing the expression of the bound proinflammatory genes. The CeD risk allele of the SNP rs917997 lowers the affinity of lnc13 for hnRNPD and chromatin, resulting in higher expression of the proinflammatory genes. Hence, patients with the risk haplotype have a higher basal expression of CeD related inflammatory genes, thereby increasing their predisposition to develop the disease [8]. In contrast to its role in CeD, human pancreatic islets harboring the T1D-associated SNP risk genotype in lnc13 (CC) showed higher STAT1 expression than islets harboring the heterozygous genotype (CT). Upregulation of lnc13 in pancreatic beta cells increased activation of the pro-­ inflammatory STAT1 pathway, which correlated with increased production of chemokines in an allele-specific manner. Lnc13 was shown to induce STAT1 protein levels by stabilizing its mRNA via allele specific interactions with the protein PCBP2 (Poly(rC)-binding protein 2) in the cytoplasm. Viral infections, have been proposed as a trigger for T1D [70], and it was observed that a viral mimic induces lnc13 translocation from the nucleus to the cytoplasm, pro-

151

moting the interaction of STAT1 mRNA with PCBP2 [31]. In summary, it appears that lnc13 and the inflammation associated SNP rs917997 contribute to the pathology of celiac disease and type 1 diabetes by regulating expression of certain immune related genes in a cell and allele-specific manner, highlighting the importance of studying lncRNA function in the context of specific disease and cell type.

8.3

 CR5AS: Influences HIV-1 C Viral Load

Human immunodeficiency virus (HIV) infection causes a chronic, generalized immune activation characterized by increased levels of different immune cell type activation markers that can be linked to disease progression [17]. HIV-1, uses CD4 receptor and the co-receptors CXCR4 and CCR5 to enter target cells [4]. An inactivating mutation in the CCR5 gene, the major coreceptor of the virus, has been linked to resistance to HIV-1 infection due to the loss of the receptor from the surface [56, 84]. Different SNP haplotypes have been associated with CCR5 expression levels and thus, with disease susceptibility [30, 66]. When studying the association between HIV-1 virus load and genetic variations, the SNP rs1015164 located 3 kb away from CCR5, downstream of the CCRL2 gene, showed genome wide significance independent of other SNPs in the region, and its effect was unrelated to that of the inactivating mutation in CCR5 [61]. The SNP rs1015164 was found to map to the 5′ region of the noncoding RNA transcript CCR5AS, that was found to be expressed in peripheral blood lymphocytes and CD4+T cells. Interestingly, this SNP was an eQTL for the lncRNA not only in CD4+T cells, but also in other tissues such as the colon or heart. Coexpression analysis revealed that the expression of this lncRNA positively correlated with that of CCR5, and silencing and overexpression experiments confirmed that CCR5AS was involved in the expression of cell surface expression of CCR5 [49].

A. Castellanos-Rubio and S. Ghosh

152

Although some lncRNAs have been related to different aspects of HIV infection [10, 78], CCR5AS is the only one so far regulated by a SNP associated with outcome of HIV-1 disease. In silico analysis revealed that SNP rs2027820 located in the first intron of the CCR5AS lncRNA, which is in perfect linkage disequilibrium (LD) with the associated rs1015164 SNP and lies within a binding site for the transcription factor ATF1, is predicted to alter the binding potential of ATF1 to the lncRNA. ATF1 is a cyclic AMP dependent transcription factor that has been shown to act as a molecular sensor of inflammatory stimuli in metastatic melanomas [62]. Although the role of ATF1 in HIV-1 infection has not been described so far, the authors of this study concluded that the SNP rs2027820, in LD with rs1015164, influenced the expression of CCR5AS by regulating its transcription via ATF1 binding. Higher expression of CCR5AS, characteristic of individuals harboring the A risk allele, leads to higher CCR5 surface expression which will, in turn influence viral loads. Downstream mechanistic characterization of the lncRNA function showed that CCR5AS binds to the Raly protein (RNA-binding protein Raly). Raly had been previously found bound to the cores of HIV-1 virions suggesting that it could be involved in either virus assembly or early infection events [85]. However, unrelated to previous findings, Raly was reported to regulate the stability of CCR5 mRNA by binding to its 5′UTR. The sequestering of Raly by CCR5AS protected the CCR5 induced degradation by Raly binding. In the context of the risk allele, higher expression of CCR5AS will sequester more Raly, thereby avoiding CCR5 degradation, inducing a higher surface expression of CCR5, thereby facilitating viral entry. CCR5 has also been implicated in the development of other inflammatory and autoimmune diseases such as atherosclerosis and IBD [45, 63]. Although the SNP rs2027820 does not seem to be related to these other diseases, the fact that it can regulate the expression of lncRNA CCR5AS in colon and heart suggests that this lncRNA may also have a role in the pathogenesis of other inflammatory disorders.

8.4

 EG3: Type 1 Diabetes M and Rheumatoid Arthritis Immune Activation

Rheumatoid arthritis (RA) is an autoimmune disease in which chronic inflammation leads to joint destruction. In type 1 diabetes (T1D) autoimmune reactivity leads to destruction of pancreatic β cells and loss of insulin production. Genome-­ wide association studies have identified SNPs implicating CD4+ T-cell function in the autoimmune process of both diseases [24, 41]. MEG3 is a maternally expressed, imprinted lncRNA located in the Immunochip region 14q32. MEG3 has been implicated in embryonic development, and it regulates tumor suppression via p53 binding [23, 111]. MEG3 has also been implicated in T cell activation and induction of inflammatory pathways mainly via its ability to competitively inhibit different miRNAs [52, 97]. Specifically, MEG3 has been involved in the regulation of Treg/Th17 balance in asthma and immune thrombocytopenic purpura by acting as a sponge for miRNAs that are involved in the inflammatory processes of these diseases [51, 79]. Moreover, experimental evidence suggests that MEG3 is involved in the inflammatory response characteristic of T1D and RA via apoptosis related mechanisms [57, 100]. Different SNPs within MEG3 lncRNA have been associated with T1D and have also been suggested to be associated with rheumatoid arthritis (RA) [47, 96, 102]. In a fine mapping study in which 76 loci were tested for T1D and RA to find causal variants, two SNPs significantly associated to T1D were found within MEG3 (signals did not reach statistical significance for RA). Genome wide data analysis together with EMSA and luciferase experiments suggested that the associated SNP rs34552516, located within an intron, shows allele specific nuclear protein binding in T cells pointing to an involvement of this SNP in regulating the function of the lncRNA [47, 102]. This SNP seems to be an eQTL for the gene DLK1, located nearby and coexpressed with MEG3 in the esophagus, but further studies have not been performed to characterize this variant.

8  Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs

The SNP rs941576 located in intron 6 of MEG3 has also been analyzed for its association with T1D and RA, again showing signs of association with T1D, and misleading results in the case of RA [3, 47, 94, 96]. A study done in the Egyptian population found significant association of the SNP with RA, and decreased levels of MEG3 in the serum of patients. Although no exact causality of this SNP has been described, RA patients harboring the risk allele showed even lower levels of MEG3. Correlation analysis with different factors measured in serum samples, led to the conclusion that MEG3 is involved in apoptosis and angiogenesis via a HIF-1a (Hypoxia Inducible factor 1a) induced downregulation of the lncRNA, but further mechanistic analysis is needed to confirm this hypothesis [94]. Regarding T1D, although SNP rs941576 seems clearly associated with disease risk, there are no functional studies about the allele-specific role of the SNP on the regulation of the lncRNA and its role in disease pathogenesis.

8.5

I FNG-AS1: IFNG Induction in Inflammatory Bowel Disease

Inflammatory bowel diseases (IBD) are common, chronic inflammatory gastrointestinal disorders clinically comprised of Crohn’s disease (CD) and ulcerative colitis (UC) [26]. These diseases are thought to be triggered due to inappropriate inflammatory response to intestinal microbes and foreign antigens in genetically susceptible individuals [20, 101]. Among the SNPs associated with IBD by GWAS, rs7134599, is located in the 12q15 region, 50 kb away from the 3′UTR of IFNG and within the sequence of IFNG-AS1 lncRNA (also termed NeST or Tmevpg1). IFNG-AS1 is significantly overexpressed in intestinal samples of ulcerative colitis patients and its expression appears to correlate with elevated levels of IFNG, IL1, IL6, and TNF-α in patients [72]. Higher levels of IFNG-AS1 have been also observed in other inflammatory diseases such as multiple sclerosis or Sjögren syndrome [39, 98].

153

IFNG-AS1 lncRNA was first related with the immune response in the context of susceptibility to persistent Theiler’s virus infection. It was observed that IFNG-AS1 is expressed in immune cells of mouse and human origin and it was demonstrated that it contributes to regulation of IFNG expression as part of the Th1 differentiation program. T-bet guides epigenetic remodeling of the enhancer of IFNG-AS1 lncRNA, leading to recruitment of stimulus-inducible transcription factors, such as NF-|B [13, 14]. Functional analysis performed in 293T cells suggested that IFNG-AS1 is a trans acting nuclear lncRNA that binds WDR5 (WD Repeat containing protein 5), a component of active chromatin remodeling complexes, to induce Ifng gene transcription by increasing H3K4me3 methylation of chromatin [28]. More recently, two genetic mouse models with disrupted Ifngas1 locus were generated showing that not only Ifngas1 transcription, but also chromatin organization of the locus is critical for Ifngas1 lncRNA function regarding Ifng expression regulation [74]. Moreover, IFNG-AS1 has been shown to also have functional relevance in memory T cells and NK cells [74, 88]. Both cell types, memory T cells and NK cells, play an important role in the pathogenesis and evolution of IBD, and the potential regulation of IFNG by IFNG-AS1 in these cells further support the involvement of this lncRNA in the development of IBD [69, 76]. Although experimental evidence points to an involvement of IFNG-AS1 in the pathogenesis of IBD, the functional impact of each allele of the rs7134599 SNP in disease development has not been analyzed so far. Sequence analysis shows 10 SNPs in total LD with the associated SNP within the lncRNA sequence. However, all of them are located in intronic regions and most likely do not influence the secondary structure of the lncRNA.  Some of these SNPs do overlap with enhancer histone marks, or are predicted to disturb protein binding motifs, which could affect the expression of the lncRNA and consequently the levels of IFNG, thereby contributing to disease susceptibility. Additionally, eQTL analysis using GTEx portal [58] points to a significant association of the SNP rs7134599 with IFNG-AS1

A. Castellanos-Rubio and S. Ghosh

154

expression in whole blood, which suggests that individuals with different allele genotypes will have different levels of IFNG-AS1, predisposing them to disease. Even if IFNG-AS1 is involved in immune response and inflammatory processes related to IBD pathogenesis, and in silico data suggests alteration of lncRNA expression by the IBD associated SNPs, functional studies are still needed to confirm the relevance of these SNPs in IFNG-AS1 function and their involvement in the development of disease.

8.6

LINC00305: Monocyte Activation in Rheumatoid Arthritis and Atherosclerosis

Rheumatoid arthritis (RA) is a chronic inflammatory joint disease that can cause damage to the cartilage and bone [86]. Atherosclerosis in turn, is a chronic inflammatory disease of the arterial wall whose prevalence is increased in rheumatoid arthritis patients [55]. Both are complex diseases, in which secretion of proinflammatory cytokines by monocytes plays an important role [27, 105]. GWAS have led to the identification of a substantial number of genetic loci associated to both diseases, and some of them have been related to the risk of RA patients developing atherosclerosis. Although several lncRNAs have been suggested to be involved in the development of both pathologies, only one has been described so far harboring a disease associated SNP. The associated SNP rs2850711 is located within an intronic sequence in the lncRNA LINC00305. The SNP rs2850711 was associated to atherosclerosis by GWAS and more recently to RA by a candidate gene association study [18, 55, 83, 93]. LINC00305 was found to be overexpressed in serum of RA patients, and in atherosclerotic plaques and peripheral blood mononuclear cells (PBMCs) from atherosclerosis patients, supporting its potential role in these diseases. Although no quantification of the lncRNA levels has been performed in rheumatoid joints, LINC00305 levels in the serum of RA patients seems to be an adequate biomarker for disease diagnosis, signif-

icantly correlating with clinical characteristics of the disease as the joint disease activity score. Additionally, analysis of LINC00305 in the cell types that compose the atherosclerosis plaques revealed that monocytes are the primary cell types expressing this lncRNA.  Furthermore, in vitro studies in the context of atherosclerosis, revealed that this lncRNA may be involved in the development of atherosclerotic plaques by inducing a change of phenotype in the vascular smooth muscle characteristic of the disease, that is mediated by the cytokines secreted by inflammatory cells. Stable overexpression of LINC00305 in human monocytes cocultured with human aortic smooth muscle cells showed that the muscle cells presented lower expression of their basal markers, suggesting that they were switching to the pathogenic phenotype and confirming the involvement of the lincRNA in the development of the atherosclerosis plaques [93, 109]. Mechanistic characterization of LINC00305, has revealed that its expression is induced in response to stimulation with lipopolysaccharide (LPS) in both monocytic and chondrogenic cell lines. This lincRNA has been related with NF-κB activation and subsequent inflammatory gene expression, and also with apoptosis induction via its ability to sequester miRNAs in different cell types [54, 71, 93, 107–109]. However, the exact mechanisms by which this lncRNA induces inflammation or/and apoptosis in each disease tissue has not been described and may very well be cell type specific as seen with other lncRNAs (see lnc13 above). Although it appears that LINC00305 plays a role in development of rheumatoid arthritis and atherosclerosis by contributing to the inflammatory environment and very likely by regulating apoptosis via miRNA sponging, the role of the associated SNPs in the function of the lincRNA is still mainly unexplored. The only experimental evidence so far relating the SNP genotype with the expression of the lincRNA points to an increase of lincRNA serum levels in individuals harboring the risk allele [93]. The disease associated SNP rs2850711 is transmitted in a linkage disequilibrium (LD) block composed of 16 intronic SNPs, that will probably not have effects

8  Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs

on the secondary structure of the lincRNA. The majority of these intronic SNPs likely alter binding site motifs which could regulate the stability or splicing of the lincRNA leading to different expression levels. However, further functional and mechanistic studies assessing the contribution of the SNP alleles in the regulation of the lincRNA are necessary to understand how this RA and atherosclerosis associated SNP in LINC00305 influences the inflammatory environment characteristic of these disorders.

8.7

Discussion

Genome wide association studies (GWAS) have contributed significantly to our knowledge about the genetic variation that predisposes individuals to complex immune and inflammatory disorders. Initial approaches to decipher the function of GWAS associated variants in the pathogenesis of such complex diseases were mainly focused on coding genes with already described immune function. Many such genes were located near, and randomly harbored, disease-associated SNPs. These approaches helped in the identification and characterization of few novel genes involved in the pathogenesis of complex immune disorders, as it was the case of a non-synonymous SNP in the SH2B3 gene, whose function has been related to the development of several immune disorders, such as celiac disease and type 1 diabetes [6, 110]. Other SNPs located near immune genes have been implicated in the their regulation by their location in gene promoters [32] or in immune cell enhancers [25] thereby affecting their expression of linked genes. However, a majority of the GWAS associated SNPs are located in intergenic regions, making it difficult to assign a molecular function to the genomic variations. Advances in our knowledge of the human genome have brought to light the importance of the long noncoding RNA molecules (lncRNAs) that do not translate into proteins and exert regulatory functions on other coding genes. The expression of these lncRNAs is cell-type specific and they function by interacting with other RNA,

155

DNA or protein molecules affecting several layers of genetic regulation, such as chromatin modification, mRNA biogenesis or protein synthesis. In the past few years, an increasing number of lncRNAs have been linked to the regulation of immune and inflammatory pathways. Some lncRNAs, e.g. lincRNA-Cox2, are specifically expressed in immune cells and play important roles in controlling inflammatory gene expression [7]. Other lncRNAs have been associated with either imparting a protective role, or greater risk, in inflammation or autoimmunity [9, 65]. Moreover, different lncRNAs, such as Morrbid in Hypereosinophilic Syndrome, are significantly up-regulated in patients, suggesting a role of these molecules in disease pathogenesis [48]. Thus, the involvement of lncRNAs in immune disease development is gaining increasing interest, with different lncRNAs already described to be associated with major immune-mediated diseases such as, type 1 diabetes or multiple sclerosis [31, 39]. A number of the complex disease-associated SNPs lie within lncRNAs and modify their sequences which lead to altered secondary structure or altered expression, thereby affecting their regulatory capacity. Therefore, a new field of study focused on uncovering the functional implications of immune disease associated SNPs in lncRNA regulation has been established. Expression and co-expression analyses of the SNP harboring lncRNAs in disease individuals or mouse models are generally carried out in order to elucidate their implication in the disease. These approaches have been helpful to identify disease-related lncRNAs but further molecular characterization of the lncRNA itself, together with functional studies evaluating the contribution of each SNP allele to lncRNA function are generally lacking [40, 82]. As the functional implication of disease associated SNPs on the regulation of coding genes has not always been straightforward, it is likely that the SNPs within and near lncRNAs may also exert complex, indirect effects, including expression of the lncRNA itself, their splicing, their secondary structure or their ability to interact with binding partners. Therefore, it is unlikely

156

A. Castellanos-Rubio and S. Ghosh

that there will ever be a simple way to determine 9. Castellanos-Rubio A, Kratchmarov R, Sebastian M et al (2017) Cytoplasmic form of Carlr lncRNA the functional consequences of such lncRNA-­ facilitates inflammatory gene expression upon associated SNPs. However, understanding how NF-κB activation. J Immunol 199:581–588. https:// doi.org/10.4049/jimmunol.1700023 immune disease associated SNPs within lncRNAs regulate their function and deciphering the 10. Chao T-C, Zhang Q, Li Z et  al (2019) The long noncoding RNA HEAL regulates HIV-1 replication involvement of such lncRNAs in the pathogenethrough epigenetic regulation of the HIV-1 promoter. sis of complex immune-related disorders will MBio 10:e02016–e02019. https://doi.org/10.1128/ mBio.02016-­19 likely continue to be a fruitful approach that will open up the possibility of developing novel RNA-­ 11. Chen YG, Satpathy AT, Chang HY (2017) Gene regulation in the immune system by long noncodbased therapies in the future. ing RNAs. Nat Immunol 18:962–972. https://doi. Acknowledgments The lncRNA work in the author’s laboratories has been funded by the NIH (R01DK102180 to SG) and by the Spanish Ministry of Science Innovation and Universities (PGC2018-097573-A-I00 to ACR).

