Middle Molecular Strategy: Flow Synthesis to Functional Molecules 9789811624582, 9811624585

Citation preview

Koichi Fukase Takayuki Doi   Editors

Middle Molecular Strategy Flow Synthesis to Functional Molecules

Middle Molecular Strategy

Koichi Fukase Takayuki Doi •

Editors

Middle Molecular Strategy Flow Synthesis to Functional Molecules

123

Editors Koichi Fukase Graduate School of Science Osaka University Toyanaka, Osaka, Japan

Takayuki Doi Graduate School of Pharmaceutical Sciences Tohoku University Sendai, Japan

ISBN 978-981-16-2457-5 ISBN 978-981-16-2458-2 https://doi.org/10.1007/978-981-16-2458-2

(eBook)

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 This work is subject to copyright. All rights are solely and exclusively licensed 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 Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

Recognition of biomolecules can be broadly categorized into “lock and key” interaction in the binding pocket and planar interaction on the protein surface. The contact areas of the former are estimated to be between 300 and 1000 Å2 and many small molecule drugs with a molecular weight of less than 500 interact with target proteins in this manner [1]. The typical recognition of the latter is protein-protein interaction (PPI) with relatively large contact surfaces (1500–3000 Å2) [2, 3]. Biopharmaceuticals such as antibody drugs typically have a high molecular weight can target PPIs with the large contact surface. Middle molecules (also called as medium-sized molecules, middle-size molecules, middle-sized molecules, mid-sized molecules) are categorized as a group of compounds with a molecular weight of about 500–3000 Å2. Middle molecules have high chemical diversity and contain a variety of compounds, including natural products, glycans, peptides, and nucleic acid drugs. Many natural product drugs and derivatives, such as avermectin, ivermectin, paclitaxel, cyclosporins, tacrolimus, vancomycin, and etc. are middle molecules. Middle molecules have great potential as higher bio-functional molecules because of the following characteristics. Since middle molecules have relatively large surface area in comparison to small molecules and can provide a similar contact area size as macromolecules, the strict and diverse molecular recognition is possible based on the multipoint interaction between middle molecules and target proteins, enabling both “lock and key” recognition and the protein surface recognition. Some of middle molecules are membrane permeable and orally active, whereas most of macromolecules are not. Some of middle molecules such as cyclosporins, tacrolimus, cotylenin A, and etc. simultaneously interact with several proteins to control the signaling pathway (Table 1). However, the inherent structural complexity of middle molecules was an obstacle for the practical use of middle molecules, since the synthesis of them is often difficult and generally requires multiple reaction steps. We therefore established the research project “Middle molecular strategy: Creation of higher bio-functional molecules by integrated synthesis” supported by Grant-in-Aid for Scientific Research on Innovative Areas. The project achieved efficient syntheses of various bioactive middle molecules by an innovative synthetic strategy based on v

vi

Preface

Table 1 Feature of middle molecules

Synthetic easiness Wide areas for planar recognition Multiple targeting Oral availability Membrane permeability

Small molecules easy difficult

Middle molecules difficult possible

Macromolecules easy easy

difficult easy easy

possible possible possible

easy difficult difficult

reaction integration. Further, novel bio-functional middle molecules were developed by new strategies such as function integration. The project was conducted in three areas, A01: creation of higher functional middle molecules, A02: efficient synthesis of bio-functional middle molecules, and A03: development of higher reaction integration. In A01, we investigated the efficient synthesis of bio-functional middle molecules such as glycans, nucleic acids, peptides, and lipids. Development of novel bio-functional middle molecules were also investigated by function integration, i.e., conjugation of bioactive compounds, creation of novel bio-functional molecules possessing the p electron system compounds. A02 aimed the highly efficient synthesis of bioactive middle molecules such as complex natural products. A02 studied the reaction integration based on novel concepts such as bio-conjugation on in vivo chemical synthesis in living cells and animals. A03 developed various continuous reaction processes using micro-flow and one-pot syntheses as well as practical reactions for multi-step synthesis. Synthetic transformations allowed by micro-flow methods, such as very fast reactions using unstable reactive species, were also investigated. This book summarizes the results of this project for 5 years (2015–2019) and consists of four parts: Part I Bio-functional Middle Molecules, Part II Synthetic Advances in Complex Middle Molecules, Part III Synthetic Advances in Aromatic Middle Molecules, and Part IV Synthetic Advances in Flow Chemistry for Middle Molecules. During the last five years, middle molecule drug discovery have been getting more popular. Middle molecules that can precisely control the biological responses are expected as the new drug modalities for the treatment of unsolved medical issues such as unmet medical needs. The continuous production process by using flow reactors is gaining momentum in the industrial production of pharmaceuticals and fine chemicals. I hope this book will help convey the appeal of middle molecular strategy to readers. Finally, I sincerely thank all the co-authors for their contributions to this book. Toyonaka, Japan

Koichi Fukase

Preface

vii

References 1. Cheng AC, Coleman RG, Smyth KT, Cao Q, Soulard P, Caffrey DR, Salzberg AC, Huang EC (2007) Structure-based maximal affinity model predicts small-molecule druggability. Nat Biotechnol 25:71–75. https://doi.org/10.1038/nbt1273 2. Smith MC, Gestwicki JE (2012) Features of protein-protein interactions that translate into potent inhibitors: topology, surface area and affinity. Expert Rev Mol Med 14:e16. https://doi. org/10.1017/erm.2012.10 3. Jones S, Thornton JM (1996) Principles of protein-protein interactions. Proceedings of the National Academy of Sciences of the United States of America 93(1):13–20

Contents

Part I 1

2

3

4

5

6

Conjugation Strategies for Development of Bioactive Middle Molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yoshiyuki Manabe, Atushi Shimoyama, Kazuya Kabayama, and Koichi Fukase

3

Total Synthesis, Biological Evaluation, and 3D Structural Analysis of a Cyclodepsipeptide Natural Product . . . . . . . . . . . . . . Takayuki Doi

21

Development of Middle-Size Molecules for Alkylation to Higher-Order Structures of Nucleic Acids . . . . . . . . . . . . . . . . . Fumi Nagatsugi and Kazumitsu Onizuka

35

In Situ Synthesis of Glycoconjugates on the Cell Surface: Selective Cell Imaging Using Low-Affinity Glycan Ligands . . . . . . . Shogo Nomura, Misako Taichi, and Katsunori Tanaka

55

Chemical Approach Toward Controlling of Transient Protein Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Junko Ohkanda

77

Mid-Sized Macrocyclic Peptides as a New Drug Modality . . . . . . . Yuki Goto, Masanobu Nagano, and Hiroaki Suga