References 1. Aguilo F, Di Cecilia S, Walsh MJ (2016) Long non-coding RNA ANRIL and polycomb in human cancers and cardiovascular disease. Curr Top Microbiol Immunol 394:29–39. https://doi. org/10.1007/82_2015_455 2. Altshuler D, Daly MJ, Lander ES (2008) Genetic mapping in human disease. Science (80- ) 322:881– 888. https://doi.org/10.1126/science.1156409 3. Aslam MM, John P, Fan K-H et al (2019) Exploration of shared genetic susceptibility loci between type 1 diabetes and rheumatoid arthritis in the Pakistani population. BMC Res Notes 12:544. https://doi. org/10.1186/s13104-­019-­4590-­8 4. Berger EA, Murphy PM, Farber JM (1999) CHEMOKINE RECEPTORS AS HIV-1 CORECEPTORS: roles in viral entry, tropism, and disease. Annu Rev Immunol 17:657–700. https:// doi.org/10.1146/annurev.immunol.17.1.657 5. Burton PR, Clayton DG, Cardon LR et  al (2007) Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447:661–678. https://doi.org/10.1038/ nature05911 6. Camargo A, Perry DJ, Brusko TM, Wallet MA (2018) SH2B3 (lnk) regulates type 1 diabetes immune phenotypes. Diabetes 67:224-LB. https:// doi.org/10.2337/db18-­224-­LB 7. Carpenter S, Aiello D, Atianand MK et  al (2013) A long noncoding RNA mediates both activation and repression of immune response genes. Science 341:789–792. https://doi.org/10.1126/ science.1240925 8. Castellanos-Rubio A, Fernandez-Jimenez N, Kratchmarov R et al (2016) A long noncoding RNA associated with susceptibility to celiac disease. Science (80- ) 352:91–95. https://doi.org/10.1126/ science.aad0467

org/10.1038/ni.3771 12. Chowdhury IH, Narra HP, Sahni A et  al (2017) Expression profiling of long noncoding RNA splice variants in human microvascular endothelial cells: lipopolysaccharide effects in vitro. Mediat Inflamm 2017:1–18. https://doi.org/10.1155/2017/3427461 13. Collier SP, Collins PL, Williams CL et  al (2012) Cutting edge: influence of Tmevpg1, a long intergenic noncoding RNA, on the expression of Ifng by Th1 cells. J Immunol 189:2084–2088. https://doi. org/10.4049/jimmunol.1200774 14. Collier SP, Henderson MA, Tossberg JT, Aune TM (2014) Regulation of the Th1 genomic locus from Ifng through Tmevpg1 by T-bet. J Immunol 193:3959–3965. https://doi.org/10.4049/ jimmunol.1401099 15. Concannon P, Erlich HA, Julier C et al (2005) Type 1 diabetes: evidence for susceptibility loci from four genome-wide linkage scans in 1,435 multiplex families. Diabetes 54:2995–3001. https://doi. org/10.2337/diabetes.54.10.2995 16. Cotsapas C, Hafler DA (2013) Immune-mediated disease genetics: the shared basis of pathogenesis. Trends Immunol 34:22–26. https://doi. org/10.1016/J.IT.2012.09.001 17. d’Ettorre G, Paiardini M, Ceccarelli G et al (2011) HIV-associated immune activation: from bench to bedside. AIDS Res Hum Retrovir 27:355–364. https://doi.org/10.1089/aid.2010.0342 18. Dedmon LE (2020) The genetics of rheumatoid arthritis. Rheumatology 59(10):2661–2670. https:// doi.org/10.1093/rheumatology/keaa232 19. Diaz-Gallo L-M, Martin J (2012) Common genes in autoimmune diseases: a link between immune-­ mediated diseases. Expert Rev Clin Immunol 8:107– 109. https://doi.org/10.1586/eci.11.90 20. Eckburg PB, Relman DA (2007) The role of microbes in Crohn’s disease. Clin Infect Dis 44:256– 262. https://doi.org/10.1086/510385 21. Eizirik DL, Colli ML, Ortis F (2009) The role of inflammation in insulitis and beta-cell loss in type 1 diabetes. Nat Rev Endocrinol 5:219–226. https://doi. org/10.1038/nrendo.2009.21 22. Farh KK-H, Marson A, Zhu J et al (2015) Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature 518:337–343. https://doi. org/10.1038/nature13835

8  Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs 23. Fico A, Fiorenzano A, Pascale E et al (2019) Long non-coding RNA in stem cell pluripotency and lineage commitment: functions and evolutionary conservation. Cell Mol Life Sci 76:1459–1471. https:// doi.org/10.1007/s00018-­018-­3000-­z 24. Finucane HK, Reshef YA, Anttila V et  al (2018) Heritability enrichment of specifically expressed genes identifies disease-relevant tissues and cell types. Nat Genet 50:621–629. https://doi. org/10.1038/s41588-­018-­0081-­4 25. Freudenberg J, Gregersen P, Li W (2015) Enrichment of genetic variants for rheumatoid arthritis within T-cell and NK-cell enhancer regions. Mol Med 21:180–184. https://doi.org/10.2119/ molmed.2014.00252 26. Fusunyan RD, Sanderson IR (1998) Inflammatory bowel disease. In: Encyclopedia of immunology. Elsevier, pp 1375–1381 27. Ghattas A, Griffiths HR, Devitt A et  al (2013) Monocytes in coronary artery disease and atherosclerosis: where are we now? J Am Coll Cardiol 62:1541–1551. https://doi.org/10.1016/J. JACC.2013.07.043 28. Gomez JA, Wapinski OL, Yang YW et al (2013) The NeST long ncRNA controls microbial susceptibility and epigenetic activation of the interferon-γ locus. Cell 152(4):743–754. https://doi.org/10.1016/j. cell.2013.01.015 29. Gong J, Liu W, Zhang J et al (2015) LncRNASNP: a database of SNPs in lncRNAs and their potential functions in human and mouse. Nucleic Acids Res 43:D181–D186. https://doi.org/10.1093/nar/ gku1000 30. Gonzalez E, Bamshad M, Sato N et al (1999) Race-­ specific HIV-1 disease-modifying effects associated with CCR5 haplotypes. Proc Natl Acad Sci U S A 96:12004–12009. https://doi.org/10.1073/ pnas.96.21.12004 31. Gonzalez-Moro I, Olazagoitia-Garmendia A, Colli ML et al (2020) The T1D-associated lncRNA Lnc13 modulates human pancreatic β cell inflammation by allele-specific stabilization of STAT1 mRNA.  Proc Natl Acad Sci U S A 117:9022–9031. https://doi. org/10.1073/pnas.1914353117 32. Guthridge JM, Lu R, Sun H et al (2014) Two functional lupus-associated BLK promoter variants control cell-type- and developmental-stage-specific transcription. Am J Hum Genet 94:586–598. https:// doi.org/10.1016/j.ajhg.2014.03.008 33. Gutierrez-Achury J, Zhernakova A, Pulit SL et  al (2015) Fine mapping in the MHC region accounts for 18% additional genetic risk for celiac disease. Nat Genet 47:577–578. https://doi.org/10.1038/ ng.3268 34. Halvorsen M, Martin JS, Broadaway S, Laederach A (2010) Disease-associated mutations that alter the RNA structural ensemble. PLoS Genet 6:e1001074. https://doi.org/10.1371/journal.pgen.1001074 35. Han Z, Xue W, Tao L et  al (2020) Genome-wide identification and analysis of the eQTL lncRNAs

157

in multiple sclerosis based on RNA-seq data. Brief Bioinform 21:1023–1037. https://doi.org/10.1093/ bib/bbz036 36. Hangauer MJ, Vaughn IW, McManus MT (2013) Pervasive transcription of the human genome produces thousands of previously unidentified long intergenic noncoding RNAs. PLoS Genet 9:e1003569. https://doi.org/10.1371/journal.pgen.1003569 37. Heap GA, Trynka G, Jansen RC et  al (2009) Complex nature of SNP genotype effects on gene expression in primary human leucocytes. BMC Med Genet 2:1. https://doi.org/10.1186/1755-­8794-­2-­1 38. Hindorff LA, Sethupathy P, Junkins HA et al (2009) Potential etiologic and functional implications of genome-wide association loci for human diseases and traits. Proc Natl Acad Sci U S A 106:9362–9367. https://doi.org/10.1073/pnas.0903103106 39. Hosseini A, Teimuri S, Ehsani M et  al (2019) LncRNAs associated with multiple sclerosis expressed in the Th1 cell lineage. J Cell Physiol 234:22153–22162. https://doi.org/10.1002/ jcp.28779 40. Hrdlickova B, Kumar V, Kanduri K et  al (2014) Expression profiles of long non-coding RNAs located in autoimmune disease-associated regions reveal immune cell-type specificity. Genome Med 6:88. https://doi.org/10.1186/s13073-­014-­0088-­0 41. Hu X, Kim H, Stahl E et al (2011) Integrating autoimmune risk loci with gene-expression data identifies specific pathogenic immune cell subsets. Am J Hum Genet 89:496–506. https://doi.org/10.1016/J. AJHG.2011.09.002 42. Hunt KA, Zhernakova A, Turner G et  al (2008) Newly identified genetic risk variants for celiac disease related to the immune response. Nat Genet 40:395–402. https://doi.org/10.1038/ng.102 43. Husby S, Koletzko S, Korponay-Szabó IR et al (2012) European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease. J Pediatr Gastroenterol Nutr 54:136–160. https://doi.org/10.1097/ mpg.0b013e31821a23d0 44. Jones AN, Pisignano G, Pavelitz T et al (2020) An evolutionarily-conserved RNA structure in the functional core of the lincRNA Cyrano. RNA 26(9):1234– 1246. https://doi.org/10.1261/rna.076117.120 45. Jones KL, Maguire JJ, Davenport AP (2011) Chemokine receptor CCR5: from AIDS to atherosclerosis. Br J Pharmacol 162:1453–1469. https:// doi.org/10.1111/j.1476-­5381.2010.01147.x 46. Kent WJ, Sugnet CW, Furey TS et  al (2002) The human genome browser at UCSC.  Genome Res 12:996–1006. https://doi.org/10.1101/gr.229102 47. Kiani AK, Jahngir S, John P et al (2015) Genetic link of type 1 diabetes susceptibility loci with rheumatoid arthritis in Pakistani patients. Immunogenetics 67:277–282. https://doi.org/10.1007/ s00251-­015-­0839-­0 48. Kotzin JJ, Spencer SP, McCright SJ et al (2016) The long non-coding RNA Morrbid regulates Bim and

158 short-lived myeloid cell lifespan. Nature 537:239– 243. https://doi.org/10.1038/nature19346 49. Kulkarni S, Lied A, Kulkarni V et  al (2019) CCR5AS lncRNA variation differentially regulates CCR5, influencing HIV disease outcome. Nat Immunol 20:824–834. https://doi.org/10.1038/ s41590-­019-­0406-­1 50. Kumar V, Westra H-J, Karjalainen J et  al (2013) Human disease-associated genetic variation impacts large intergenic non-coding RNA expression. PLoS Genet 9:e1003201. https://doi.org/10.1371/journal. pgen.1003201 51. Li J-Q, Hu S-Y, Wang Z-Y et al (2016) Long non-­ coding RNA MEG3 inhibits microRNA-125a-5p expression and induces immune imbalance of Treg/Th17  in immune thrombocytopenic purpura. Biomed Pharmacother 83:905–911. https://doi. org/10.1016/J.BIOPHA.2016.07.057 52. Li R, Fang L, Pu Q et al (2018) MEG3-4 is a miRNA decoy that regulates IL-1β abundance to initiate and then limit inflammation to prevent sepsis during lung infection. Sci Signal 11:eaao2387. https://doi. org/10.1126/scisignal.aao2387 53. Li X, Wu Z, Fu X, Han W (2014) LncRNAs: insights into their function and mechanics in underlying disorders. Mutat Res Rev Mutat Res 762:1–21. https:// doi.org/10.1016/j.mrrev.2014.04.002 54. Li X, Yu M, Han L et al (2019) LINC00305 represses miR-124 expression to trigger inflammatory insults in the presence of lipopolysaccharide. Artif Cells Nanomed Biotechnol 47:2352–2360. https://doi.org/ 10.1080/21691401.2019.1624559 55. Libby P, Ridker PM, Hansson GK (2011) Progress and challenges in translating the biology of atherosclerosis. Nature 473:317–325. https://doi. org/10.1038/nature10146 56. Liu R, Paxton WA, Choe S et al (1996) Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiply-exposed individuals to HIV-1 infection. Cell 86:367–377. https://doi.org/10.1016/ S0092-­8674(00)80110-­5 57. Liu Y, Yang L, Xu Q et  al (2019) Long noncoding RNA MEG3 regulates rheumatoid arthritis by ­targeting NLRC5. J Cell Physiol 234:14270–14284. https://doi.org/10.1002/jcp.28126 58. Lonsdale J, Thomas J, Salvatore M et al (2013) The genotype-tissue expression (GTEx) project. Nat Genet 45:580–585. https://doi.org/10.1038/ng.2653 59. Lu X, Ding Y, Bai Y et al (2020) Detection of allosteric effects of lncRNA secondary structures altered by SNPs in human diseases. Front Cell Dev Biol 8:242. https://doi.org/10.3389/fcell.2020.00242 60. McDowell I, Pai A, Guo C et al (2016) Many long intergenic non-coding RNAs distally regulate mRNA gene expression levels. bioRxiv 044719. https://doi. org/10.1101/044719 61. McLaren PJ, Coulonges C, Bartha I et  al (2015) Polymorphisms of large effect explain the majority of the host genetic contribution to variation of HIV-1