Part II 7

Bio-functional Middle Molecules

97

Synthetic Advances in Complex Middle Molecules

Convergent Total Synthesis of (+)-Cotylenin A . . . . . . . . . . . . . . . . 111 Masahiro Uwamori, Ryunosuke Osada, Ryoji Sugiyama, Kotaro Nagatani, and Masahisa Nakada

ix

x

Contents

8

Electrochemical Synthesis of Oligosaccharides as Middle-Sized Molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Md. Azadur Rahman, Kumpei Yano, Sujit Manmode, Yuta Isoda, Norihiko Sasaki, Toshiyuki Itoh, and Toshiki Nokami

9

Efficient Synthesis of Biologically Active Peptides Based on Micro-flow Amide Bond Formation . . . . . . . . . . . . . . . . . 139 Shinichiro Fuse

10 Design, Concise Synthesis and Self-Assembly of the Mid-Sized Molecules Exploiting Bispyrrolidinoindoline Alkaloidal Scaffolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Hiroki Oguri Part III

Synthetic Advances in Aromatic Middle Molecules

11 Efficient Synthesis of Polycyclic Aromatic Hydrocarbons Using Unreactive Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Fumitoshi Kakiuchi 12 Efficient Access to Highly Condensed Aromatic Compounds Using Reactive Molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Toshiyuki Hamura 13 Reaction Field for a Lewis Acid with a Tunable Factor for Selective Organic Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Akihito Konishi, Yoshihiro Nishimoto, and Makoto Yasuda Part IV

Synthetic Advances in Flow Chemistry for Middle Molecules

14 Organozinc Reagent in Flow Chemistry . . . . . . . . . . . . . . . . . . . . . 263 Seijiro Matsubara 15 Development and Integration of New Green Reactions . . . . . . . . . . 275 Takashi Ohshima, Hiroyuki Morimoto, and Tetsuya Kadota 16 Development of Novel Organic Electrosynthetic Processes Using Electrochemical Flow Microreactor . . . . . . . . . . . . . . . . . . . . 297 Mahito Atobe and Naoki Shida 17 Multiple Organolithium Reactions Based on Space Integration . . . 309 Aiichiro Nagaki 18 Efficient Photoreaction Using Photo-Microreactors . . . . . . . . . . . . . 321 Takahide Fukuyama 19 Recent Advances in the Integrated Microflow Synthesis of Organofluorine Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Hideki Amii

Part I

Bio-functional Middle Molecules

Chapter 1

Conjugation Strategies for Development of Bioactive Middle Molecules Yoshiyuki Manabe, Atushi Shimoyama, Kazuya Kabayama, and Koichi Fukase

Abstract Conjugation of bioactive molecules can generate molecules with new functions by simultaneously or sequentially interacting with multiple targets. These conjugated molecules are called small molecule drug conjugates (SMDCs), that are considered to be a type of middle molecules (mid-sized bioactive molecules). A similar conjugation strategy, a self-adjuvanting strategy, has been used for vaccine development. The antigen-adjuvant conjugates are taken up by the antigen-presenting cells via the interactions between innate immune receptors and their ligands, and the adjuvants activate the immune system and efficiently induce antibodies. Keywords Middle molecules Carbohydrate antigens

1.1


Conjugation
Adjuvants
Vaccines


Introduction

Living organisms are harmonious and dynamic assemblies of various bio-molecules, which interact with other bio-molecules to form a dynamic network of biological information. Bioactive compounds can interact with bio-molecules to alter and modulate the network. Drugs need to target node molecules that link appropriate number of other node molecules in the network to induce effective and controlled response, since acting on hub molecules where many nodes are connected should cause undesired side effects. Some of middle molecules can interact multiple bio-molecules simultaneously or sequentially to show complex biological responses. For example, immunosuppressive agents such as cyclosporine (1), tacrolimus (FK-506) (2), and sirolimus (rapamycin) (3) interact with two target proteins to express their activity (Fig. 1.1) [1]. Cyclosporin inhibits calcineurin by Y. Manabe
A. Shimoyama
K. Kabayama
K. Fukase (&) Department of Chemistry, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_1

3

4

Y. Manabe et al.

Fig. 1.1 Structures of immunosuppressive agents

forming a complex with cyclophilin in T-cells. Tacrolimus forms a complex with FKBP12 (FK506 binding protein 12) and inhibits calcineurin. Sirolimus forms a complex with the FKBP12 protein to binds to mTOR (mammalian target of rapamycin). Although it is difficult to de novo design molecules that can simultaneously interact with multiple targets, conjugation of bioactive molecules can generate middle molecules with new functions. In fact, conjugation is a common strategy to develop biopharmaceuticals such as antibody-drug conjugates (ADCs) consisting in monoclonal antibodies or antibody fragments linked to small molecule cytotoxic drugs. Small molecule drug conjugates (SMDCs) are a new class of middle molecules which consists of small molecules that target specific cancer cells and potent antitumor agents (Fig. 1.2). Vintafolide (4) is a complex of folic acid, which targets the folate receptor expressed on cancer cells, and vinblastine, a potent antitumor alkaloid [2]. Folic acid-tubulysin conjugate EC1456 (5) showed significant anti-proliferative activity against FR-positive tumors [3]. Targeted radionuclide therapy, which uses radionuclides instead of anticancer agents, is also promising as a new anticancer therapy. Lutathera® (177Lu-DOTA[Tyr3]-octreotide (6)) was developed for the treatment of neuroendocrine tumors [4]. Targeted a-particle therapy (or TAT) uses a-emitting radionuclides such as 225 Ac and 211At, which show highly cytotoxic effects because of the very high linear energy of a-particle. Clinical tests with 225Ac-PSMA-617 (7) demonstrated the extremely potent therapeutic effect on metastatic castration-resistant prostate cancer [5]. To date, various immunotherapies have been developed, including passive immunotherapy (typified by antibody therapy) and active immunotherapy (typified by vaccines). Chemical conjugation has contributed to the development of novel immunotherapeutic agents such as the above-mentioned antibody-drug conjugates (ADCs) and self-adjuvanting conjugated vaccines. In this chapter, the authors focus the self-adjuvanting conjugated vaccines as examples of conjugated middle molecules.

1 Conjugation Strategies for Development of Bioactive …

5

Fig. 1.2 Structures of SMDCs and targeted radiopharmaceuticals

Immunity serves as a biological defense against foreign substances such as bacteria and viruses and is classified in terms of innate immunity and acquired immunity. Innate immunity is activated by the recognition of pathogen-associated molecular patterns (PAMPs) by innate immune receptors [i.e., pattern recognition receptors (PRRs)]. Acquired immunity consists of a highly specialized defense system that is acquired through contact with pathogens and produces immunological memory. The antibody response is part of acquired immunity. Vaccines, which utilize acquired immunity, increase resistance to pathogens by memorizing them. Attenuated or inactivated pathogens and their toxins have historically been used as vaccines. Vaccines consisting of pure antigens such as recombinant proteins or chemically synthesized peptides have also been enthusiastically studied in order to induce specific immunoresponses. These molecules are well defined and homogeneous, minimizing the risk of unexpected side effects and making them suitable for clinical use. Since the immune response cannot be effectively induced by administering the antigen alone, it is necessary to co-administer an adjuvant, which is a substance that enhances the host’s immune response to the antigen. Innate immune receptor ligands act as effective adjuvants [6]. The self-adjuvant strategy provides an attractive approach for inducing an antigen-specific immune response by using a well-defined adjuvant that integrates

6

Y. Manabe et al.

with the antigen without the need for co-administration of additional adjuvants and carrier proteins. This chapter reviews vaccine adjuvants, including our contribution for developing lipid A adjuvants, and then outlines recent advances in self-adjuvanting conjugated vaccines.

1.2

FDA-Approved Adjuvants and Development of Lipid a Adjuvants

Adjuvants are broadly categorized to delivery systems and immunostimulators. The former promotes the uptake of vaccine antigens and also functions as a carrier for presenting the antigens to the immune system. Adjuvants MF59 (Novartis) and AS03 (GlaxoSmithKline), which are oil-in-water emulsions made of squalene and emulsifiers, have been used in current vaccines such as Focetria influenza vaccine (Novartis) and Arepanrix influenza vaccine (GlaxoSmithKline). On the other hand, various innate immune activators have been investigated as adjuvants. However, since many innate immune activators have side effects typified by a strong pro-inflammatory action in addition to a useful adjuvant action, it is necessary to modulate their immunostimulating responses. Alum, an adjuvant consisting of aluminum salts, has been used since the 1920s. Suspending the antigen in an alum emulsion promotes antigen uptake by antigen-presenting cells (ACPs). Recent studies have shown that alum promotes the innate immune response by activating NOD-like receptor protein 3 (NALP3) inflammasomes [7, 8]. Ishii and Desmet report that host DNA released from dying cells by the action of alum mediates adjuvant activity [9, 10]. GlaxoSmithKline has developed lipid A adjuvants that show moderate immunostimulating activity via Toll-like receptor 4 (TLR4). A monophosphoryl lipid A derivative, 3D-MPL (12), is an adjuvant derived from Salmonella lipopolysaccharide and shows anti-viral responses (Fig. 1.3) [11]. AS04 and AS01B (GlaxoSmithKline) are clinically approved adjuvants containing 3D-MPL with Al(OH)3 (AS04) or QS-21 saponin (AS01B), respectively. Oligonucleotides containing unmethylated CpG sequences are the ligand of Toll-like receptor-9. Dynavax Technologies developed CpG 1018 composed of 22-mer oligonucleotide containing CpG motifs as an adjuvant for immunization against hepatitis B virus (HBV) [12]. We have been investigating the synthetic and functional studies on immunopotentiators such as lipopolysaccharide (LPS) from Gram-negative bacteria and its active entity lipid A. LPS, also known as an endotoxin, has extremely strong immunostimulatory and pro-inflammatory effects. LPS induces various cytokines, nitric oxide (NO), active oxygen, leukocyte migration, antibacterial peptide production, and lymphocyte activation to protect the body against bacterial infection. On the other hand, endotoxin triggers severe systemic illness that can cause

1 Conjugation Strategies for Development of Bioactive …

7

Fig. 1.3 Structures of lipid A and derivatives

multiple organ failure, low blood pressure, and septic shock. The innate immune receptor TLR4/MD-2 is responsible for the recognition of lipid A and LPS. We and other groups have found that the acyl and phosphate groups in lipid A regulate the interaction with the TLR4/MD-2 complex and significantly influence the activity. Downstream of TLR4/MD-2 has several signaling pathways such as NF-jB pathway that induces inflammatory cytokines involved in acute inflammation such as TNF-a, IL-1, and IL-6, and IRF3 pathway that induces interferon. We have studied the structure-activity relationship study of lipid A to reveal that the lipid A signals can be controlled by suitable structural modification in lipid A. Escherichia coli lipid A (8) effectively induces dimerization of TLR4/MD-2 to show strong immunopotentiating and pro-inflammatory effects. Biosynthetic precursor type lipid A, lipid IVa (9), has moderate immunostimulatory effects in mice but has antagonistic effects in humans. Monophosphoryl lipid A (MPL) (10), which lacks 1-O-phosphate, shows a much lower ability in TLR4/MD-2 dimerization than lipid A 8. Consequently, 10 shows much weaker activity in the production of inflammatory cytokines such as TNF-a and IL-6. MPL 10 promotes the intracellular translocation of TLR4/MD-2 to induce interferon b similar to lipid A 8 [13]. IL-18-inducing ability of MPL 10 lacking 1-position phosphate is weaker than that of E. coli lipid A 8, whereas the MPL 11 lacking 4-position phosphate shows IL-18-inducing ability similar to E. coli LPS [14]. As described above, GlaxoSmithKline developed 3D-MPL (12). The inflammatory activity of 3D-MPL is weak but 12 was reported to selectively activate the IRF3 pathway [11]. AS04, a mixture of 3D-MPL and aluminum salt, is

8

Y. Manabe et al.

used as an adjuvant for the human papillomavirus (HPV) vaccine Cervarix (cervical cancer preventive vaccine) and the HBV vaccine (Fendrix). We have shown that some parasitic and pathogenic bacteria avoid the innate immune system. Yersinia pestis expresses the antagonist lipid IVa (9) at 37 °C, which is the body temperature of mammals, and inhibits the host’s innate immune response to establish Y. pestis infection [15]. Lipid A (13) of Helicobacter pylori, which inhabits the stomach and causes gastric ulcer, acts as an antagonist, whereas the immunostimulatory activity of 14 is extremely weak. We found that lipid As 13 and 14 derived from these parasitic bacteria strongly induces IL-12 and IL-18, which are involved in the parasitism of parasitic bacteria and the induction of chronic inflammation [16]. Since the combination of IL-12 and IL-18 induces IFNc from T-cells to induce antitumor and antiallergic effects, lipid As from parasitic bacteria are adjuvant candidates that can activate T-cell responses. Recently, Kiyono and Kunisawa found that the genus Alcaligenes are symbiotic within the dendritic cells of the Peyer’s patches of immune tissue in the intestinal tract. We found LPS from Alcaligenes faecalis induces production of IgA important for mucosal immunity, but rarely elicits inflammation [17]. We recently demonstrated that synthetic A. faecalis lipid A promoted antigen-specific IgA and IgG production and enhanced Th17 response in vaccination of female BALB/c mice without excessive inflammation [18, 19]. A. faecalis lipid A was thus identified to be a promising and safe vaccine adjuvant candidate. Various types of lipid As are expected to be effective adjuvants for the self-adjuvanting strategy described below.