A. Castellanos-Rubio and S. Ghosh virus load. Proc Natl Acad Sci U S A 112:14658– 14663. https://doi.org/10.1073/pnas.1514867112 62. Melnikova VO, Mourad-Zeidan AA, Lev DC, Bar-­ Eli M (2006) Platelet-activating factor mediates MMP-2 expression and activation via phosphorylation of cAMP-response element-binding protein and contributes to melanoma metastasis. J Biol Chem 281:2911–2922. https://doi.org/10.1074/jbc. M508683200 63. Mencarelli A, Cipriani S, Francisci D et  al (2016) Highly specific blockade of CCR5 inhibits leukocyte trafficking and reduces mucosal inflammation in murine colitis. Sci Rep 6:30802. https://doi. org/10.1038/srep30802 64. Miao Y-R, Liu W, Zhang Q, Guo A-Y (2018) lncRNASNP2: an updated database of functional SNPs and mutations in human and mouse lncRNAs. Nucleic Acids Res 46:D276–D280. https://doi. org/10.1093/nar/gkx1004 65. Mowel WK, Kotzin JJ, McCright SJ et  al (2018) Control of immune cell homeostasis and function by lncRNAs. Trends Immunol 39(1):55–69 66. Mummidi S, Bamshad M, Ahuja SS et  al (2000) Evolution of human and non-human primate CC chemokine receptor 5 gene and mRNA.  Potential roles for haplotype and mRNA diversity, differential haplotype-specific transcriptional activity, and altered transcription factor binding to polymorphic nucleotides in the pathogenesis of HIV-1 and simian immunodeficiency virus. J Biol Chem 275:18946– 18961. https://doi.org/10.1074/jbc.M000169200 67. Myhr CB, Hulme MA, Wasserfall CH et  al (2013) The autoimmune disease-associated SNP rs917997 of IL18RAP controls IFNγ production by PBMC. J Autoimmun 44:8–12. https://doi.org/10.1016/J. JAUT.2013.06.001 68. Ning S, Yue M, Wang P et al (2017) LincSNP 2.0: an updated database for linking disease-associated SNPs to human long non-coding RNAs and their TFBSs. Nucleic Acids Res 45:D74–D78. https://doi. org/10.1093/nar/gkw945 69. Noble A, Durant L, Hoyles L et al (2020) Deficient resident memory T cell and CD8 T cell response to commensals in inflammatory bowel disease. J Crohns Colitis 14:525–537. https://doi.org/10.1093/ ecco-­jcc/jjz175 70. Op de Beeck A, Eizirik DL (2016) Viral infections in type 1 diabetes mellitus--why the β cells? Nat Rev Endocrinol 12:263–273. https://doi.org/10.1038/ nrendo.2016.30 71. Owens GK, Kumar MS, Wamhoff BR (2004) Molecular regulation of vascular smooth muscle cell differentiation in development and disease. Physiol Rev 84:767–801. https://doi.org/10.1152/ physrev.00041.2003 72. Padua D, Mahurkar-Joshi S, Law IKM et al (2016) A long noncoding RNA signature for ulcerative colitis identifies IFNG-AS1 as an enhancer of inflammation. Am J Physiol 311:G446–G457. https://doi. org/10.1152/ajpgi.00212.2016

8  Functional Implications of Intergenic GWAS SNPs in Immune-Related LncRNAs 73. Parkes M, Cortes A, van Heel DA, Brown MA (2013) Genetic insights into common pathways and complex relationships among immune-mediated diseases. Nat Rev Genet 14:661–673. https://doi. org/10.1038/nrg3502 74. Petermann F, Pękowska A, Johnson CA et al (2019) The magnitude of IFN-γ responses is fine-tuned by DNA architecture and the non-coding transcript of Ifng-as1. Mol Cell 75:1229–1242.e5. https://doi. org/10.1016/J.MOLCEL.2019.06.025 75. Plaza-Izurieta L, Castellanos-Rubio A, Irastorza I et  al (2011) Revisiting genome wide association studies (GWAS) in coeliac disease: replication study in Spanish population and expression analysis of candidate genes. J Med Genet 48:493–496. https:// doi.org/10.1136/jmg.2011.089714 76. Poggi A, Benelli R, Venè R et al (2019) Human gut-­ associated natural killer cells in health and disease. Front Immunol 10:961. https://doi.org/10.3389/ fimmu.2019.00961 77. Popadin K, Gutierrez-Arcelus M, Dermitzakis ET, Antonarakis SE (2013) Genetic and epigenetic regulation of human lincRNA gene expression. Am J Hum Genet 93:1015–1026. https://doi.org/10.1016/j. ajhg.2013.10.022 78. Postler TS, Pantry SN, Desrosiers RC, Ghosh S (2017) Identification and characterization of a long non-coding RNA up-regulated during HIV-1 infection. Virology 511:30–39. https://doi.org/10.1016/j. virol.2017.08.006 79. Qiu Y, Wu Y, Lin M et  al (2019) LncRNA-­ MEG3 functions as a competing endogenous RNA to regulate Treg/Th17 balance in patients with asthma by targeting microRNA-17/RORγt. Biomed Pharmacother 111:386–394. https://doi. org/10.1016/J.BIOPHA.2018.12.080 80. Ren C, An G, Zhao C et  al (2018) Lnc2Catlas: an atlas of long noncoding RNAs associated with risk of cancers. Sci Rep 8:1909. https://doi.org/10.1038/ s41598-­018-­20232-­4 81. Ricaño-Ponce I, Wijmenga C (2013) Mapping of immune-mediated disease genes. Annu Rev Genomics Hum Genet 14:325–353. https://doi. org/10.1146/annurev-­genom-­091212-­153450 82. Ricaño-Ponce I, Zhernakova DV, Deelen P et  al (2016) Refined mapping of autoimmune disease associated genetic variants with gene expression suggests an important role for non-coding RNAs. J Autoimmun 68:62–74. https://doi.org/10.1016/J. JAUT.2016.01.002 83. Rodríguez-Rodríguez L, López-Mejías R, García-Bermúdez M et  al (2012) Genetic markers of cardiovascular disease in rheumatoid arthritis. Mediat Inflamm 2012:1–14. https://doi. org/10.1155/2012/574817 84. Samson M, Libert F, Doranz BJ et  al (1996) Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene. Nature 382:722–725. https://doi. org/10.1038/382722a0

159

85. Santos S, Obukhov Y, Nekhai S et al (2012) Virus-­ producing cells determine the host protein profiles of HIV-1 virion cores. Retrovirology 9:65. https:// doi.org/10.1186/1742-­4690-­9-­65 86. Smolen JS, Aletaha D, McInnes IB (2016) Rheumatoid arthritis. Lancet 388:2023–2038. https://doi.org/10.1016/S0140-­6736(16)30173-­8 87. Sparber P, Filatova A, Khantemirova M, Skoblov M (2019) The role of long non-coding RNAs in the pathogenesis of hereditary diseases. BMC Med Genet 12:42. https://doi.org/10.1186/s12920-­019-­0487-­6 88. Stein N, Berhani O, Charpak-Amikam Y et al (2019) IFNG-AS1 enhances interferon gamma production in human natural killer cells. iScience 11:466–473. https://doi.org/10.1016/j.isci.2018.12.034 89. Tak YG, Farnham PJ (2015) Making sense of GWAS: using epigenomics and genome engineering to understand the functional relevance of SNPs in non-coding regions of the human genome. Epigenetics Chromatin 8:57. https://doi.org/10.1186/ s13072-­015-­0050-­4 90. Tan JY, Smith AAT, Ferreira da Silva M et al (2017) cis-Acting complex-trait-associated lincRNA expression correlates with modulation of chromosomal architecture. Cell Rep 18:2280–2288. https:// doi.org/10.1016/J.CELREP.2017.02.009 91. Törn C, Hadley D, Lee H-S et al (2015) Role of type 1 diabetes-associated SNPs on risk of autoantibody positivity in the TEDDY study. Diabetes 64:1818– 1829. https://doi.org/10.2337/db14-­1497 92. Visscher PM, Wray NR, Zhang Q et  al (2017) 10 years of GWAS discovery: biology, function, and translation. Am J Hum Genet 101:5–22. https://doi. org/10.1016/j.ajhg.2017.06.005 93. Wahba A, Ibrahim M, Mesbah N et  al (2020a) Serum LINC00305 expression and its genetic variant rs2850711 are associated with clinical and laboratory features of rheumatoid arthritis. Br J Biomed Sci 77:142–147. https://doi.org/10.1080/09674845.2 020.1744942 94. Wahba AS, Ibrahim ME, Mesbah NM et al (2020b) Long non-coding RNA MEG3 and its genetic variant rs941576 are associated with rheumatoid arthritis pathogenesis in Egyptian patients. Arch Physiol Biochem:1–8. https://doi.org/10.1080/13813455.20 20.1784951 95. Wall JD, Pritchard JK (2003) Haplotype blocks and linkage disequilibrium in the human genome. Nat Rev Genet 4:587–597. https://doi.org/10.1038/ nrg1123 96. Wallace C, Smyth DJ, Maisuria-Armer M et  al (2010) The imprinted DLK1-MEG3 gene region on chromosome 14q32.2 alters susceptibility to type 1 diabetes. Nat Genet 42:68–71. https://doi. org/10.1038/ng.493 97. Wang J, Liu X, Hao C et al (2019) MEG3 modulates TIGIT expression and CD4 + T cell activation through absorbing miR-23a. Mol Cell Biochem 454:67–76. https://doi.org/10.1007/s11010-­018-­3453-­2

160 98. Wang J, Peng H, Tian J et  al (2016) Upregulation of long noncoding RNA TMEVPG1 enhances T helper type 1 cell response in patients with Sjögren syndrome. Immunol Res 64:489–496. https://doi. org/10.1007/s12026-­015-­8715-­4 99. Wang KC, Chang HY (2011) Molecular mechanisms of long noncoding RNAs. Mol Cell 43(6):904–914 100. Wang Z, Ding L, Zhu J et  al (2018) Long non-­ coding RNA MEG3 mediates high glucose-induced endothelial cell dysfunction. Int J Clin Exp Pathol 11:1088–1100 101. Weinstock JV (2006) Helminths and mucosal immune modulation. Ann N Y Acad Sci 1072:356– 364. https://doi.org/10.1196/annals.1326.033 102. Westra H-J, Martínez-Bonet M, Onengut-Gumuscu S et al (2018) Fine-mapping and functional studies highlight potential causal variants for rheumatoid arthritis and type 1 diabetes. Nat Genet 50:1366– 1374. https://doi.org/10.1038/s41588-­018-­0216-­7 103. Wray NR, Goddard ME, Visscher PM (2007) Prediction of individual genetic risk to disease from genome-wide association studies. Genome Res 17:1520–1528. https://doi.org/10.1101/gr.6665407 104. Wu X, Chen H, Xu H (2015) The genomic landscape of human immune-mediated diseases. J Hum Genet 60:675–681. https://doi.org/10.1038/jhg.2015.99 105. Yang X, Chang Y, Wei W (2020) Emerging role of targeting macrophages in rheumatoid arthritis: focus

A. Castellanos-Rubio and S. Ghosh on polarization, metabolism and apoptosis. Cell Prolif 53:e12854. https://doi.org/10.1111/cpr.12854 106. Yates AD, Achuthan P, Akanni W et  al (2019) Ensembl 2020. Nucleic Acids Res 48:D682–D688. https://doi.org/10.1093/nar/gkz966 107. Zhang B-Y, Jin Z, Zhao Z (2017a) Long intergenic noncoding RNA 00305 sponges miR-136 to regulate the hypoxia induced apoptosis of vascular endothelial cells. Biomed Pharmacother 94:238–243. https:// doi.org/10.1016/J.BIOPHA.2017.07.099 108. Zhang C, Kang K, Li X, Xie B (2015) MicroRNA-136 promotes vascular muscle cell proliferation through the ERK1/2 pathway by targeting PPP2R2A in atherosclerosis. Curr Vasc Pharmacol 13:405–412 109. Zhang D-D, Wang W-T, Xiong J et al (2017b) Long noncoding RNA LINC00305 promotes inflammation by activating the AHRR-NF-κB pathway in human monocytes. Sci Rep 7:46204. https://doi. org/10.1038/srep46204 110. Zhernakova A, Elbers CC, Ferwerda B et al (2010) Evolutionary and functional analysis of celiac risk loci reveals SH2B3 as a protective factor against bacterial infection. Am J Hum Genet 86:970–977. https://doi.org/10.1016/j.ajhg.2010.05.004 111. Zhou Y, Zhong Y, Wang Y et  al (2007) Activation of p53 by MEG3 non-coding RNA.  J Biol Chem 282:24731–24742. https://doi.org/10.1074/jbc. M702029200

9

Long Noncoding RNAs as Therapeutic Targets Jacob B. Pierce, Haoyang Zhou, Viorel Simion, and Mark W. Feinberg

Abstract

Long noncoding RNAs (lncRNAs) have emerged as critical regulators of cellular functions including maintenance of cellular homeostasis as well as the onset and progression of disease. LncRNAs often exhibit cell-, tissue-, and disease-specific expression patterns, making them desirable therapeutic targets. LncRNAs are commonly targeted using oligonucleotide therapeutics, and advances in

J. B. Pierce Department of Medicine, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA Feinberg School of Medicine, Northwestern University, Chicago, IL, USA H. Zhou Department of Medicine, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA Department of Cardiology, The Second Xiangya Hospital of Central South University, Changsha, Hunan, China V. Simion · M. W. Feinberg (*) Department of Medicine, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA e-mail: [email protected]

oligonucleotide chemistry including C2 ribose sugar modifications such as 2′-fluoro, 2′-O-methyl, and 2-O-methoxyethyl modifications; 2′4′-constrained nucleotides such as locked nucleic acids and constrained 2′-O-ethyl (cEt) nucleotides; and phosphorothioate bonds have dramatically improved efficacy of oligonucleotide therapies. Novel delivery platforms such as viral vectors and nanoparticles have also improved pharmacokinetic properties of oligonucleotides targeting lncRNAs. Accumulating pre-clinical studies have utilized these strategies to therapeutically target lncRNAs and alter progression of many different disease states including Snhg12 and Chast in cardiovascular disease, Mirt2 and HOTTIP in sepsis and autoimmune disease, and Malat1 and HOXB-AS3 in cancer. Emerging oligonucleotide conjugation methods including the use of peptide nucleic acids hold promise to facilitate targeting to specific tissue types. Here, we review recent advances in lncRNA therapeutics and provide examples of how lncRNAs have been successfully targeted in pre-clinical models of disease. Finally, we detail remaining challenges facing the lncRNA field and how advances in delivery platforms and oligonucleotide chemistry might help overcome these barriers to catalyze the translation of pre-clinical studies to successful pharmaceutical development.

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_9

161

162

Keywords

J. B. Pierce et al.

used viral vectors is adenovirus, a naked, double-­ stranded DNA virus with a capacity of carrying Therapeutics · RNA · Delivery platforms · ~8 kb of exogenous transcript [5]. LncRNAs can Oligonucleotides · Cardiovascular disease be both therapeutically inhibited (e.g. with small hairpin RNA [shRNA] or small interfering RNA [siRNA]) or overexpressed (e.g. with lncRNA 9.1 Introduction transcripts) using adenoviral vectors. Transduction using adenoviral vectors is highly Long noncoding RNAs (lncRNAs) are increas- efficient in the majority of cell types of interest, ingly being recognized as critical regulators of resulting in very robust gene expression. the onset and progression of a wide range of dis- However, genetic material delivered via adenovieases. Because of their often tissue- and disease-­ rus does not integrate into the genome and is specific expression patterns, they are attractive expressed only transiently in transduced cells. targets for therapeutic intervention in a variety of Adenovirus is also highly prevalent in the envichronic disease states including cardiovascular ronment, resulting in rapid and robust adaptive disease [1, 2], inflammatory and autoimmune immune responses, including neutralizing antidiseases [3], and cancer [4]. To date, many thera- bodies, in most individuals treated with prevalent peutic interventions targeting lncRNAs have serotypes of adenovirus. For this reason, adenobeen described in pre-clinical in vivo models of viral vectors typically used modified genotypes disease, but there has been little progression into of serotype 5 (Ad5), although many different pharmaceutical development in part due to tech- serotypes including chimeric serotypes have been nical limitations of targeted therapeutic delivery developed for use as vaccines, oncolytics, and and toxicity with first generation inhibitors. gene therapies [6, 7]. Adenovirus therapeutics However, recent advances in nucleic acid thera- can be both replication deficient or competent peutics are poised to revolutionize lncRNA-based depending on the goals of therapy. pharmaceutical interventions. In this chapter, we review advances in viral and non-viral oligonucleotide delivery platforms and small interfering 9.2.2 Adeno-Associated Virus RNA (siRNA) and antisense oligonucleotide (ASOs) chemistry that can significantly improve Adeno-associated virus (AAV) is a naked, single-­ the efficacy and limit the off-target effects of stranded DNA virus with a capacity of carrying lncRNA therapeutics (Fig.  9.1). We highlight ~5  kb exogenous transcript [5]. Despite its examples of lncRNAs that have been therapeuti- smaller cloning capacity, AAV has several advancally exploited in pre-clinical models of cardio- tages over adenovirus with regard to potential vascular disease (CVD), inflammatory disease, gene therapy applications. Perhaps most notably, and cancer and detail some of the existing chal- AAV can induce long-term expression changes in lenges in translating lncRNA therapeutics from target genes without integration into the host the bench to the bedside. genome [5]. AAV can also be targeted to specific tissues in vivo. Eleven different serotypes of AAV naturally exist, each with different tropisms for particular tissues [8]. Engineering of capsid pro9.2 Viral Delivery Platforms teins also allows for even more targeted delivery for RNA Therapeutics of viral particles. AAV-based gene therapies are highly effica9.2.1 Adenovirus cious with several being FDA approved. For Recombinant viral delivery platforms are emerg- example, AVXS-101 is FDA-approved for the ing as effective and safe therapeutic gene-therapy treatment of spinal muscle atrophy, a monogenic delivery strategies. One of the most commonly disease of SMN1. An AAV9 vector containing

9  Long Noncoding RNAs as Therapeutic Targets

163

Fig. 9.1  Schematic of lncRNA interventions and delivery platforms. Several different interventions designed to both overexpress and knockdown of target lncRNAs have been utilized in pre-clinical models of disease such as siRNA, antisense oligonucleotides (ASOs), peptide nucleic acids (PNAs), and direct delivery of the lncRNA transcript. RNA interventions often utilize both viral (adenovirus, AAV, lentivirus) and nonviral delivery platforms (liposomes, lipid nanoparticles, gold nanoparticles).