1.3

Self-adjuvanting Vaccines

Recently, self-adjuvanting strategy has received much attention for vaccine development. Self-adjuvanting vaccines use covalently linked or physically associated antigen-adjuvant conjugates [20–59]. The antigen-adjuvant conjugates are efficiently taken up by antigen-presenting cells via the interaction with innate immune receptors with their ligand adjuvants and promote antigen presentation. The adjuvant parts then activate the immune system to promote cytokines induction, and antibody production. Lipidic conjugates can form self-aggregate particles that are readily taken up antigen-presenting cells. Self-adjuvanting vaccines have been actively studied for the development of carbohydrate-based vaccines [20–25, 29–31, 34–45, 47–49]. Since glycans cover the cell surface and are closely related to the state of cells, they function as the “face” of the cells. Tumor-related carbohydrate antigens (TACAs), which are overexpressed in tumor cells, can be used for cancer vaccine therapy. Pathogens such as bacteria and viruses display unique glycans on their surface. These glycans are attractive targets for vaccine therapy against many infectious diseases. However, these glycan antigens are T-cell-independent antigens and have low immunogenicity, development of carbohydrate-based vaccines is challenging. The

1 Conjugation Strategies for Development of Bioactive …

9

self-adjuvanting strategy can overcome these difficulties. The following is a review of previous reports describing the findings of the self-adjuvant vaccine.

1.3.1

TLR2 Ligand-Conjugated Vaccines

TLR2 recognizes lipopeptides from microorganisms and activates the immune system. Although the lipopeptide itself has not been approved as an adjuvant, Trumenba®, which is approved as a vaccine against Neisseria meningitidis group B disease, has a lipoprotein with TLR2-stimulating activity that acts as a self-adjuvanting vaccine [60]. Well-known TLR2 ligands include macrophage-activating diacylated lipopeptide 2 (MALP-2) derived from mycoplasma, diacylated Pam2CSK4, and triacylated Pam3CSK4. We also reported that N-terminal diacylated lipopeptides on lipoproteins from Staphylococcus aureus, a Gram-positive bacterium, induced TLR2 activation in a sequence-dependent manner [61] Seya et al. presented evidence showing that diacylated lipopeptides, including Pam2CSK4, exerted antitumor effects through dendritic cell-mediated activation of natural killer T (NKT) cells [62]. These results indicate that lipopeptides effectively induce antitumor immune responses and are promising adjuvant candidates for anticancer vaccines. MUC1 is a glycoprotein expressed on the cell surface. MUC1 on tumor cells has a lower glycosylation rate and a shorter glycan chain length. In 2007, Boons et al. synthesized a three-component conjugate vaccine 15 composed of MUC1 glycopeptide overexpressed in human breast cancer cells, Pam3CSK4 as a TLR2 ligand, and a T-cell epitope, which activates T-cell immunoresponses and induces a class switch from IgM to IgG antibodies [20]. The synthesized three-component conjugate vaccine 15 potently induced the production of IgG antibodies to MUC1. They also showed that the additional adjuvant was not necessary in the self-adjuvant strategy (Fig. 1.4). Kunz, Li et al. synthesized MUC1 glycopeptide Pam3CSK4 conjugates 16–17 and confirmed that they induced antigen-specific immune responses (Fig. 1.5) [21].

Fig. 1.4 Structure of MUC1 self-adjuvanting glycopeptide vaccine 1

10

Y. Manabe et al.

Fig. 1.5 Structures of MUC1 self-adjuvanting glycopeptide vaccines 2

The three-component vaccines 19 and 20 induced much higher antibody production than the antigen alone (Fig. 1.5) [22]. Similar to the report by Boons et al., co-administration of an additional adjuvant was not required. Kunz et al. also introduced various T-helper cell epitopes or multiple T-helper cell epitopes into the MUC1 glycopeptide-based self-adjuvanting vaccine and evaluated their effects [23, 24]. We reported a self-adjuvanting vaccine based on the sialyl-Tn (STn) antigen, which is a TACA abundantly expressed on several epithelium-derived tumors [25]. The Sialyl-Tn (STn) antigen is a mucin-derived TACA that is highly expressed in breast cancer, lung cancer, colon cancer, gastric cancer, ovarian cancer, etc., but is rarely expressed in normal cells. Therefore, the STn-KLH complex (Theratope) was developed as a vaccine for metastatic breast cancer, but had failed in phase III clinical trial. STn-recognizing antibody and KLH-recognizing antibody were induced in the Theratope-administered patients, but clinically relevant antibodies that recognized mucin were not effectively produced [63]. We applied the self-adjuvanting strategy to develop an STn-based vaccine. We replaced the STn monomer with a clustered (trimeric) STn antigen to enhance the tumor specificity by mimicking the in vivo STn-expression pattern on the pancreatic and ovarian cancers. Cancer vaccine candidate 28 having TriSTn, lipopeptide adjuvant (Pam3CSK4), and T-cell epitope was thus synthesized. Since it was found that the antigenicity of STn was improved by converting the 5-position acetyl group of sialic acid to a propionyl group [64], a propionyl-modified form 29 was also synthesized (Fig. 1.6).

1 Conjugation Strategies for Development of Bioactive …

11

Fig. 1.6 Synthesis of STn anticancer vaccine candidates

Glycosylation of azide sialic acid donor 21 was carried out with the galactosaminyl threonine derivative 22. In order to obtain a selectivity by utilizing the solvent effect of propionitrile, the reaction should be carried out at low temperature (−78 °C). However, NIS-TfOH, which is commonly used for glycosylation with thioglycoside did not promote the glycosylation at −78 °C. We found that the combination of ICl and In(OTf)3 is much more reactive and effectively promotes the glycosylation using thioglycoside donors at −78 °C [65]. The glycosylation of 21 and 22 proceeded smoothly at −78 °C to give 23. After deriving this to TriSTn 24 and 25, T-cell epitope 26 was coupled by click reaction in the presence of a copper catalyst, and then Pam3CSK4 27 was coupled by thioether ligation to afford 28 and 29, respectively. Antibodies against the clustered STn antigens were efficiently induced by immunization of 28 or 29 to mouse, again suggesting that the self-adjuvanting strategy provides an effective option for developing TACA-based anticancer vaccines. HER2 is one of the promising tumor-associated antigens (TAAs) of cancers and has been used as a target for immunotherapy using anti-HER2 monoclonal antibody trastuzumab. Cancer vaccine therapies using several anti-HER2 peptide vaccines, including E75, GP2, and AE37 have also been investigated for the treatment of breast cancer. The CH401 peptide, YQDTILWKDIFHKNNQLALT, is an epitope of a chimeric antibody CH401 mAb, which exhibited efficient antibody-dependent cellular cytotoxicity against HER2-positive human adenocarcinoma. The CH401 peptide contains B- and T-cell epitopes and can induce specific IgG antibody production against HER2. A 9-mer epitope (mini fragment of CH401, MFCH401: DTILWKDIF) was predicted as the immunodominant fragment in the CH401

12

Y. Manabe et al.

Fig. 1.7 Structures of Pam3CSK4-MFCH401 conjugates

Fig. 1.8 Structure of Pam3CSK4-CH401 conjugate

peptide. Therefore, we synthesized Pam3CSK4-MFCH401 conjugate 30 and conjugates having MFCH401 tandem repeats 31 and 32 (Fig. 1.7). Pam3CSK4MFCH401 30 effectively induced antibodies against HER2-overexpressing human BT474 cells [26]. This result clearly shows that MFCH401 is a promising epitope, indicating the usefulness of the self-adjuvant vaccine. We also synthesized Pam3CSK4-CH401 conjugate 33, which showed much more potent activity than 30. This result indicated that CH401 contains both B-cell and T-cell epitopes [27] (Fig. 1.8). Khan, Payne, and Moyle also independently reported that conjugates with TLR2 ligands enhanced the production of antigen-specific antibodies [28–34]. It should be noted that Payne and co-workers reported that a self-adjuvanting vaccine consisting of MALP-2 and a MUC1 glycopeptide induced induce high titers of IgG antibodies against MUC1 in C57BL/6 mice in the absence of a T-cell epitope [31]. As described here, self-adjuvanting vaccines developed using TLR2 ligands universally gave successful results.

1.3.2

TLR4 Ligand-Conjugated Vaccines

Considering 3D-MPL (12) is a FDA-approved adjuvant, lipid A derivatives are also promising adjuvants for self-adjuvanting vaccines. It should be noted that slight modification in lipid A, e.g., introduction of hydrophobic linker or PEG linker sometimes abolished the activity of lipid A. We therefore developed fluorescenceand biotin-labeled lipid As possessing a hydrophilic glutaryl-glucose linker, which maintain the bioactivity and also prevent self-aggregation [66]. Guo et al. used MPL of Neisseria meningitidis lipid A, which has symmetrically distributed acyl group on the two sugar units, as an adjuvant. Carbohydrate

1 Conjugation Strategies for Development of Bioactive …

13

Fig. 1.9 Structures of TLR4-ligand-conjugated vaccines

antigens, including GM3 (TACA), STn (TACA), and a-2,9-oligosialic acid (a meningococcal antigen) were introduced to 1-position of MPL to afford 34–36 (Fig. 1.9) [35–37]. These conjugate vaccines efficiently promoted antibody production. Guo also reported the MPL-based self-adjuvanting vaccine composed of a tetrasaccharide of mycobacterial lipoarabinomannan (LAM) linked 6’-position of MPL [38]. Given that the 6’-position of lipid A is linked to the polysaccharide chain in LPS, modification at the 6’-position seems to be reasonable design for avoiding any potential influence on the binding between MPL and TLR4. Developing lipid A derivatives with simplified structures (while retaining the activity) will also contribute to the development of a self-adjuvanting vaccine, based on TLR4 activation, by reducing the difficulty associated with synthesis. Indeed, Jiang et al. reported the synthesis of a conjugate 37 between the Thomsen– Friedenreich antigen and RC-529, a structure-simplified TLR4 agonist (Fig. 1.9) [39]. Code synthesized the conjugate 38 composed of TLR4 ligand, CRX-527, and the model peptide antigen, which initiated specific CD8+ T-cell-mediated killing of antigen-loaded target cells [40].

1.3.3

CD1d Ligand-Conjugated Vaccines

CD1d ligands, which are presented to NKT cells and activate the immune system, have been used as adjuvants for self-adjuvanting vaccines. De Libero reported the conjugated vaccine 39 composed of the capsular polysaccharide of Streptococcus

14

Y. Manabe et al.

Fig. 1.10 Self-adjuvanting vaccines using a CD1d ligand

pneumoniae and a-galactosylceramide, a representative CD1d ligand (Fig. 1.10) [41]. The conjugated vaccine 39 activated NKT cells and produced antibodies against the carbohydrate antigen. Hermans, Painter et al. synthesized a conjugate 40 between a peptide antigen derived from ovalbumin (OVA) and an a-galactosylceramide derivative (Fig. 1.10) [42]. The OVA peptide (antigen) was designed to be cleaved from the conjugate. This vaccine robustly produced antibodies against the OVA peptide, and significantly suppressed the growth of melanoma cells expressing the OVA peptide. Hermans, Painter et al. also synthesized conjugates of the OVA peptide with a-galactosylceramide that were connected via a disulfide or maleimido linkage [43]. In this case, the antigen peptide can be released by disulfide exchange with intracellular glutathione. Guo and Li reported the synthesis of a-galactosylceramide-conjugated STn and MUC1, respectively. These conjugate vaccines induced the production of antibodies against the corresponding antigens [44, 45].

1.3.4

Self-assembling Self-adjuvanting Vaccines

As described above, chemical conjugation is a reliable and effective strategy for preparing self-adjuvanting vaccine. However, chemical conjugation is sometimes tedious and the yields are not always high enough. In fact, self-adjuvanting vaccines can be divided into two main types: self-assembling vaccines and adjuvant-conjugated vaccines. Self-assembling of antigen enables multiple antigen presentation to improve the metabolic stability and the antigen uptake into the immune cells. Lipidated amphiphilic peptides spontaneously self-assemble to afford various forms of aggregates [50, 51]. Fibril forming peptides, such as self-assembling Q11 (QQKFQFQFEQQ) [52–55], KFE8 (FKFEFKFE) [56, 57], and RADA16 (RADARADARADARADA) [58, 59] have been also applied to self-adjuvanting vaccines to enhance the antigenicity of the attached epitopes.