Example 1: Chast was packaged in an AAV9 vector and delivered to untreated C57BL/6  N mice via intravenous injection [42]. Example 2: Locked nucleic acid (LNA) gapmers were utilized to target HOXB-AS3 using a lipid nanoparticle vector and delivered to NOD-scid IL2Rgammanull (NSG) mice treated with AML patient leukemic blasts via intravenous and intraperitoneal injection [67]. (Created with Biorender.com)

human SMN1 was delivered in a single dose to infants with spinal muscle atrophy and resulted in prolonged survival and improved motor function [9]. Similar to adenovirus, a major limitation of AAV technology is the induction of both innate and adaptive immune systems upon administration [10]. Additionally, a majority of the population has naturally occurring humoral immunity against AAV; natural exposure to AAV in the environment leads to circulating anti-AAV antibodies which can neutralize the virus [11]. Recent advances in capsid engineering have been able to mitigate many of these limitations imposed by the host immune system [12, 13].

genomic RNA is then reverse transcribed into DNA by viral enzymes. Unlike adenovirus and AAV, lentivirus genomic material subsequently integrates into the genome, resulting in permanent genomic alterations and long-term gene expression. This delivery platform has been successfully exploited therapeutically in the cancer field with the production of chimeric antigen receptor (CAR) T cells [14]. Ex vivo T cells are transduced using lentivirus with synthetic T cell receptor genes targeting particular cell surface receptors to enhance anti-tumor activity. For instance, CAR T cells have been used to target CD19, a B cell surface receptor, in large B-cell lymphoma [15]. Lentiviral delivery platforms are limited by the lack of tissue specificity and concerns regarding long-term side effects and immunogenicity associated with viral genomic integration [5]. Finally, there is a small but theoretical risk for viral reconstitution following systemic administration of replication-defective virus [16].

9.2.3 Lentivirus Lentivirus, a subtype of retroviruses, is an enveloped, single-stranded RNA virus with a capacity of carrying ~8  kb of exogenous transcript [5]. Following cellular entry, the single-stranded

J. B. Pierce et al.

164

9.3

 on-viral Delivery Platforms N for RNA Therapeutics

9.3.1 Liposomes Liposomes are lipid vesicles comprised of a phospholipid bilayer surrounding an aqueous, hydrophobic core. Because of their ease of production, biocompatibility, and chemical versatility, liposomes (and lipid-based delivery platforms more broadly) are attractive delivery platforms for oligonucleotide-based therapeutics. Due to the presence of both hydrophobic (between lipid bilayers) and hydrophilic compartments (within aqueous core), liposomes are an incredibly versatile platform for delivery of many different classes of therapeutics. However, extensive modifications to lipid composition and surface chemistry are often needed in order to optimize the pharmacokinetics of these therapies in vivo [17]. Drug delivery in lysosomes generally involves cellular uptake into cells via the endolysosomal pathway. Thus, a major challenge in delivering therapeutic oligonucleotides to the cytosol is avoiding lysosomal nuclease degradation in a process called endosomal escape. Advances in lipid chemistry have fostered the use of pH and temperature sensitive liposomes, allowing for targeted release of therapeutic cargo from liposomes and improved drug delivery. pH-sensitive lysosomes utilize modified lipid molecules that release liposomal contents upon exposure to the lower pH of lysosomal compartment. Temperature-sensitive lysosomes are typically used concurrent with radiation therapy in the treatment of cancer [17]. Due to their size and high molecular weight, liposomes carrying oligonucleotide therapeutics are at risk for clearance by the reticuloendothelial system, resulting in high levels of splenic accumulation. However, the inclusion of polyethylene glycol (PEG), a synthetic neutrally charged polymer, reduces immune system recognition of liposomes and nanoparticles, resulting in dramatically prolonged half-life in circulation and overall improved pharmacokinetic properties of liposomal formulations [18]. PEG is thought to reduce the binding of opsonins, circulating proteins used

by the reticuloendothelial system to enhance phagocytosis, to nanoparticles, thereby reducing activation of the innate immune system and clearance by the mononuclear phagocytic cells in the liver and spleen [19]. In order to avoid issues with endosomal release, a range of non-­liposomal platforms are being developed to facilitate delivery of non-coding RNAs in a non-endocytic manner.

9.3.2 Nanoparticles A major advance in RNA therapeutic delivery has been the development of nanoparticle-based delivery platforms. Lipid nanoparticles (LNPs) were developed in response to limited stability of oligonucleotides contained within lysosomes and their associated poor pharmacokinetic properties. Several different classes of LNPs exist including lipoplexes or cationic liposomes, stable, nucleic acid-lipid particles, lipopolyplexes, and membrane-­ core nanoparticles [20]. Fundamentally, each of these classes is based upon advances in lipid synthesis in which cationic lipids (rather than anionic phospholipids) are used to form mono- or bilayered vesicles containing the anionic RNAs. In LNPs, anionic RNA molecules are conjugated to carrier molecules including inorganic nanoparticles, polycationic polymers, or the cationic lipids themselves, increasing the stability of RNAs upon systemic delivery and uptake into target cells [18]. Cationic lipids in LNPs increase efficiency of oligonucleotide delivery to target cells by facilitating lysosomal escape due to interactions with anionic lipids in the endolysosomal membranes [21]. Commercially available products such as Lipofectamine® (Thermo Fisher) are cationic lipids that form stable LNPs and are commonly used for in vitro delivery of oligonucleotides. Inorganic nanoparticles have also been developed for the purposes of delivering oligonucleotide therapeutics to target tissues. Recently, spherical nucleic acids (SNAs) have demonstrated promise as an inorganic nanoparticle delivery platform. SNAs are made of two parts: a solid or hollow core nanoparticle surrounded by a

9  Long Noncoding RNAs as Therapeutic Targets

shell of conjugated nucleic acids [22]. While original SNAs were comprised of a gold-based core with thiol-conjugated nucleic acids, advances in SNA chemistry have lead to the development of alternative metal-based (e.g. platinum or silver) or non-metal-based cores (e.g. lipsosomes or proteins). Gold-based SNAs with conjugated antisense oligonucleotides have been successfully employed to knockdown lncRNA Malat1 (metastasis-associated lung adenocarcinoma transcript 1) in a mouse cancer model, reducing lung metastasis and improving mouse survival [23].

9.3.3 Polymers Oligonucleotides can also form electrostatic interactions with inorganic compounds, which has been exploited in the development of cationic polymers. Perhaps one of the most well-known polymers is polyethyleneimine (PEI), an amino-­ containing polymer (either in linear or branched form) [18]. Following uptake via the endolysosomal pathways by target cells, the amino groups in PEI give it large capacity for proton buffering. As a consequence of the strong osmotic gradient, chloride ions and water enter and disrupt the lysosomal compartment and cause release of the oligonucleotide cargo into the cell. Commercially available in vivo-jetPEI® (Polyplus-transfection® SA) has been successfully employed to deliver 2′-O-methylated (discussed below) and 5′ capped lncRNA SNHG12 to the vessel wall via intravenous injection and dramatically reduced atherosclerosis [24]. Dendrimers are branching inorganic polymers that are also commonly used for in vivo delivery of oligonucleotide therapeutics [18].

9.4

Knockdown Strategies

9.4.1 Small Interfering RNA Small interfering RNA (siRNA) is commonly used to knockdown target gene expression, a strategy known as RNA interference (RNAi).

165

SiRNA is comprised of a double-stranded RNA molecule including guide strand and a passenger strand. In RNAi, siRNA is introduced into the cell where the double-stranded RNA molecule associates with endogenous host RNA-induced silencing complex (RISC) proteins. Once bound to RISC, the passenger strand dissociates from the guide strand, which is subsequently able to bind to target RNAs and facilitate RISC-mediated degradation. This strategy is highly effective at limiting host gene expression and is frequently used for the silencing of cytoplasmic lncRNAs. However, because siRNA necessarily involves cytosolic cellular proteins to degrade target RNAs, it is limited in its use for investigating the many lncRNAs that localize to the nucleus. For this reason, antisense oligonucleotides (ASOs) were developed and are now the preferred method of lncRNA silencing both in vitro and in vivo.

9.4.2 Antisense Oligonucleotides Antisense oligonucleotides (ASOs) are single-­ stranded oligonucleotides delivered to cells to induce target RNA degradation. In contrast to siRNA, ASOs bind directly to target RNAs via Watson-Crick base-pairing. The resulting nucleic acid duplexes are recognized by endogenous ribonucleases (RNase), and target RNAs are degraded. Endogenous RNase H, present in both the cytoplasm and the nucleus, selectively binds DNA/RNA duplexes and degrades the associated RNA.  ASOs frequently take advantage of the DNA/RNA duplex substrate selectivity of endogenous RNase H by employing a chimeric RNA-­ DNA-­RNA structure whereby a complimentary DNA segment is flanked by two short RNA segments. The resulting ASOs are referred to as “gapmers,” resembling an anti-sense RNA oligonucleotide with a DNA “gap” in the middle. These are commonly described according to the lengths of the RNA-DNA-RNA segments. For example, 10 DNA nucleotides flanked on the 5′ and 3′ ends by 5 RNA nucleotides would be described as a “5–10–5” gapmer ASO. A distinct advantage of ASOs is their ability to enter the nucleus and induce degradation of nuclear RNA

J. B. Pierce et al.

166

transcripts. This is particularly important in the study of lncRNAs, as the majority of lncRNA transcripts are retained in the nucleus.

9.5

 ucleic Acid Modifications N in RNA-Based Therapeutics

Advances in oligonucleotide chemistry have led to the development of increasingly potent and nontoxic RNA therapeutics, particularly with ASOs. While chemical modifications to the nucleobase and alterations in oligonucleotide backbone structure have been developed, the most common modifications to oligonucleotide structure occur in the ribose sugar. Generally, the goal of sugar modifications has been to stabilize 3-dimensional conformations of the ribose sugar ring that promote RNA/RNA duplex formation. These modifications are primarily focused on modifications to C2′ of the ribose sugar ring and have dramatically altered the efficacy of target-RNA knockdown by reducing innate immune activation, improving resistance to nuclease degradation, and increasing target binding affinity (Fig.  9.2) [25]. Common chemical modifications include 2′-fluoro (2′-F), 2′-O-methyl (2′-O-Me), and 2′-O-methoxyethyl (2′-O-MOE) modifications. For example, inclisiran, an investigational siRNA targeting proprotein convertase subtilisin/kexin type 9 (PCSK9) that lowers serum cholesterol, includes both 2′-O-Me and 2′-F modifications to the guide strand to improve pharmacokinetics and efficacy [26, 27]. Similarly, IONIS-ANGPTL3-LRx is a 20-mer (5–10–5 gapmer structure) ASO therapeutic targeting Angptl3 mRNA that contains 2′-O-MOE modifications to the 5 flanking RNA nucleotides on both the 5′ and 3′ ends [28]. ASOs also commonly utilize 2′4′-constrained nucleosides to further stabilize favorable ribose sugar conformations. Locked nucleic acids (LNA) are frequently used modifications to gapmers that increase the potency of targeted lncRNA knockdown. The structure of LNAs resembles 2′-O-Me nucleotides in which the methyl group has been covalently bound to C4’ in the ribose sugar ring, forming a methylene

bridge between C2′ and C4′ (Fig. 9.2). However, LNA-based lncRNA knockdown strategies are unlikely to be translated directly into human therapeutics due to the unacceptable hepatotoxicity associated with LNA gapmers [29, 30]. High-­ affinity off-target protein binding unrelated to RNase H1-mediated target RNA degradation is likely responsible for increased hepatocyte apoptosis and resulting transaminase elevations seen in in vivo LNA gapmer studies [31, 32]. Transaminase elevations have also been observed in phase I clinical trials of LNA ASOs, suggesting limited therapeutic window of LNA ASOs [33]. Finally, modifications to the sugar-­ phosphate linkages have also been developed to increase potency and limit susceptibility to endogenous nucleases. ASOs commonly replace phosphodiester bonds with phosphorothioate bonds in which one of the non-bridging oxygens is been replaced with sulfur [34]. Recent research in RNA-based therapeutics has been focused on the reduction in unfavorable off-target protein interactions and subsequent hepatotoxicity. For example, alternative 2′,4′-constrained nucleosides such as constrained 2′-O-­ MOE (cMOE) and 2′-O-ethyl (cEt) nucleotides have also been developed with apparent reductions in hepatotoxicity relative to LNA oligonucleotides [35, 36]. Additionally, targeted modifications to the deoxyribose oligonucleotide gap in ASOs have demonstrated marked reductions in hepatoxicity associated with ASO treatment including methoxypropylphosphonate linkages and the interruption of the continuous deoxynucleotide sequence with 2′-O-MOE ribonucleotide monomers (Fig. 9.2) [31, 32]. Finally, RNA therapeutics can be conjugated to targeted moieties that bind to cell type-specific receptors. In doing so, therapeutic oligonucleotides can be selectively internalized by target cell types, creating a more favorable pharmacodynamic profile and reducing undesirable off-target effects. For example, both inclisiran and IONIS-ANGPTL3-­ LRx target cholesterol synthesis mechanisms in hepatocytes and are bound to hepatocyte-specific triantennary N-acetylglycosamine (GalNAc) moieties [26–28]. This targeting strategy is par-

9  Long Noncoding RNAs as Therapeutic Targets

167

Fig. 9.2  Chemical modifications to RNA therapeutics. Chemical modifications are often incorporated into RNA-­ based therapeutics to improve and optimize their pharmacokinetic properties. Modifications often occur in three

varieties: modifications to the oligonucleotide backbone (highlighted in green), alterations to C2′ of the ribose sugar ring (highlighted in purple), and 2′4′-constrained nucleotides (highlighted in blue)

ticularly attractive for lncRNA-based RNA ­therapeutics due to common cell- and diseasespecific lncRNA expression patterns. Although RNA-­based interventions targeting lncRNAs that utilize these modifications have been described in pre-clinical in vivo studies, there have not been any RNA-based therapeutics targeting lncRNAs to date.

which purines and pyrimidines are easily connected [37]. PNAs are resistant to cleavage by enzymes or chemicals and typically do not degrade within the cell. PNAs are usually incorporated to nanoparticles to provide additional properties such as for imaging [38]. For example, attachment of peptide nucleic acid antimiRs to a peptide with a low pH-induced transmembrane structure (pHLIP) successfully targeted the tumor microenvironment to transport anti-miRs in solid tumors pH ~6 [39]. Analogous paradigms are being developed for PNA-mediated delivery in other pathobiological conditions such as excessive inflammation or metabolic alterations [40, 41]. In theory, conjugation of ASOs to antibodies might provide similar tissue-specific delivery strategies.