1 Conjugation Strategies for Development of Bioactive …

15

Fig. 1.11 Self-adjuvanting vaccines (V1-V6) for in vivo mouse immunization

We combined both methods and designed co-assembling vaccines comprising in a lipidated CH401 peptide antigen and lipid adjuvants as breast cancer vaccine candidates (Fig. 1.11) [27]. We prepared six vaccines V1–V6 including self-assembling vaccine V1 containing N-palmitoylated CH401 43 and the conjugated vaccine V2 consisting of Pam3CSK4-CH401 33 as references. Co-assembling vaccines V3–V5 were prepared from N-palmitoylated CH401 43 with lipophilic adjuvants, Pam3CSK4 44, a-GalCer 41, or synthetic E. coli lipid A 8, respectively. Co-assembling vaccine V3 promoted a greater anti-CH401 antibody production than V1, indicating Pam3CSK4 44 functioned as an adjuvant in the co-assembly. In contrast, antibody titers of V3 were lower than V2. Pam3CSK4-CH401 33 possessed both self-assembling property for multivalent antigen presentation as well as adjuvanticity derived from a chemically linked TLR2 ligand. We also observed the mixture of non-lipidated CH401 and Pam3CSK4 44 formed a co-assembling structure. Consequently, it promoted IgG production. a-GalCer co-assembling V4 strikingly promoted antibody production, while lipid A co-assembling V5 induced the same level of antibody production to V2. Hence, a-GalCer 41 was found to be more suitable adjuvant for CH401 peptide antigen than Pam3CSK4 and E. coli lipid A. V6 containing palmitoylated Th epitope 45 resulted to be the similar level of antibody production to V3. Since CH401 peptide itself possessed Th epitope, addition of 45 might not influence the immunogenicity. Overall, our co-assembling vaccine was handy prepared and performed self-adjuvanting property. Such next generation vaccine platform opens the door for more concise and effective vaccine development. As described above, conjugation of bioactive molecules can generate new biological functions that the original components do not have. Conjugated molecules are capable of multiple interactions with the target proteins to regulate their functions in spatiotemporal manner. Multiple targeting strategies based on molecular conjugation are expected to reduce side effects and provide high efficacy.

16

Y. Manabe et al.

References 1. Azzi JR, Sayegh MH, Mallat SG (2013) Calcineurin inhibitors: 40 years later, can’t live without. J Immunol 191:5785–5791. https://doi.org/10.4049/jimmunol.1390055 2. Vergote I, Leamon CP (2015) Vintafolide: a novel targeted therapy for the treatment of folate receptor expressing tumors. Adv Med Oncol 7:206–218. https://doi.org/10.1177/ 1758834015584763 3. Reddy JA, Dorton R, Bloomfield A, Nelson M, Dircksen C, Vetzel M, Kleindl P, Santhapuram H, Vlahov IR, Leamon CP (2018) Pre-clinical evaluation of EC1456, a folate-tubulysin anti-cancer therapeutic. Sci Rep 8:8943. https://doi.org/10.1038/s41598-01827320-5 4. Hennrich U, Kopka K (2019) Lutathera®: the first FDA- and EMA-approved radiopharmaceutical for peptide receptor radionuclide therapy. Pharmaceuticals (Basel) 12114. https://doi. org/10.3390/ph12030114 5. Kratochwil C, Bruchertseifer F, Giesel FL, Weis M, Verburg FA, Mottaghy F, Kopka K, Apostolidis C, Haberkorn U, Morgenstern A (2016) 225Ac-PSMA-617 for PSMA-targeted a-radiation therapy of metastatic castration-resistant prostate cancer. J Nucl Med 57:1941– 1944. https://doi.org/10.2967/jnumed.116.178673 6. Coffman RL, Sher A, Seder RA (2010) Vaccine adjuvants: putting innate immunity to work. Immunity 33:492–503. https://doi.org/10.1016/j.immuni.2010.10.002 7. Eisenbarth SC, Colegio OR, O’Connor W, Sutterwala FS, Flavell RA (2008) Crucial role for the Nalp3 inflammasome in the immunostimulatory properties of aluminium adjuvants. Nature 453:1122–1126. https://doi.org/10.1038/nature06939 8. Hornung V, Bauernfeind F, Halle A, Samstad EO, Kono H, Rock KL, Fitzgerald KA, Latz E (2008) Silica crystals and aluminum salts activate the NALP3 inflammasome through phagosomal destabilization. Nat Immunol 9:847–856. https://doi.org/10.1038/ni.1631 9. Kuroda E, Ishii KJ, Uematsu S, Ohata K, Coban C, Akira S, Aritake K, Urade Y, Morimoto Y (2011) Silica crystals and aluminum salts regulate the production of prostaglandin in macrophages via NALP3 inflammasome-independent mechanisms. Immunity 34:514–526. https://doi.org/10.1016/j.immuni.2011.03.019 10. Marichal T, Ohata K, Bedoret D, Mesnil C, Sabatel C, Kobiyama K, Lekeux P, Coban C, Akira S, Ishii KJ, Bureau F, Desmet CJ (2011) DNA released from dying host cells mediates aluminum adjuvant activity. Nat Med 17:996–1002. https://doi.org/10.1038/nm.2403 11. Mata-Haro V, Cekic C, Martin M, Chilton PM, Casella CR, Mitchell TC (2007) The vaccine adjuvant monophosphoryl lipid A as a TRIF-biased agonist of TLR4. Science 316:1628– 1632. https://doi.org/10.1126/science.1138963 12. Campbell JD (2017) Development of the CpG adjuvant 1018: a case study. Methods Mol Biol 1494:15–27. https://doi.org/10.1007/978-1-4939-6445-1_2 13. Tanimura N, Saitoh S, Ohto U, Akashi-Takamura S, Fujimoto Y, Fukase K, Shimizu T, Miyake K (2014) The attenuated inflammation of MPL is due to the lack of CD14-dependent tight dimerization of the TLR4/MD2 complex at the plasma membrane. Int Immunol 26:307– 314. https://doi.org/10.1093/intimm/dxt071 14. Fujimoto Y, Shimoyama A, Saeki A, Kitayama N, Kasamatsu C, Tsutsui H, Fukase K (2013) Innate immunomodulation by lipophilic termini of lipopolysaccharide; synthesis of lipid As from Porphyromonas gingivalis and other bacteria and their immunomodulative responses. Mol BioSyst 9:987–996. https://doi.org/10.1039/c3mb25477a 15. Montminy SW, Khan N, McGrath S, Walkowicz MJ, Sharp F, Conlon JE, Fukase K, Kusumoto S, Sweet C, Miyake K, Akira S, Cotter RJ, Goguen JD, Lien E (2006) Virulence factors of Yersinia pestis are overcome by a strong lipopolysaccharide response. Nat Immunol 7:1066–1073. https://doi.org/10.1038/ni1386 16. Shimoyama A, Saeki A, Tanimura N, Tsutsui H, Miyake K, Suda Y, Fujimoto Y, Fukase K (2011) Chemical synthesis of Helicobacter pylori lipopolysaccharide partial structures and