9.5.1 P  eptide Nucleic Acid-Based Therapeutics In an attempt to improve non-specific tissue biodistribution and endolysosomal trafficking, novel oligonucleotide conjugation methods, such as peptide nucleic acids (PNAs), are emerging to facilitate targeting ASOs to specific tissue types. Delivery strategies that utilize peptides as ‘zipcodes’ to specific cell types or even intracellular motifs may facilitate cell entry using non-­ endocytic pathways and can be tailored to local microenvironments. PNAs contain a distinct N-2-aminoethylglycine neutral backbone to

9.6

LncRNA Therapeutics in CVD

As highlighted in this chapter, lncRNAs play critical roles in dictating the onset and progression of cardiovascular disease. Therapeutic

J. B. Pierce et al.

168

manipulation of lncRNA expression has been investigated in several pre-clinical in vivo models of cardiovascular disease including atherosclerosis, heart failure, and myocardial infarction. LncRNA Chast (cardiac hypertrophy-associated transcript) was significantly upregulated in cardiomyocytes in transverse aortic constriction (TAC)-operated mice and human heart tissue from patients with aortic stenosis [42]. Mechanistically, Chast reduces expression of Pleckstrin homology domain–containing protein family M member 1 (PLEKHM1) to promote cardiomyocyte hypertrophy. Intravenous injection of AAV9 containing full-length Chast transcript resulted in >10-fold overexpression of Chast in mouse cardiomyocytes. Mouse cardiomyocytes were consequently larger with increased expression of B-type natriuretic peptide, a frequently used biomarker of volume overload and heart failure. Markers of fibrosis were also increased in mice following AAV9-mediated Chast overexpression. Conversely, weekly intraperitoneal injections of LNA gapmer Chast knockdown completely abrogated increased Chast expression following TAC.  When LNA gapmer injections were started concurrently with TAC operation, fibrosis at 5-weeks post-­operation was reduced by ~75%, similar to sham-operated mice. In addition, LNA gapmer Chast-treated mice demonstrated reduced heart weight and pathologic dilation and increased ejection fraction [42] (Table 9.1). Several lncRNAs have been therapeutically manipulated to limit myocardial fibrosis following myocardial infarction (MI). Wisper (Wisp2 super-enhancer–associated RNA) is enriched in cardiac fibroblasts and plays an important role in myocardial fibrosis [43]. One month after ligation of the left anterior descending artery (LAD), Wisper was upregulated 20-fold in the mouse myocardium. After LNA gapmer treatment via intraperitoneal injection, however, Wisper expression was reduced by ~75%, and myocardial fibrosis was reduced from ~13% to ~5% [42]. Echocardiographic measures of cardiac function were also improved following LNA gapmer treatment. LncRNA UIHTC (upregulated in hypothermia treated cardiomyocytes) was also

investigated as a therapeutic target in mice ­following MI. Similar to Wisper, direct intramyocardial injection of AAV9-based UIHTC resulted in reduced fibrosis and improved myocardial function [44]. RNA-sequencing of the aortic intima in Ldlr−/− mice revealed that the lncRNA Snhg12 was significantly downregulated during atherosclerosis progression [24]. Mechanistic studies revealed that Snhg12 positively regulated DNA repair machinery and, in turn, dramatically impacted cellular senescence. Reduced DNA damage in the vascular endothelium as a result of Snhg12 prevented vascular senescence, LDL transcytosis (permeability to LDL), and atherosclerotic lesion formation. Biweekly intravenous injections of LNA gapmer Snhg12 achieved nearly 50% reduction in Snhg12 expression in the intima and increased atherosclerotic lesion size in the aortic sinus by 240%. Lesion burden in the descending aorta was also increased by 170%. Conversely, 5′ capped and 2′-O-Me Snhg12 was directly delivered to the vessel wall via intravenous injection and resulted in over 4-fold overexpression in the aortic intima. Atherosclerotic lesion burden was reduced by 34% and 40% in the aortic sinus and descending aorta, respectively [24]. This series of experiments demonstrated important proof-of-concept that in addition to siRNA and ASOs, full-length lncRNA transcripts can be therapeutically delivered to the vessel wall with important implications for a wide range of vascular disease states. It will be important to determine the therapeutic index and potential toxicity associated with intravenous delivery of lncRNAs, though it is a promising field of future research in lncRNA therapeutics.

9.7

LncRNA Therapeutics in Sepsis

Sepsis, an extreme systemic inflammatory and immune response to infection, is a potentially life-threatening medical emergency. A diverse set of epigenetic reprogramming is known to facilitate this response [45]. LncRNAs play a vital role on innate and adaptive immune response and are

Disease process Atherosclerosis

Cardiac hypertrophy

Cardiac fibrosis

Cardiac fibrosis

Sepsis

Sepsis-induced cardiomyopathy

LncRNA Snhg12

Chast

Wisper

UIHTC

Mirt2

KCNQ1OT1

Saline

Overexpression

Knockdown

Ldlr−/− mice

Untreated C57BL/6 N mice

Transverse aortic constriction-treated C57BL/6 N mice LAD ligation-treated C57BL/6 mice

LPS-treated Sprague-Dawley rats

LPS-treated C57BL/6 mice

LAD ligation Sprague-Dawley rats

AAV9

Knockdown

In vivo model Apoe−/− mice

Overexpression

Adenovirus

Adenovirus

AAV9

Overexpression

Overexpression

Saline

Knockdown

Saline

Vector or Overexpression or delivery method knockdown Overexpression Polymer

Table 9.1  Therapeutic manipulation of select lncRNAs using in vivo models of human disease

LncRNA expression plasmid

Unmodified lncRNA transcript

Unmodified lncRNA transcript

LNA gapmer

LNA gapmer

Unmodified lncRNA transcript

LNA gapmer

5′ capped, 2-O-Me lncRNA transcript

Intervention

Intravenous

Intravenous

Intramyocardial

Intraperitoneal

Intraperitoneal

Intravenous

Intravenous

Route of delivery Intravenous

Phenotype Refs. [24] Decreased DNA damage, cellular senescence, and atherosclerosis Increased DNA damage, cellular senescence, and atherosclerosis [42] Enhanced cardiac hypertrophy, fibrosis, left ventricular dilation Reduced cardiac hypertrophy, fibrosis, left ventricular dilation [43] Reduced cardiac fibrosis, improved echocardiographic measures of myocardial function [44] Reduced cardiac fibrosis, improved echocardiographic measures of myocardial function [48] Improved survival in LPS-induced endotoxemia and reduced systemic inflammation [49] Improved echocardiographic measures of myocardial function and reduced systemic inflammation (continued)

9  Long Noncoding RNAs as Therapeutic Targets 169

Rheumatoid arthritis

Liver regeneration

Acute myeloid leukemia

HOTAIR

LncPHx2

HOXB-AS3

Meg3

Rheumatoid arthritis Rheumatoid arthritis

Knockdown

Knockdown

Partial hepatectomy in Balb/c mice

NOD-scid IL2Rgammanull (NSG) mice treated with AML patient leukemic blasts

Overexpression

Overexpression

Cationic LNPs

Not described

Lentivirus

Lentivirus

Lentivirus

Lentivirus

Knockdown

Knockdown

Lentivirus

Lentivirus

PBS

Overexpression

Knockdown

Knockdown

Vector or Overexpression or delivery method knockdown Knockdown Lentivirus

Sprague-Dawley rats

Adjuvant-induced arthritis in Lewis rats Sprague-Dawley rats

LPS-treated C57BL/6 mice Adjuvant-induced arthritis in Lewis rats

Mouse mammary tumor virus-PyMT

Breast cancer

Sepsis-evoked acute lung injury Rheumatoid arthritis

In vivo model LPS-treated Balb/c mice

Disease process Sepsis

PVT1

HOTTIP

Neat1

LncRNA Malat1

Table 9.1 (continued)

cEt gapmer with phosphorothioate bonds LNA gapmer with phosphorothioate bonds

Unmodified lncRNA transcript

Unmodified lncRNA transcript

shRNA

shRNA

Unmodified lncRNA transcript

cEt gapmer with phosphorothioate bonds shRNA

Intervention shRNA

Intravenous and intraperitoneal

Subcutaneous

Subcutaneous

Subcutaneous

Intravenous

Intravenous

Intravenous

Intratracheal

Subcutaneous

Route of delivery Intravenous

Worsened arthritis severity and higher metalloproteinase expression Improved arthritis severity and lower metalloproteinase expression Limited progression of arthritis Enhanced proliferation and suppressed inflammation in chondrocytes Enhanced proliferation and suppressed inflammation in chondrocytes Enhanced liver regeneration and reduced tranaminitis following hepatectomy Improved survival of mice with AML and reduced proliferative capacity and protein synthesis in AML blasts

Reduced lung injury

Phenotype Reduced skeletal muscle injury and systemic inflammation Reduced tumor burden and number of lung metastases

[67]

[66]

[64]

[63]

[60]

[59]

[53]

Refs. [50, 65]

170 J. B. Pierce et al.

9  Long Noncoding RNAs as Therapeutic Targets

171

regulated at the transcriptional and post-­ Acute lung injury (ALI) and peritonitis are transcriptional level. Therapeutic targeting common manifestations in patients with sepsis. lncRNAs in sepsis or sepsis-associated organ Neat1, a conserved lncRNA, promotes activation dysfunction may provide new avenues for con- of inflammasomes and stabilizes the mature castrolling stage-specific inflammatory responses pase-­1 to promote interleukin-1β production in and tissue damage. macrophages. Neat1 knockout mice show Preventing organ failure and improving sur- reduced peritonitis and attenuated lung inflamvival remain major challenges in patients with mation in vivo [52]. Neat1 ablation also restrained sepsis. Mirt2, an abundantly expressed lncRNA LPS-evoked activation of HMGB1-RAGEin macrophages, tethers and attenuates ubiquiti- NF-κB pathway and significantly alleviated lung nation of TRAF6, thus inhibiting activation of tissue injury in a LPS-induced sepsis-ALI model. NF-κB and MAPK pathways and limiting pro- Neat1 knockdown in vivo was achieved using a duction of proinflammatory cytokines [46, 47]. lentivirus containing sh-Neat1 through intratraIn the LPS-induced systemic inflammatory cheal injection 7  days prior to LPS treatment mouse model, Mirt2 recombinant adenovirus [53], an effect highlighting successful extra-­ overexpressed Mirt2  in various tissues. After 6 hepatic tissue delivery. hours of LPS injection, exogenous Mirt2 significantly reduced pro-inflammatory cytokines in the plasma. Only 12.5% of the Ad-EV (negative con- 9.8 LncRNA Therapeutics trol) treated mice survived, whereas Ad-Mirt2 in Autoimmunity administration showed remarkable protection and led to 50% survival with effectively relieving Because lncRNAs are emerging as critical reguLPS-induced endotoxemia and organ dysfunc- lators in both innate and adaptive immunity, they tion, especially the lung and liver [48]. Delivery have also played prominent roles in autoimmune of this lncRNA under more advanced sepsis con- diseases, especially rheumatoid arthritis (RA). ditions will be of interest. RA affects about 1.3 million adults in the United Sepsis-induced cardiomyopathy (SIC) is a States. Worldwide, it is estimated to occur in up common complication of sepsis that is associated to 1 percent of the population [54]. Epigenetic with increased mortality. The lncRNA modification, environmental factors, and suscepKCNQ1OT1 interacts with miR-192-5p and ame- tibility genes alter post-transcriptional regulation liorates the progression of SIC through the miR- during the pre-arthritis phase inducing systemic 192-5p/XIAP axis. Using recombinant adenovirus loss of tolerance and triggering clinical manifescarrying a KCNQ1OT1 overexpression plasmid tations of RA [55]. injected by tail vein in rats 2 weeks prior to LPS RA-FLS (Rheumatoid arthritis fibroblast-like treatment, preserved left ventricular ejection frac- synoviocyte), a major regulator in joint health, tion and reduced TNF-α, IL-1β, and IL-6 expres- abundantly proliferates and invades cartilage and sion in the myocardium [49]. Another lncRNA bone during RA pathology and have been identiMALAT1 was found to interact with EZH2 and fied as a promising target for RA treatment [56, stimulated AKT-1 phosphorylation and decreased 57]. LncRNA HOTTIP, highly expressed and BRCA1 expression and export from the nucleus localized in the nucleus of RA-FLS, induced [50]. Adenoviral BRCA1 therapy is known to methylation of the SFRP1 through binding with reduce systemic inflammatory responses and Dnmt3b and activating the Wnt signaling pathimprove survival in experimental sepsis [51]. way, thereby regulating proliferation, inflammaIntravenous delivery of septic mice with lentivirus tory responses, and apoptosis of RA-FLS. In RA containing sh-MALAT1 resulted in ~50% silenc- of rats [58], silencing HOTTIP by tail vein injecing of MALAT1  in skeletal muscle tissues and tion of a lentivirus containing the pSIH1-H1-­ alleviated skeletal muscle injury and serum levels copGFP vector carrying sh-HOTTIP alleviates of inflammatory factors in 2 days. the progression of RA, while overexpressing

J. B. Pierce et al.

172

HOTTIP by a lentivirus containing LV5-GFP vector carrying oe-HOTTIP exacerbates it [59]. Furthermore, lncRNA PVT1, highly expressed in synovial tissues of RA in rats and predominately localized to the nucleus, also regulates proliferation, inflammation, and apoptosis of RA-FLS through binding to the sirt6 promoter and inhibiting sirt6 transcription via alterations in methylation. Knockdown PVT1 alleviates the progression of RA in rats using tail vein injected sh-PVT1 lentiviral vectors [60]. As the only cell type in the cartilage [61], chondrocytes are an indispensable contributor in RA through releasing multiple enzymes of extracellular matrix degradation, facilitating angiogenesis, enhancing inflammation and immune responses [62]. LncRNA MEG3, downregulated in LPS-treated chondrocytes, alleviates RA through binding to miR-141 and deactivating the AKT/mTOR signaling pathway. In RA rats, using lentiviral vectors carrying MEG3 to induce expression of MEG3 by 3.5 fold enhanced chondrocyte proliferation, while suppressed inflammation in cartilage [63]. Another lncRNA HOTAIR is also significantly reduced in LPS-­ treated chondrocytes. By directly targeting the miR-138, HOTAIR overexpression significantly inhibited the activation of NF-κB in LPS-treated chondrocytes by suppressing p65  in the cell nucleus and decreased inflammation. Finally, RA rats that were treated by subcutaneous injection of LV-HOTAIR (recombinant lentiviruses carrying sequences encoding HOTAIR) promoted chondrocyte proliferation and suppressed inflammation by reducing Th17 cells in peripheral blood mononuclear cells and inactivating NF-κB signaling in vivo [64].

9.9

LncRNA Therapeutics in Cancer

Some of the most innovative lncRNA therapeutics have been pioneered by the field of oncology. LNA gapmers have become the gold-standard for ASO-based knockdown in in vivo studies but were associated with unacceptable hepatotoxicity and transaminitis, limiting their translational

potential. To address this limitation, alternative constrained nucleotide conformations have been utilized therapeutically in pre-clinical cancer models. In one study that used the mouse mammary tumor virus-PyMT model of human luminal B breast cancer, authors investigated the role of Malat1 in breast cancer metastasis. Arun et al. subcutaneously injected mice with 16-mer gapmer targeting Malat1 modified with alternative cEt-constrained nucleotides and phosphorothioate bonds prior to the spontaneous development of mammary tumors that normally occur in this model [65]. Remarkably, cEt gapmer treatment achieved ~60% intra-tumor knockdown of Malat1 with two independent gapmers. In mice treated with Malat1 cEt gapmers, tumor growth was 50% slower than that in mice treated with scrambled ASO controls. Treated mice also had over 50% reduction in metastatic burden in the lungs, the primary site of metastasis in this in vivo model. Tumors in Malat1 ASO-treated mice showed a highly differentiated, cystic/ductular histological phenotype resembling Malat1−/− genomic knockout mice, while tumors in control ASO-treated mice were solid with poorly differentiated cells. In another study, Huang et al. also utilized a 3-10-3 gapmer ASO that contained cEt-­ constrained ribonucleotides and phosphorothioate linkages to study the role of LncPHx2 in liver regeneration following partial hepatectomy [66]. LncPHx2 was upregulated more than 10-fold in the livers of Balb/c mice following partial hepatectomy and limited liver regeneration. Knockdown of LncPHx2 with subcutaneously-­ delivered gapmers resulted in nearly complete silencing of LncPHx2, faster recovery of liver weight, and attenuated transaminitis following partial hepatectomy [66]. Novel delivery platforms have also been utilized in pre-clinical models of cancer. Recently, Papaioannou et  al. optimized a formulation of cationic LNPs for in vivo delivery of LNA gapmers modified with phosphorothioate bonds targeting the lncRNA HOXB-AS3 [67]. HOXB-AS3 is a lncRNA within the HOX region of developmental genes that is not normally expressed in healthy bone marrow. In NPM1-mutated acute myeloid leukemia (AML), HOXB-AS3 is dramat-

9  Long Noncoding RNAs as Therapeutic Targets

ically upregulated in leukemic blasts as a compensatory mechanism to increase ribosome and protein synthesis. As a result of the selective expression of HOXB-AS3 in leukemic blasts, it is an attractive therapeutic target to limit protein synthesis and proliferative capacity in leukemic blasts. Papaioannou et al. used a mouse model of AML in which NOD-scid IL2Rgammanull (NSG) mice were engrafted with AML patient blasts. The resulting leukemic mice were subsequently treated with both intravenous and intraperitoneal injections of cationic LNPs of HOXB-AS3 LNA gapmers. Remarkably, more than half of HOXB-AS3 gapmer-treated mice survived beyond the end of the follow-up compared to 100% mortality in control gapmer-treated mice with a median survival of just 63  days. Near complete silencing of HOXB-AS3 was achieved with the cationic LNP-based gapmer delivery method with limited toxicity, highlighting the therapeutic potential of this delivery platform and knockdown strategy [67].