1 Conjugation Strategies for Development of Bioactive …

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

17

their selective proinflammatory responses. Chem Eur J 17:14464–14474. https://doi.org/10. 1002/chem.201003581 Shibata N, Kunisawa J, Hosomi K, Fujimoto Y, Mizote K, Kitayama N, Shimoyama A, Mimuro H, Sato S, Kishishita N, Ishii KJ, Fukase K, Kiyono H (2018) Lymphoid tissue-resident Alcaligenes LPS induces IgA production without excessive inflammatory responses via weak TLR4 agonist activity. Mucosal Immunol 11:693–702. https://doi.org/10. 1038/mi.2017.103 Yoshii K, Hosomi K, Shimoyama A, Wang Y, Yamaura H, Nagatake T, Suzuki H, Lan H, Kiyono H, Fukase K, Kunisawa J (2020) Chemically synthesized Alcaligenes lipid A shows a potent and safe nasal vaccine adjuvant activity for the induction of Streptococcus pneumoniae-specific IgA and Th17 mediated protective immunity. Microorganisms 8:1102. https://doi.org/10.3390/microorganisms8081102 Wang Y, Hosomi K, Shimoyama A, Yoshii K, Yamaura H, Nagatake T, Nishino T, Kiyono H, Fukase K, Kunisawa J (2020) Adjuvant activity of synthetic lipid A of alcaligenes, a gut-associated lymphoid tissue-resident commensal bacterium, to augment antigen-specific IgG and Th17 responses in systemic vaccine. Vaccines (Basel) 8:395. https://doi.org/10.3390/ vaccines8030395 Ingale S, Wolfert MA, Gaekwad J, Buskas T, Boons G-J (2007) Robust immune responses elicited by a fully synthetic three-component vaccine. Nat Chem Biol 3:663–667. https://doi. org/10.1038/nchembio.2007.25 Kaiser A, Gaidzik N, Becker T, Menge C, Groh K, Cai H, Li YM, Gerlitzki B, Schmitt E, Kunz H (2010) Fully synthetic vaccines consisting of tumor-associated MUC1 glycopeptides and a lipopeptide ligand of the Toll-like receptor 2. Angew Chem Int Ed Engl 49:3688–3692. https://doi.org/10.1002/anie.201000462 Cai H, Sun ZY, Huang ZH, Shi L, Zhao YF, Kunz H, Li YM (2013) Fully synthetic self-adjuvanting thioether-conjugated glycopeptide-lipopeptide antitumor vaccines for the induction of complement-dependent cytotoxicity against tumor cells. Chem Eur J 19:1962– 1970. https://doi.org/10.1002/chem.201203709 Cai H, Chen MS, Sun ZY, Zhao YF, Kunz H, Li YM (2013) Self-adjuvanting synthetic antitumor vaccines from MUC1 glycopeptides conjugated to T-Cell epitopes from tetanus toxoid. Angew Chem Int Ed 52:6106–6110. https://doi.org/10.1002/anie.201300390 Palitzsch B, Hartmann S, Stergiou N, Glaffig M, Schmitt E, Kunz H (2014) A fully synthetic four-component antitumor vaccine consisting of a mucin glycopeptide antigen combined with three different T-Helper-Cell epitopes. Angew Chem Int Ed 53:14245–14249. https://doi.org/ 10.1002/anie.201406843 Chang TC, Manabe Y, Fujimoto Y, Ohshima S, Kametani Y, Kabayama K, Nimura Y, Lin CC, Fukase K (2018) Syntheses and immunological evaluation of self-adjuvanting clustered N-Acetyl and N-Propionyl Sialyl-Tn combined with a T-helper cell epitope as antitumor vaccine candidates. Angew Chem Int Ed 57:8219–8224. https://doi.org/10.1002/ anie.201804437 Feng Q, Manabe Y, Kabayama K, Aiga T, Miyamoto A, Ohshima S, Kametani Y, Fukase K (2019) Syntheses and Functional studies of self-adjuvanting anti-HER2 cancer vaccines. Chem Asian J 14:4268–4273. https://doi.org/10.1002/asia.201901002 Aiga T, Manabe Y, Ito K, Chang TC, Kabayama K, Ohshima S, Kametani Y, Miura A, Furukawa H, Inaba H, Matsuura K, Fukase K (2020) Immunological evaluation of co-assembling a lipidated peptide antigen and lipophilic adjuvants: self-adjuvanting anti-breast-cancer vaccine candidates. Angew Chem Int Ed Engl. https://doi.org/10.1002/ anie.202007999 Khan S, Weterings JJ, Britten CM, de Jong AR, Graafland D, Melief CJ, van der Burg SH, van der Marel G, Overkleeft HS, Filippov DV, Ossendorp F (2009) Chirality of TLR-2 ligand Pam3CysSK4 in fully synthetic peptide conjugates critically influences the induction of specific CD8+ T-cells. Mol Immunol 46:1084–1091. https://doi.org/10.1016/j.molimm.2008. 10.006

18

Y. Manabe et al.

29. Wilkinson BL, Day S, Malins LR, Apostolopoulos V, Payne RJ (2011) Self-adjuvanting multicomponent cancer vaccine candidates combining per-glycosylated MUC1 glycopeptides and the toll-like receptor 2 Agonist Pam3CysSer. Angew Chem Int Ed 50:1635–1639. https:// doi.org/10.1002/anie.201006115 30. Wilkinson BL, Day S, Chapman R, Perrier S, Apostolopoulos V, Payne RJ (2012) Synthesis and immunological evaluation of self-assembling and self-adjuvanting tricomponent glycopeptide cancer-vaccine candidates. Chem Eur J 18:16540–16548. https://doi.org/10. 1002/chem.201202629 31. McDonald DM, Wilkinson BL, Corcilius L, Thaysen-Andersen M, Byrne SN, Payne RJ (2014) Synthesis and immunological evaluation of self-adjuvanting MUC1-macrophage activating lipopeptide 2 conjugate vaccine candidates. Chem Commun 50:10273–10276. https://doi.org/10.1039/c4cc03510k 32. Ashhurst AS, McDonald DM, Hanna CC, Stanojevic VA, Britton WJ, Payne RJ (2019) Mucosal vaccination with a self-adjuvanted lipopeptide is immunogenic and protective against mycobacterium tuberculosis. J Med Chem 62:8080–8089. https://doi.org/10.1021/acs. jmedchem.9b00832 33. Xu Z, Rivera-Hernandez T, Chatterjee O, Walker MJ, Moyle PM (2020) Semisynthetic, self-adjuvanting vaccine development: efficient, site-specific sortase A-mediated conjugation of Toll-like receptor 2 ligand FSL-1 to recombinant protein antigens under native conditions and application to a model group A streptococcal vaccine. J Control Release 317:96–108. https://doi.org/10.1016/j.jconrel.2019.11.018 34. McDonald DM, Hanna CC, Ashhurst AS, Corcilius L, Byrne SN, Payne RJ (2018) Synthesis of a self-adjuvanting MUC1 vaccine via diselenide-selenoester ligation-deselenization. ACS Chem Biol 13:3279–3285. https://doi.org/10.1021/acschembio.8b00675 35. Wang Q, Zhou Z, Tang S, Guo Z (2012) Carbohydrate-monophosphoryl lipid a conjugates are fully synthetic self-adjuvanting cancer vaccines eliciting robust immune responses in the mouse. ACS Chem Biol 7:235–240. https://doi.org/10.1021/cb200358r 36. Zhou Z, Mondal M, Liao G, Guo Z (2014) Synthesis and evaluation of monophosphoryl lipid A derivatives as fully synthetic self-adjuvanting glycoconjugate cancer vaccine carriers. Org Biomol Chem 12:3238–3245. https://doi.org/10.1039/c4ob00390j 37. Liao G, Zhou Z, Suryawanshi S, Mondal MA, Guo Z (2016) Fully synthetic self-adjuvanting alpha-2,9-Oligosialic acid based conjugate vaccines against group C meningitis. ACS Cent Sci 2:210–218. https://doi.org/10.1021/acscentsci.5b00364 38. Wang L, Feng S, Wang S, Li H, Guo Z, Gu G (2017) Synthesis and immunological comparison of differently linked lipoarabinomannan oligosaccharide-monophosphoryl lipid A conjugates as antituberculosis vaccines. J Org Chem 82:12085–12096. https://doi.org/10. 1021/acs.joc.7b01817 39. Lewicky JD, Ulanova M, Jiang ZH (2016) Synthesis of a TLR4 agonist-carbohydrate antigen conjugate as a self-adjuvanting cancer vaccine. ChemistrySelect 1:906–910. https://doi.org/ 10.1002/slct.201600230 40. Reintjens NRM, Tondini E, de Jong AR, Meeuwenoord NJ, Chiodo F, Peterse E, Overkleeft HS, Filippov DV, van der Marel GA, Ossendorp F, Codée JDC (2020) Self-adjuvanting cancer vaccines from conjugation-ready lipid A analogues and synthetic long peptides. J Med Chem 63:11691–11706. https://doi.org/10.1021/acs.jmedchem.0c00851 41. Cavallari M, Stallforth P, Kalinichenko A, Rathwell DC, Gronewold TM, Adibekian A, Mori L, Landmann R, Seeberger PH, De Libero G (2014) A semisynthetic carbohydrate-lipid vaccine that protects against S. pneumoniae in mice. Nat Chem Biol 10:950–956. https://doi. org/10.1038/nchembio.1650 42. Anderson RJ, Compton BJ, Tang CW, Authier-Hall A, Hayman CM, Swinerd GW, Kowalczyk R, Harris P, Brimble MA, Larsen DS, Gasser O, Weinkove R, Hermans IF, Painter GF (2015) NKT cell-dependent glycolipid-peptide vaccines with potent anti-tumour activity. Chem Sci 26:5120–5127. https://doi.org/10.1039/c4sc03599b