9.10 Challenges with lncRNA Therapeutics Despite the many pre-clinical studies that have investigated lncRNAs as therapeutic targets, translating these findings to viable therapeutics has proven challenging. First, eukaryotic cells have evolved robust intracellular defense systems for sensing and responding to foreign nucleic acids. Both extra- and intracellular nucleic acid sensing proteins are capable of eliciting innate immune responses and rapidly targeting foreign nucleic acids for degradation [25]. Additionally, nucleic acids in circulation are rapidly metabolized by the liver and kidneys. Even when novel delivery platforms such as lipid-based formulations and nanoparticles are utilized to improve pharmacokinetic properties of oligonucleotide therapeutics, targeting oligonucleotides to diseased tissues remains a challenge to the field of lncRNA therapeutics. While high levels of hepatotoxicity were observed with the first generation of LNA gapmers in both pre-clinical studies and phase I clinical trials [29–32], emerging

173

c­ hemistries such as cEt gapmers exhibit improved toxicity profiles [35, 36]. Use of emerging cellspecific delivery platforms such as PNAs may extend targeting in tissues beyond the liver or in response to pathobiological stimuli. Future studies that modulate lncRNA-expressing loci using CRISPR-Cas9 or other gene editing methods in combination with viral, LNPs, or nanoparticle-­ based delivery platforms will be of interest, especially for hereditary diseases. Finally, unlike other classes of noncoding RNAs such as microRNAs, lncRNAs are often thousands of base pairs long, making packaging and delivery to target tissues difficult. Collectively, these challenges have limited the translation of accumulating studies demonstrating the importance of lncRNAs in many different disease states to meaningful therapeutics and clinical trials.

9.11 Conclusions LncRNAs are critical regulators of important cellular functions and disease phenotypes that have been successfully targeted in numerous pre-­ clinical models of disease including CVD, autoimmune disease, and cancer. Recent advances in oligonucleotide delivery platforms and ASO chemistry seek to overcome the current limitations of oligonucleotide therapeutics including limited pharmacokinetic profiles, off-target protein binding, and unacceptable hepatotoxicity. These advances are poised to foster a revolution in oligonucleotide therapies targeting lncRNAs in many different chronic and devastating diseases.

References 1. Pierce JB, Feinberg MW (2009) Long noncoding RNAs in atherosclerosis and vascular injury: pathobiology, biomarkers, and targets for therapy. Arterioscler Thromb Vasc Biol 40:2002–2017 2. Lu D, Thum T (2019) RNA-based diagnostic and therapeutic strategies for cardiovascular disease. Nat Rev Cardiol 16(11):661–674 3. Wu G-C, Pan H-F, Leng R-X et al (2015) Emerging role of long noncoding RNAs in autoimmune diseases. Autoimmun Rev 14(9):798–805

174 4. Arun G, Diermeier SD, Spector DL (2018) Therapeutic targeting of long non-coding RNAs in cancer. Trends Mol Med 24(3):257–277 5. van Haasteren J, Li J, Scheideler OJ, Murthy N, Schaffer DV (2020) The delivery challenge: fulfilling the promise of therapeutic genome editing. Nat Biotechnol:1–11 6. Wold SMW, Toth K (2013) Adenovirus vectors for gene therapy, vaccination and cancer gene therapy. Curr Gene Ther 13(6):421–433 7. Shaw AR, Suzuki M (2019) Immunology of adenoviral vectors in cancer therapy. Mol Ther Methods Clin Dev 15:418–429 8. Kotterman MA, Schaffer DV (2014) Engineering adeno-associated viruses for clinical gene therapy. Nat Rev Genet 15(7):445–451 9. Mendell JR, Al-Zaidy S, Shell R et al (2017) Single-­ dose gene-replacement therapy for spinal muscular atrophy. N Engl J Med 377(18):1713–1722 10. Ferreira V, Twisk J, Kwikkers K et al (2014) Immune responses to intramuscular administration of alipogene tiparvovec (AAV1-LPLS447X) in a phase II clinical trial of lipoprotein lipase deficiency gene therapy. Hum Gene Ther 25(3):180–188 11. Calcedo R, Vandenberghe LH, Gao G, Lin J, Wilson JM (2009) Worldwide epidemiology of neutralizing antibodies to adeno-associated viruses. J Infect Dis 199(3):381–390 12. Tse LV, Klinc KA, Madigan VJ et al (2017) Structure-­ guided evolution of antigenically distinct adeno-­ associated virus variants for immune evasion. Proc Natl Acad Sci 114(24):E4812–E4821 13. Maheshri N, Koerber JT, Kaspar BK, Schaffer DV (2006) Directed evolution of adeno-associated virus yields enhanced gene delivery vectors. Nat Biotechnol 24(2):198–204 14. Milone MC, O’Doherty U (2018) Clinical use of lentiviral vectors. Leukemia 32(7):1529–1541 15. Neelapu SS, Locke FL, Bartlett NL et  al (2017) Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell lymphoma. N Engl J Med 377(26):2531–2544 16. Waehler R, Russell SJ, Curiel DT (2007) Engineering targeted viral vectors for gene therapy. Nat Rev Genet 8(8):573–587 17. Mallick S, Choi JS (2014) Liposomes: versa tile and biocompatible nanovesicles for efficient biomolecules delivery. J Nanosci Nanotechnol 14(1):755–765 18. Antimisiaris S, Mourtas S, Papadia K (2017) Targeted si-RNA with liposomes and exosomes (extracellular vesicles): how to unlock the potential. Int J Pharm 525(2):293–312 19. Li S-D, Huang L (2010) Stealth nanoparticles: high density but sheddable PEG is a key for tumor targeting. J Control Release 145(3):178 20. Xia Y, Tian J, Chen X (2016) Effect of surface properties on liposomal siRNA delivery. Biomaterials 79:56–68 21. Wang Y, Miao L, Satterlee A, Huang L (2015) Delivery of oligonucleotides with lipid nanoparticles. Adv Drug Deliv Rev 87:68–80

J. B. Pierce et al. 22. Kapadia CH, Melamed JR, Day ES (2018) Spherical nucleic acid nanoparticles: therapeutic potential. BioDrugs 32(4):297–309 23. Gong N, Teng X, Li J, Liang X-J (2018) Antisense oligonucleotide-conjugated nanostructure-targeting lncRNA MALAT1 inhibits cancer metastasis. ACS Appl Mater Interfaces 11(1):37–42 24. Haemmig S, Yang D, Sun X et al (2020) Long noncoding RNA SNHG12 integrates a DNA-PK-mediated DNA damage response and vascular senescence. Sci Transl Med 12:eaaw1868 25. Dowdy SF (2017) Overcoming cellular barriers for RNA therapeutics. Nat Biotechnol 35(3):222 26. Ray KK, Landmesser U, Leiter LA et  al (2017) Inclisiran in patients at high cardiovascular risk with elevated LDL cholesterol. N Engl J Med 376(15):1430–1440 27. Ray KK, Stoekenbroek RM, Kallend D et  al (2019) Effect of 1 or 2 doses of inclisiran on low-density lipoprotein cholesterol levels: one-year follow-up of the ORION-1 randomized clinical trial. JAMA Cardiol 4(11):1067–1075 28. Graham MJ, Lee RG, Brandt TA et  al (2017) Cardiovascular and metabolic effects of ANGPTL3 antisense oligonucleotides. N Engl J Med 377(3):222–232 29. Swayze EE, Siwkowski AM, Wancewicz EV et  al (2007) Antisense oligonucleotides containing locked nucleic acid improve potency but cause significant hepatotoxicity in animals. Nucleic Acids Res 35(2):687–700 30. Burdick AD, Sciabola S, Mantena SR et  al (2014) Sequence motifs associated with hepatotoxicity of locked nucleic acid—modified antisense oligonucleotides. Nucleic Acids Res 42(8):4882–4891 31. Shen W, De Hoyos CL, Migawa MT et  al (2019) Chemical modification of PS-ASO therapeutics reduces cellular protein-binding and improves the therapeutic index. Nat Biotechnol 37(6):640–650 32. Migawa MT, Shen W, Wan WB et  al (2019) Site-­ specific replacement of phosphorothioate with alkyl phosphonate linkages enhances the therapeutic profile of gapmer ASOs by modulating interactions with cellular proteins. Nucleic Acids Res 47(11):5465–5479 33. Bianchini D, Omlin A, Pezaro C et  al (2013) Firstin-­ human phase I study of EZN-4176, a locked nucleic acid antisense oligonucleotide to exon 4 of the androgen receptor mRNA in patients with castration-­ resistant prostate cancer. Br J Cancer 109(10):2579–2586 34. Iannitti T, Cesar Morales-Medina J, Palmieri B (2014) Phosphorothioate oligonucleotides: effectiveness and toxicity. Curr Drug Targets 15(7):663–673 35. Seth PP, Vasquez G, Allerson CA et al (2010) Synthesis and biophysical evaluation of 2′, 4′-constrained 2′ O-methoxyethyl and 2′, 4′-constrained 2′ O-ethyl nucleic acid analogues. J Org Chem 75(5):1569–1581 36. Seth PP, Siwkowski A, Allerson CR et al (2009) Short antisense oligonucleotides with novel 2′− 4′ conformationaly restricted nucleoside analogues show improved potency without increased toxicity in animals. J Med Chem 52(1):10–13

9  Long Noncoding RNAs as Therapeutic Targets 37. Nielsen PE, Egholm M (1999) An introduction to peptide nucleic acid. Curr Issues Mol Biol 1(1–2):89–104 38. Gupta A, Mishra A, Puri N (2017) Peptide nucleic acids: advanced tools for biomedical applications. J Biotechnol 259:148–159 39. Cheng CJ, Bahal R, Babar IA et al (2015) MicroRNA silencing for cancer therapy targeted to the tumour microenvironment. Nature 518(7537):107–110 40. Montagner G, Bezzerri V, Cabrini G et al (2017) An antisense peptide nucleic acid against Pseudomonas aeruginosa inhibiting bacterial-induced inflammatory responses in the cystic fibrosis IB3-1 cellular model system. Int J Biol Macromol 99:492–498 41. Wancewicz EV, Maier MA, Siwkowski AM et  al (2010) Peptide nucleic acids conjugated to short basic peptides show improved pharmacokinetics and antisense activity in adipose tissue. J Med Chem 53(10):3919–3926 42. Viereck J, Kumarswamy R, Foinquinos A et al (2016) Long noncoding RNA Chast promotes cardiac remodeling. Sci Transl Med 8(326):326ra322 43. Micheletti R, Plaisance I, Abraham BJ et al (2017) The long noncoding RNA Wisper controls cardiac fibrosis and remodeling. Sci Transl Med 9(395):eaai9118 44. Zhang J, Yu L, Xu Y et  al (2018) Long noncoding RNA upregulated in hypothermia treated cardiomyocytes protects against myocardial infarction through improving mitochondrial function. Int J Cardiol 266:213–217 45. Hawiger J (2018) Heartfelt sepsis: microvascular injury due to genomic storm. Kardiol Pol 76(8):1203–1216 46. Kishimoto K, Matsumoto K, Ninomiya-Tsuji J (2000) TAK1 mitogen-activated protein kinase kinase kinase is activated by autophosphorylation within its activation loop. J Biol Chem 275(10):7359–7364 47. Lamothe B, Besse A, Campos AD, Webster WK, Wu H, Darnay BG (2007) Site-specific Lys-63-linked tumor necrosis factor receptor-associated factor 6 auto-ubiquitination is a critical determinant of I kappa B kinase activation. J Biol Chem 282(6):4102–4112 48. Du M, Yuan L, Tan X et al (2017) The LPS-inducible lncRNA Mirt2 is a negative regulator of inflammation. Nat Commun 8(1):2049 49. Sun F, Yuan W, Wu H et  al (2020) LncRNA KCNQ1OT1 attenuates sepsis-induced myocardial injury via regulating miR-192-5p/XIAP axis. Exp Biol Med (Maywood):1535370220908041 50. Yong H, Wu G, Chen J et al (2020) lncRNA MALAT1 accelerates skeletal muscle cell apoptosis and inflammatory response in sepsis by decreasing BRCA1 expression by recruiting EZH2. Mol Ther Nucleic Acids 19:97–108 51. Teoh H, Quan A, Creighton AK et al (2013) BRCA1 gene therapy reduces systemic inflammatory response and multiple organ failure and improves survival in experimental sepsis. Gene Ther 20(1):51–61 52. Zhang P, Cao L, Zhou R, Yang X, Wu M (2019) The lncRNA Neat1 promotes activation of inflammasomes in macrophages. Nat Commun 10(1):1495

175 53. Zhou H, Wang X, Zhang B (2020) Depression of lncRNA NEAT1 antagonizes LPS-evoked acute injury and inflammatory response in alveolar epithelial cells via HMGB1-RAGE signaling. Mediat Inflamm 2020:8019467 54. Helmick CG, Felson DT, Lawrence RC et al (2008) Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part I. Arthritis Rheum 58(1):15–25 55. McInnes IB, Schett G (2011) The pathogenesis of rheumatoid arthritis. N Engl J Med 365:2205–2219 56. Bottini N, Firestein GS (2013) Duality of fibroblast-­ like synoviocytes in RA: passive responders and imprinted aggressors. Nat Rev Rheumatol 9(1):24–33 57. Mor A, Abramson SB, Pillinger MH (2005) The fibroblast-like synovial cell in rheumatoid arthritis: a key player in inflammation and joint destruction. Clin Immunol 115(2):118–128 58. Gjertsson I, Laurie KL, Devitt J et al (2009) Tolerance induction using lentiviral gene delivery delays onset and severity of collagen II arthritis. Mol Ther 17(4):632–640 59. Hu X, Tang J, Hu X et  al (2020) Silencing of long non-coding RNA HOTTIP reduces inflammation in rheumatoid arthritis by demethylation of SFRP1. Mol Ther Nucleic Acids 19:468–481 60. Zhang CW, Wu X, Liu D et  al (2019) Long non-­ coding RNA PVT1 knockdown suppresses fibroblast-­ like synoviocyte inflammation and induces apoptosis in rheumatoid arthritis through demethylation of sirt6. J Biol Eng 13:60 61. Adán N, Guzmán-Morales J, Ledesma-Colunga MG et al (2013) Prolactin promotes cartilage survival and attenuates inflammation in inflammatory arthritis. J Clin Invest 123(9):3902–3913 62. Tseng CC, Chen YJ, Chang WA et al (2020) Dual role of chondrocytes in rheumatoid arthritis: the chicken and the egg. Int J Mol Sci 21(3):1071 63. Li G, Liu Y, Meng F et  al (2019) LncRNA MEG3 inhibits rheumatoid arthritis through miR-141 and inactivation of AKT/mTOR signalling pathway. J Cell Mol Med 23(10):7116–7120 64. Zhang HJ, Wei QF, Wang SJ et  al (2017) LncRNA HOTAIR alleviates rheumatoid arthritis by targeting miR-138 and inactivating NF-kappaB pathway. Int Immunopharmacol 50:283–290 65. Arun G, Diermeier S, Akerman M et  al (2016) Differentiation of mammary tumors and reduction in metastasis upon Malat1 lncRNA loss. Genes Dev 30(1):34–51 66. Huang L, Damle SS, Booten S et  al (2015) Partial hepatectomy induced long noncoding RNA inhibits hepatocyte proliferation during liver regeneration. PLoS One 10(7):e0132798 6 7. Papaioannou D, Petri A, Dovey OM et al (2019) The long non-coding RNA HOXB-AS3 regulates ribosomal RNA transcription in NPM1mutated acute myeloid leukemia. Nat Commun 10(1):1–15

Part V Challenges and Future Directions

Challenges and Future Directions for LncRNAs and Inflammation

10

Haley Halasz and Susan Carpenter

Abstract

Keywords

Until somewhat recently, the complexity of the human genome has not been well understood. With advancements in sequencing technology, we now know that nearly the whole genome is transcribed but a very small portion of those transcripts code for proteins. As the research of non-coding genes and transcripts has evolved rapidly in the last decade, it has become clear that many of them serve important biological functions in many previously well-studied cell processes. As the previous chapters in this book have reviewed, the field of noncoding RNA research has provided new insights into specific disease states, especially those driven by inflammation. Understanding the basic mechanisms of non-coding RNAs in the context of inflammation has led to prospective therapeutics that may overcome many of the challenges faced in diagnosing and treating inflammatory diseases. In this final chapter we discuss the current state of the field of non-coding RNA therapeutics and how it may evolve to overcome the short cummings we currently face with diagnosing and treating inflammatory diseases.

Noncoding RNAs · Inflammation · Diagnostics · Therapeutics

H. Halasz · S. Carpenter (*) Department of Molecular, Cell and Developmental Biology, University of California Santa Cruz, Santa Cruz, CA, USA e-mail: [email protected]

10.1 Introduction- Noncoding RNAs and Inflammation A well-adapted immune response is defined by acute activation of immune cells that results in transient release of inflammatory mediators. This rapid response aids in ridding the body of pathogens and repairing tissue damage. This process must be tightly regulated, and the acute response must be adequately resolved once the perceived threat is no longer present. Prolonged activation of inflammatory signaling cascades can lead to chronic expression and secretion of inflammatory mediators. Even at low levels, chronically circulating inflammatory mediators can instigate mal-­ adaptions in a variety of tissues throughout the body [1]. It is now well understood and accepted that unresolved inflammation drives cardiovascular, lung, neurodegenerative, and metabolic diseases, including cancer and stroke [1–4]. All of which the CDC has defined as leading causes of death (https://www.cdc.gov/nchs/fastats/leading-­ causes-­of-­death.htm). The invasion of a pathogen leads to infiltration and activation of innate immune cells such as macrophages and lymphocytes. Pattern

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0_10

179

180

Recognition Receptors (PRRs) present in the membranes of these cells specifically recognize invading pathogens. Engagement of these receptors leads to transcriptional activation and translation of many pro-inflammatory chemokines and cytokines. In a well-functioning immune response these inflammatory signals get resolved once the pathogen has been eliminated. But other factors at the tissue-environmental interface can influence immune function and cause constitutive activation of inflammatory signals in the absence of a pathogen [5]. Constantly elevated levels of cytokines can disrupt the function of many crucial cellular processes, such as mitochondrial function, protein folding, DNA repair, cell differentiation and cell regeneration [2, 3, 5]. While many mechanisms by which chronic inflammation drives diseases have been defined, there is still a great need to better control such aberrant processes. While aging, genetics, and environmental pressures seem to be the main factors that determine immune function, developing a more complete understanding of the complex signaling networks that lead to chronic inflammation is central to preventing and treating inflammatory diseases. As many of the chapters in this book have demonstrated, we can say with growing certainty that non-coding genes play an essential role in inflammatory disease pathology. While the non-­ coding RNA field is still somewhat in its infancy, advancements in screening technologies and transcriptomics have helped us identify hundreds of non-coding genes involved in various cell types and cell processes [6–9]. Recent studies have shown that over 80% of the genome is ­transcribed but less than 3% of those transcripts are coding, revealing newfound complexity of the human genome (refer to Chap. 2). This means that most of the transcriptome is non-coding, and despite increasing evidence that many of these transcripts have specialized functions, especially in diseased states, only roughly 2% of non-­coding transcripts have been characterized. Long non-­ coding RNAs (lncRNAs) account for the largest class of non-coding RNA transcripts and are defined as being greater than 200 nucleotides in length. These genes have now been found to par-

H. Halasz and S. Carpenter

ticipate in diverse regulatory functions ranging from signaling molecules to transcriptional modifiers. The previous chapters in this book have described in detail the mechanisms of many well-­ studied lncRNAs. These lncRNAs have been found to play crucial roles in mediating inflammatory processes underlying autoimmune diseases, cardiovascular diseases, and cancer (refer to Chaps. 4, 5, 6, and 7). As we continue to better understand the basic mechanisms of action by which lncRNAs fine-tune many complex inflammatory signaling pathways and develop better biochemical assays for studying lncRNA structure and function (refer to Chap. 3), we anticipate that lncRNAs will provide novel and desirable therapeutic targets (refer to Chaps. 7 and 9). Collectively the field of non-coding RNA research may offer promising new resolutions to many of the challenges we face in diagnosing and treating chronic inflammatory conditions.

10.2 Current State of Inflammatory Disease Diagnostics and Treatment One particular challenge in diagnosing inflammatory diseases has been the detection of biomarkers that can truly distinguish a diseased state in an individual [10]. Standard diagnostic blood tests can somewhat reliably assess systemic inflammation through circulating levels of interleukins (IL)-6, -8, -10, tumor necrosis factor-α (TNFα), C-reactive protein, and fibrinogen, etc.… [11] and while these markers are useful for flagging general inflammation or risk for a particular disease, they lack the specificity needed to make a clear diagnosis. Further diagnostic methods that may accompany blood tests involve rather invasive practices such as imaging, endoscopy, and tissue biopsies, but are currently necessary to confirm specific disease states. Many inflammatory pathologies require multiple diagnostic approaches and the time to diagnosis, let alone treatment, can be quite long. This can pose a challenge for patients already experiencing debilitating or severe symptoms and may make

10  Challenges and Future Directions for LncRNAs and Inflammation

181

certain treatment options less effective once a 10.3 Clinical Potential for Non-­ diagnosis is determined. coding RNAs Drug development and treatment of diseases driven by inflammation has been somewhat Developments in genomics and transcriptomics, underwhelming. Current pharmacological treat- have led to a much clearer understanding of the ment strategies for inflammatory diseases still individual variation or stratification that exists rely mainly on non-steroidal anti-inflammatory within a certain diseased population. A “one-­ drugs (NSAIDs), disease-modifying anti-­size-­fits-all” approach to treating disease is simrheumatic drugs (DMARDs), glucocorticoids ply not optimal. In 2015 the U.S. government and biologic agents12. While each of these classes launched the Precision Medicine Initiative to of drugs work through slightly different mecha- fund more research and implement a clinical revnisms, they all work by inhibiting immune sig- olution that emphasizes individual, personalized naling globally and ubiquitously. Fortunately, healthcare. This includes a push to start employthese drugs have been effective at relieving ing more advanced sequencing and biomarker symptoms in patients, but inhibition and suppres- detection technologies that yield in-depth inforsion of these important signaling cascades in a mation about healthy and diseased individuals non-specific manner can be very detrimental to [14]. A more personalized approach to medicine immune homeostasis in the long term, and these means considering a person’s genetics, microbidrugs fail to address the underlying biology driv- ome, metabolism, lifestyle, and health history to ing disease [12, 13]. This again highlights the determine the biological cause of the disease and need to better understand the diverse biological tailor treatments to address that cause. The focus roles of many of these targets and the various on precision medicine has dawned a new aspects of physiology they control, to develop approach to drug development and diagnostics, more effective treatments with much better speci- and a push to detect subclinical disease before the ficity and precision. With the mounting evidence onset of overt or debilitating symptoms. that lncRNAs can fine-tune and very tightly conRNA-based diagnostics and therapeutics have trol immune related pathways in a very tissue begun to show great promise in precision medispecific manner, these molecules may act as cine. As mentioned earlier, such a small portion much more promising therapeutic targets (Chap. of the genome is translated into proteins and a 9). majority of proteins have been deemed undrugWith the advancement of the complete human gable [15]. An eruption of evidence for functional genome sequence also came the ability to better non-coding transcripts has really expanded the study gene variants associated with disease. potential for tailored healthcare. The highly tisGenome-wide association studies (GWAS) have sue and context specific expression of non-­coding suggested the relationship between specific sin- transcripts and their deregulation in diseased gle nucleotide polymorphisms (SNPs) and states makes them very attractive drug targets. ­specific diseases. These studies have served as a Growing evidence suggests that many non-­ valuable guide to better understand the biology of coding RNAs are secreted from cells and are promany diseases, with the hope of identifying more tected in extracellular vesicles [16]. Therefore specific drug targets [14]. Although, GWAS stud- the tissue-specific expression profiles of ncRNAs ies have also revealed the fact that most SNPs are often reflected in bodily fluids, making them associated with disease are present in non-coding very attractive as diagnostic markers as well [17, regions, therefore the biology of many of these 18] (Chap. 7). Having this type of diagnostic SNPs are still unknown (Chap. 8). Hence, better power would not only eliminate the need for understanding the actions of non-coding variants invasive tissue biopsies but could also help detect could reveal drug targets better tailored to spe- diseases in their infancy, leading to much better cific diseases. management and prevention.

182

To date, targeting ncRNAs therapeutically has relied on various RNA interference (RNAi) methods. These therapeutic strategies include the use of anti-sense oligos (ASOs), aptamers, small interfering RNAs (siRNAs), and micro-RNAs (miRNAs). These strategies work to modulate transcription or post-transcriptional RNA processing of their targets (Chap. 9). The biggest challenges with these methods have been related to RNA instability, difficult delivery across cell membranes, and the immunogenicity or cell toxicity of exogenous nucleic acids. Advancements in nanotechnology have facilitated stabilization, more efficient delivery, and better safety profile of therapeutic nucleotides, mainly overcoming such challenges [15]. While RNAi therapeutics have had success modulating their target genes, the only RNAi drugs that have been FDA approved so far, target transcripts of coding genes. For instance, Alnylam Pharmaceuticals has developed FDA approved siRNA therapies for treating rare genetic diseases that include hereditary amyloidosis (patisiran), acute hepatic porphyria (givosiran), and primary hyperoxaluria type 1 (lumasiran). Ionis Pharmaceuticals and Serepta Therapeutics have successfully developed ASOs to treat spinal muscular atrphy (nusinersen), hereditary amyloidosis (inotersen), and Duchenne muscular distrophy (eteplirsen) respectively. Currently Haya Therapeutics is dedicated to targeting lncRNAs that drive fibrotic disease and have developed an ASO that successfully targets the lncRNA Wisper in the treatment of cardiac fibrosis, which is now undergoing clinical trial [19]. Perhaps the most promising aspect of the translational potential of lncRNAs is the fact that they can be detected in circulation, and their expression profile often corresponds with incipient disease. With advancements in RNA sequencing technology, the ability to detect circulating RNA in blood and other bodily fluids may offer a truly non-invasive way to not only diagnose presence of disease with high tissue specificity, but also prognose the potential for developing disease (Chaps. 7 and 9). Companies such as Freenome, Grail, and Guardant Health have now developed liquid biopsy tests ready for the clinic

H. Halasz and S. Carpenter

that can detect the earliest stages of cancer through circulating tumor DNA [20]. This technology relies on DNA that gets released into the blood from cancer cell death (tumor shedding), but there is now evidence that live cancer cells secrete RNA through extracellular vesicles [21]. In this sense, circulating RNA may offer a more reliable and even more specific biomarkers of cancer and other inflammatory conditions.

10.4 C  onclusion- Future Insights for Non-Coding RNA Therapeutics and Inflammation While lncRNA research is advancing rapidly, our understanding of this class of molecules is still somewhat limited. Given that such a large portion of the genome produces non-coding transcripts, we must continue to rely on high throughput screening methods that can efficiently identify functional lncRNAs in specific cell types and under certain disease conditions. As we continue to identify functional lncRNAs in the context of inflammation, we will continue to gain a more complete understanding of the basic molecular mechanisms by which lncRNAs modulate inflammatory processes. Defining these mechanisms has begun to highlight the translational potential for lncRNAs yet getting lncRNA therapeutics into clinical trials has been somewhat challenging. This is primarily because non-­ coding RNAs tend to be poorly conserved, making the jump from preclinical studies in animals into human clinical trials challenging. But with the implement of high throughput methods across species, we have the power to quickly identify molecules that may lack sequence conservation but maintain functional conservation. Hopefully this will help focus our basic research efforts on lncRNAs with the utmost clinical potential. Additionally, rapidly evolving transcriptomic technologies have facilitated the possibility to detect circulating lncRNAs. The hope is that these lncRNAs will provide highly specific diagnostic markers for a range of conditions. As mentioned above, being able to detect subtle, chronic

10  Challenges and Future Directions for LncRNAs and Inflammation

183

Quinlan P, Thompson A, Rajapakshe K, Coarfa C, Gunaratne PH, Marchion DC, Magliocco AM, Tsai KY, Flores ER (2020) Pan-cancer analysis reveals TAp63-regulated oncogenic lncRNAs that promote cancer progression through AKT activation. Nat Commun 11(1):5156. https://doi.org/10.1038/ s41467-­020-­18973-­w 8. Liu SJ et  al (2017) CRISPRi-based genome-scale identification of functional long noncoding RNA loci in human cells. Science 355:eaah7111 9. Covarrubias S et  al (2017) CRISPR/Cas-based screening of long non-coding RNAs (lncRNAs) in macrophages with an NF-κB reporter. J Biol Chem 292:20911–20920 10. Huang Y, Gulshan K, Nguyen T, Wu Y (2017) Biomarkers of cardiovascular disease. Dis Markers References 2017:1–2 11. Pearson TA et  al (2003) Markers of inflammation 1. Barnig C, Bezema T, Calder PC, Charloux A, and cardiovascular disease: application to clinical Frossard N, Garssen J, Haworth O, Dilevskaya K, and public health practice: a statement for healthcare Levi-Schaffer F, Lonsdorfer E, Wauben M, Kraneveld professionals from the Centers for Disease Control AD, te Velde AA (2019) Activation of resolution pathand Prevention and the American Heart Association. ways to prevent and fight chronic inflammation: lesCirculation 107:499–511 sons from asthma and inflammatory bowel disease. 12. Li P, Zheng Y, Chen X (2017) Drugs for autoimmune Front Immunol 10:1699. https://doi.org/10.3389/ inflammatory diseases: from small molecule comfimmu.2019.01699 pounds to anti-TNF biologics. Front Pharmacol 8:460 2. Zuo L, Prather ER, Stetskiv M, Garrison DE, Meade 13. Her M, Kavanaugh A (2016) Alterations in immune JR, Peace TI, Zhou T (2019) Inflammaging and oxidafunction with biologic therapies for autoimmune distive stress in human diseases: from molecular mechaease. J Allergy Clin Immunol 137:19–27 nisms to novel treatments. Int J Mol Sci 20(18):4472. 14. Dugger SA, Platt A, Goldstein DB (2018) Drug develhttps://doi.org/10.3390/ijms20184472 opment in the era of precision medicine. Nat Rev 3. Franceschi C, Garagnani P, Parini P, Giuliani C, Drug Discov 17:183–196 Santoro A (2018) Inflammaging: a new immune– 15. Damase TR et al (2021) The limitless future of RNA metabolic viewpoint for age-related diseases. Nat Rev therapeutics. Front Bioeng Biotechnol 9:628137 Endocrinol. 14(10):576–590. https://doi.org/10.1038/ 16. Everaert C et  al (2019) Performance assessment of s41574-­018-­0059-­4 total RNA sequencing of human biofluids and extra 4. Lin T, Liu GA, Perez E, Rainer RD, Febo M, Cruz-­ cellular vesicles. Sci Rep 9:17574 Almeida Y, Ebner NC (2018) Systemic inflamma17. Bolha L, Ravnik-Glavač M, Glavač D (2017) tion mediates age-related cognitive deficits. Front Long noncoding RNAs as biomarkers in canAging Neurosci 10:236. https://doi.org/10.3389/ cer. Dis Markers. 2017:7243968. https://doi. fnagi.2018.00236 org/10.1155/2017/7243968 5. Renz H, von Mutius E, Brandtzaeg P, Cookson WO, 18. Yuan L et al (2020) Long non-coding RNAs towards Autenrieth IB, Haller D (2011) Gene-environment precision medicine in gastric cancer: early diagnosis, interactions in chronic inflammatory disease. Nat treatment, and drug resistance. Mol Cancer 19:96 Immunol. 12(4):273–277. https://doi.org/10.1038/ 19. Micheletti R et  al (2017) The long noncoding RNA ni0411-­273 Wisper controls cardiac fibrosis and remodeling. Sci 6. Xue C, Zhang X, Zhang H, Ferguson JF, Wang Y, Transl Med 9:eaai9118 Hinkle CC, Li M, Reilly MP (2017) De novo RNA 20. Roy D, Tiirikainen M (2020) Diagnostic power of sequence assembly during in  vivo inflammatory DNA methylation classifiers for early detection of stress reveals hundreds of unannotated lincRNAs in cancer. Trends Cancer 6:78–81 human blood CD14+ monocytes and in adipose tis 21. Larson MH et al (2021) A comprehensive charactersue. Physiol Genomics. 49(6):287–305 ization of the cell-free transcriptome reveals tissue 7. Napoli M, Li X, Ackerman HD, Deshpande AA, and subtype-specific biomarkers for cancer detection. Barannikov I, Pisegna MA, Bedrosian I, Mitsch J, Nat Commun 12:2357

levels of inflammation or autoimmunity at their onset in an individual, has the power to revolutionize how we manage and prevent inflammatory conditions. So while the lncRNA field has gained momentum, we have really only begun to shine a light on this so-called RNA “dark matter”, but what we have uncovered so far offers great promise and we look forward to seeing more progress made in both basic and translational lncRNA research.

Index

A Abdominal aortic aneurysm (AAA), 78 Acute lung injury (ALI), 171 Acute myeloid leukemia (AML), 172 Adeno-associated virus (AAV), 162, 163 Adenovirus, 162 Adipose tissue, 37 Affinity column approach, 131 Alternative ORFS (altORFs), 84 Antinuclear antibodies (ANA), 59 Antisense Non-coding RNA in the INK4 Locus (ANRIL), 55, 56, 86 Antisense oligonucleotides (ASOs), 80, 162, 165, 182 Aortic aneurysm, 71 Apolipoprotein (APO), 80 Aptamers, 182 Atherosclerosis, 71, 154 Atherosclerotic plaques, 154 Atomic force microscopy (AFM) studies, 27 Autoimmune diabetes, 104 Autoimmune disease, 102 Autoimmunity, 171, 172, 183 B B-cell activating factor (BAFF), 60 β-cell proteins, 101 Biologic agents, 181 Biomarkers, 87–89 blood plasma, 131, 132 cancer, 127 classes, 127, 128 definition, 127 inflammation, 127 performance, 128, 129 planning, 131 sample collection and handling, 129 study design, 129 Biomolecules, 126 Blood glucose regulation, 101 Blood plasma, 129 Bone marrow edema (BME), 36

C Cajal-Retzius neurons, 17 Cancer, 124, 125, 162, 172, 173 Cancer susceptibility candidate 2 (CASC2), 55 Cardiac fibrosis, 71 Cardiac mesoderm enhancer associated noncoding RNA (CARMEN), 74 Cardiogenesis, 73 Cardiomyocytes, 25, 73 Cardiovascular disease (CVD), 162, 167, 168 biomarkers, 87–89 cholesterol homeostasis, 80–82 economic status, 71 endothelial cell growth, 79, 80 evolution, 89 human transcriptome, 72 lineage commitment and development, 73 macrophage lipid metabolism, 74, 76, 78 microarrays, 89 micropeptides, 84, 85 migration, 78, 79 non-coding RNA molecules, 72 regulation, 72 RNA-protein complexes, 72 SNPs, 85–87 sprouting, 79, 80 Celiac disease (CeD), 149–151 Cell type, 126 Cholesterol homeostasis, 80–82 Chronic inflammatory disease, 154 Cis-acting DNA regulatory elements, 12 Coding sequences (CDSs), 84 Comparative Annotation Toolkit (CAT), 17 Competing endogenous RNA (ceRNA), 41, 104 Congenital heart disease (CHD), 71 Coronary artery disease (CAD), 71 Correlative recurrent expression of predicted elements (CREPE), 89 C reactive protein (CRP), 59, 180 Crohn’s disease (CD), 153 Cross validation (CV), 137 Cryogenic electron microscopy (cryo-EM), 28 Cytokines, 180

© Springer Nature Switzerland AG 2022 S. Carpenter (ed.), Long Noncoding RNA, Advances in Experimental Medicine and Biology 1363, https://doi.org/10.1007/978-3-030-92034-0

185

Index

186 D Damage-associated molecular patterns (DAMPs), 38 Decapping enzyme 2 (DCP2), 151 Density gradient centrifugation (DG), 131 Destabilization of the medial meniscus (DMM), 41 Direct RNA sequencing, 19 Disease-modifying anti-rheumatic drugs (DMARDs), 181 Double stranded RNA (dsRNA), 122

MEG3, 152, 153 polymorphisms, 147 protein-coding genes, 148 transcriptome sequencing, 148 Genomics, 181 Glucocorticoids, 181 Glutamic acid decarboxylase 65 (GAD65), 101 G-protein coupled receptor 119 (GPR119), 76 Growth arrest-specific 5 (GAS5), 41, 48, 51, 59

E E330013P06, 76 Encyclopedia of DNA Elements (ENCODE) project, 13, 14 Endothelial cells (ECs), 73, 74 Erythrocyte sedimentation rate (ESR), 59 Exosome Total Isolation Chip (ExoTIC), 132 Exosomes, 126 Expectation Maximization (EM) models, 135 Expression quantitative trait loci (eQTL), 103 Extracellular matrix (ECM), 36, 40 Extracellular vesicles (EVs) exosomal lncRNAs, 127 intercellular communication, 126 lncRNAs, 126 types, 126

H Hepatocytes, 74, 87 High-density lipoprotein (HDL), 74 H19 imprinted maternally expressed transcript (H19), 51 Histone deacetylase 1 (HDAC1), 58, 151 Hormones, 97 HOXA transcript at the distal tip (HOTTIP), 54 HOX transcript antisense RNA (HOTAIR), 40, 41 Human carotid aortic SMCs (HCASMCs), 79 Human embryonic stem cells (hESCs), 123 Human endogenous retroviruses (HERVs), 122 Human immunodeficiency virus (HIV) infection, 151, 152 Human transcriptome, 14 Hypertension, 71 Hypoxia-inducible factor 2 alpha (HIF-2α), 38

F FANTOM project, 13 Fibrinogen, 180 Fibroblast-like synoviocytes (FLS), 36 Fluorescence correlation spectroscopy (FCS), 27 G GENCODE project, 14 Genome wide association studies (GWAS), 86, 101, 181 autoimmunity, 155 biology of disease, 147 CCR5AS, 151, 152 expression and co-expression analyses, 155 functional implication, 155 genetic and environmental factors, 147–148 genetic and epigenetic processes, 148 genetic variants, 148 IFNG-AS1, 153, 154 immune and inflammatory disorders, 155 immune-disease, 149 immune disease-associated SNPs, 148 immune-mediated diseases, 148 immune mediated disorders, 147 Immunochip platform, 148 inflammation, 155 LINC00305, 154 linkage disequilibrium (LD), 149 Lnc13, 149–151 mechanisms, 150 mechanisms of action, 148

I IGHCgamma1 (IGHCy1), 56 Illumina platform, 132, 133 Immune cells, 179 Inducible NO synthase (iNOS), 38 Inflammation, 124, 179, 180, 183 Inflammatory, 102 Inflammatory bowel diseases (IBD), 153, 154 Inflammatory diseases, 124, 162, 180, 181 Insulin-dependent diabetes mellitus (IDDM), 98 Insulinoma-associated autoantigen 2 (IA2), 101 Interleukin 6 (IL-6), 36 Ischemic stroke, 71 Islet-specific glucose-6-phosphatase catalytic subunit related protein (IGRP), 101 Isoforms and RNA modifications, 19 J Janus Kinase/Signal Transducers and Activators of Transcription (JAK/STAT), 36 JUPITER (Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin) clinical trial, 86 Juvenile diabetes, 98 K k-mers, 89 K Nearest Neighbors (KNN), 137

Index L Least Absolute Shrinkage and Selection Operator (LASSO), 137 Leave one out CV (LOOCV), 137 Left anterior descending artery (LAD), 168 Lentivirus, 163 Leukocytes, 101 LeXis, 81 Linkage disequilibrium (LD), 149, 152, 154 Lipid nanoparticles (LNPs), 164 Lipopolysaccharide (LPS), 154 Liposomes, 164 Lnc-HC, 81 LncHR1, 82 LncRNA conservation, 5 LncRNA downregulated in liver cancer (Lnc-DILC), 56 Locked nucleic acids (LNA), 166 Logistic regression, 137 Long intergenic ncRNA p21 (lincRNA-p21), 56 Long non-coding RNA molecules (lncRNAs), 12, 72, 155 alignment, 134 biomarkers, 6 in cardiovascular disease, 6 cell proliferation and migration, 6 classification, 4 conservation, 4 count normalization, 135 discovery, 18 dysregulation, 5, 6 elements, 16 evolution and function, 16 features, 16 genes, 15 human brain, 17 immune cell processes, 5 inflammation, 6 mechanisms of action, 122 multi-mapping reads, 135 overview, 7 PCA3, 6 production, 17 property, 16 quality control, 134 quantification, 134 research, 5 RNA-seq analysis, 134, 135 RNA selection, 133 roles, 5, 6 statistical modeling, 135, 137 structure, 5 TEs (see Transposable elements (TEs)) Lymphocytes, 179 M Macrophages, 36, 74, 75, 179 Maternally expressed gene 3 (MEG3), 53, 54 Matrix metalloproteinases (MMPs), 36 Mesenchymal stem cells (MSC), 54

187 Messenger RNAs (mRNAs), 11 Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1), 39, 40, 59, 80, 82 MeXis, 76 Microarrays, 12, 89 Micropeptides, 84, 85 MicroRNAs (miRNAs), 72, 121, 182 Microvesicles, 126 Mitogen-Activated Protein Kinase (MAPK), 36 Molecular mechanism, lncRNAs CRISPR/Cas9 knock out, 25 genome wide studies, 23 high resolution structural biology techniques, 23 high resolution techniques, 28 knock down studies, 25 loss of function studies, 25 phenotypes, 25 structural tools, 28 structure-function relationships, 23 proteins, 24 ribonucleoprotein complexes, 24, 25 riboswitch RNAs, 24 RNA processing, 25 RNA systems, 24 self splicing introns, 24 spliceosome, 25 2-D structural studies, 26, 27 Molecular signalling pathways, 39 Molecules, 126 Monocytes, 36 Mononuclear phagocytic cells, 164 Myeloid-derived suppressor cells (MDSCs), 60 Myocardial infarction (MI), 71 Myoregulin (MLN), 85 N Nanoparticles, 164, 165 Nanoparticle tracking analysis (NTA), 132 Nanopore sequencing technology, 133 Neurotransmitters, 97 NF-κB signaling, 124 Nitric oxide (NO), 38 Non-coding RNAs, 179–182 Noncoding transcriptome, 4 Non-negative Matrix Factorization (NMF), 137 Nonobese diabetic (NOD) mouse model, 101 Non-small cell lung cancer (NSCLC), 127 Non-steroidal anti-inflammatory drugs (NSAIDs), 181 Nuclear Enriched Abundant Transcript 1 (NEAT1), 52, 59, 60, 76, 81 Nuclear magnetic resonance imaging (NMR), 28 Nucleic acids, 173 Nutrients, 97 O Oligoadenylate synthase (OAS) proteins, 59 Oligonucleotides, 165 Open reading frames (ORFs), 84

Index

188 Oral squamous cell carcinoma (OSCC), 125 Organism’s transcriptome, 19 Osteoarthritis (OA), 42–45 B- and T- cells, 38 bioinformatics analysis, 38 cartilage, 37, 38 catabolic factors, 38 chondrocyte-fibroblast crosstalk, 37 chondrocytes, 37 immune cell regulation, 38 joint inflammation, 37 NF-kB signalling pathway, 38 pro-inflammatory cytokines, 38 risk factor, 37 synovial immune cells, 37 synovial joints, 37 Osteopontin (OPN), 39, 53 Oxidized LDL (oxLDL), 76 P Pancreatic islets, 98, 100, 101, 103, 105, 106, 108 Pattern recognition receptors (PRRs), 38, 179–180 Peptide nucleic acids (PNAs), 167 Peripheral blood mononuclear cells (PBMCs), 41, 59, 87, 154 Peritonitis, 171 pH-induced transmembrane structure (pHLIP), 167 Phosphatidylinositide-3-Kinase/AKT/mammalian Target of Rapamycin (PI3K/AKT/mTOR), 36 Plasmacytoma variant translocation 1 (PVT1), 54, 55 Platlet-derived growth factor-BB (PDGF-BB), 51 Pleckstrin homology domain–containing protein family M member 1 (PLEKHM1), 168 Pluripotent stem cell, 17 Polycomb group (PcG) proteins, 123 Polycomb repressive complex 2 (PRC2), 73, 123 Polyethylene glycol (PEG), 164 Polyethyleneimine (PEI), 165 Polymers, 165 POU class 3 homeobox 3 (POU3F3), 127 Precision Medicine Initiative, 181 Principal component analysis (PCA), 137 Programmed death-1 ligand (PD-L1), 106 Pro-inflammatory chemokines, 180 Proinsulin, 100–101 Proprotein convertase subtilisin/kexin type 9 (PCSK9), 166 Protein phosphatase 2A (PP2A), 58 Proteoglycans, 36 Q Quantitative polymerase chain reaction (qPCR), 138 R Radiographs, 36 Random Forests (RF), 137 RAS signaling, 125 Reactive oxygen species (ROS), 105 Recursive Feature Elimination (RFE), 137

Rheumatoid arthritis (RA), 46–48, 152–154, 171 Rheumatoid arthritis fibroblast-like synoviocyte (RA-FLS), 171 Rheumatology chronic inflammatory autoimmune condition, 35 pro-inflammatory cytokines, 36 pro-inflammatory microenvironment, 36 risk factors, 36 Ribonucleoprotein (RNP), 52 Ribosomal RNAs (rRNAs), 133 RNA-based therapies, 7 RNA-induced silencing complex (RISC), 165 RNA interference (RNAi), 165, 182 RNA-sequencing (RNA-seq), 12, 84 Rodent genomes, 12 S Sarcoplasmic reticulum (SR), 85 Sarcoplasmic reticulum calcium ATPase (SERCA) activity, 85 scRNA-Seq data analysis, 17 Sepsis, 168, 171 Sepsis-induced cardiomyopathy (SIC), 171 Serum exosomes, 41 Serum starvation, 51 Single nucleotide polymorphisms (SNPs), 72, 85–87, 101, 181 Small angle X-ray scattering (SAXS), 24, 27, 74 Small hairpin RNA (shRNA), 162 Small interfering RNAs (siRNAs), 76, 162, 165, 182 Small noncoding RNAs, 121 Small nucleolar RNA host gene 1 (SNHG1), 56 Small regulatory polypeptide of amino acid response (SPAAR), 85 Smooth muscle cells, 73, 78, 79 Soluble IL-6 receptor (sIL-6R), 37 Spherical nucleic acids (SNAs), 164 Spinal muscle atrophy (SMA), 7 Sprague Dawley (SD) rat model, 39 Support vector machines (SVM), 137 Systemic lupus erythematous (SLE), 124 Systemic lupus erythematosus (SLE) blood-brain barrier, 58 ceRNA networks, 58 chronic autoimmune disease, 57 clinical markers, 59 co-expression analysis, 58 heterogeneity, 57 inflammatory pathways, 58 kidney injury, 58 lncRNAs, 49–51 pathophysiology, 57, 58 T-cells, 58 whole transcriptome profiling, 58 T Tanshinone IIA (Tan IIA), 48 Taurine up-regulated 1 (TUG1), 55, 60 Temporomandibular joint OA (TMJ-OA), 41 TGFβ signalling pathway, 53

Index Therapeutic targets challenges, 173 chemical modifications, 167 chronic disease, 162 delivery platforms, 163 interventions, 163 in vivo models, 169–170 nucleic acid, 162 tissue- and disease-specific expression patterns, 162 Thrombosis, 71 Tissue inhibitors of metalloproteinases (TIMPs), 36 TNF and HNRNPL related immunoregulatory LncRNA (THRIL), 57, 61 TNF receptors (TNFRs), 37 Toll-like receptors (TLRs), 4, 37, 104 Transcriptomics, 181 Transcripts-per-million (TPM), 135 Transmission electron microscopy (TEM), 132 Transposable elements (TEs), 16, 72 classes, 122 embryonic and adult tissues, 123, 124 host genes, 122 internal RNA polymerase II promoter, 122 lncRNAs, 122 mobile DNA elements, 122 open reading frames (ORFs), 122 Transverse aortic constriction (TAC), 168 TRIB1 associated locus (TRIBAL), 86 Trimmed means of M (TMM), 135 Tumor necrosis factor-α (TNFα), 36, 180 2’-fluoro (2’-F), 166 2’-O-methoxyethyl (2’-O-MOE), 166 2’-O-methyl (2’-O-Me), 166 Type 1 diabetes (T1D), 149–153 autoimmunity, 109 β-cell responses, 105–108 cell types, 108 cellular processes, 98 cellular responses, 110 chronic exposure, 97 expression analyses, 98

189 genetic and expression studies, 101, 102 genetic and transcriptome studies, 108 immune cell responses, 103–105, 109 immune cells, 98 inflammatory and autoimmune responses, 109 molecular pathways, 110 molecules, 110 pancreatic β cells, 98, 100, 101 pathological inflammation, 109 polymorphisms, 110 protein-coding genes, 98, 109 transcriptome analysis, 98 Type 2 diabetes (T2D), 108 U Ulcerative colitis (UC), 153 Ultracentrifugation (UC), 131 Ultrasonography, 36 Unique molecular identifier (UMI), 134 Urothelial carcinoma-associated 1 (UCA1), 55, 60 U.S. National Institutes of Health (NIH), 127 V Vascular smooth muscle cells (VSMCs), 74 W Wnt signalling pathway, 54 X X-inactive specific transcript (XIST), 52, 53, 60 X-ray crystallography, 25, 28 Z Zinc finger antisense 1 (ZFAS1), 127 Zinc transporter 8 (ZnT8), 101 ZNFX1 antisense RNA1 (ZFAS1), 57