1 Conjugation Strategies for Development of Bioactive …

19

43. Compton BJ, Tang CW, Johnston KA, Osmond TL, Hayman CM, Larsen DS, Hermans IF, Painter GF (2015) Synthesis and activity of 6”-deoxy-6”-thio-a-GalCer and peptide conjugates. Org Lett 17:5954–5957. https://doi.org/10.1021/acs.orglett.5b02836 44. Yin XG, Chen XZ, Sun WM, Geng XS, Zhang XK, Wang J, Ji PP, Zhou ZY, Baek DJ, Yang GF, Liu Z, Guo J (2017) IgG antibody response elicited by a fully synthetic two-component carbohydrate-based cancer vaccine candidate with a-galactosylceramide as built-in adjuvant. Org Lett 19:456–459. https://doi.org/10.1021/acs.orglett.6b03591 45. Chen PG, Hu HG, Sun ZY, Li QQ, Zhang BD, Wu JJ, Li WH, Zhao YF, Chen YX, Li YM (2020) Fully synthetic invariant NKT cell-dependent self-adjuvanting antitumor vaccines eliciting potent immune response in mice. Mol Pharm 17(2):417–425. https://doi.org/10.1021/ acs.molpharmaceut.9b00720 46. Gential GPP, Hogervorst TP, Tondini E, van de Graaff MJ, Overkleeft HS, Codée JDC, van der Marel GA, Ossendorp F, Filippov DV (2019) Peptides conjugated to 2-alkoxy-8-oxo-adenine as potential synthetic vaccines triggering TLR7. Bioorg Med Chem Lett 29:1340–1344. https://doi.org/10.1016/j.bmcl.2019.03.048 47. De Silva RA, Wang Q, Chidley T, Appulage DK, Andreana PR (2009) Immunological response from an entirely carbohydrate antigen: design of synthetic vaccines based on Tn-PS A1 conjugates. J Am Chem Soc 131:9622–9623. https://doi.org/10.1021/ja902607a 48. Berti F, Adamo R (2013) Recent mechanistic insights on glycoconjugate vaccines and future perspectives. ACS Chem Biol 8:1653–1663. https://doi.org/10.1021/cb400423g 49. Shi M, Kleski KA, Trabbic KR, Bourgault J-P, Andreana PR (2016) Sialyl-Tn polysaccharide A1 as an entirely carbohydrate immunogen: synthesis and immunological evaluation. J Am Chem Soc 138:14264–14272. https://doi.org/10.1021/jacs.6b05675 50. BenMohamed L, Wechsler SL, Nesburn AB (2002) Lipopeptide vaccines–yesterday, today, and tomorrow. Lancet Infect Dis 2:425–431. https://doi.org/10.1016/S1473-3099(02)00318-3 51. Rouaix F, Gras-Masse H, Mazingue C, Diesis E, Ridel PR, Estaquier J, Capron A, Tartar A, Auriault C (1994) Effect of a lipopeptidic formulation on macrophage activation and peptide presentation to T cells. Vaccine 12:1209–1214. https://doi.org/10.1016/0264-410X(94) 90245-3 52. Rudra JS, Sun T, Bird KC, Daniels MD, Gasiorowski JZ, Chong AS, Collier JH (2012) Modulating adaptive immune responses to peptide self-assemblies. ACS Nano 6:1557–1564. https://doi.org/10.1021/nn204530r 53. Rudra JS, Tian YF, Jung JP, Collier JH (2010) A self-assembling peptide acting as an immune adjuvant. Proc Natl Acad Sci USA 107(2):622–627. https://doi.org/10.1073/pnas.0912124107 54. Jung JP, Nagaraj AK, Fox EK, Rudra JS, Devgun JM, Collier JH (2009) Co-assembling peptides as defined matrices for endothelial cells. Biomaterials 30:2400–2410. https://doi.org/ 10.1016/j.biomaterials.2009.01.033 55. Tian YF, Hudalla GA, Han H, Collier JH (2013) Controllably degradable b-sheet nanofibers and gels from self-assembling depsipeptides. Biomater Sci 1:1037–1045. https://doi.org/10. 1039/C3BM60161G 56. Marini DM, Hwang W, Lauffenburger DA, Zhang S, Kamm RD (2002) Left-Handed Helical Ribbon Intermediates in the Self-Assembly of a b-Sheet Peptide. Nano Lett 2:295–299. https://doi.org/10.1021/nl015697g 57. Sieminski AL, Semino CE, Gong H, Kamm RD (2008) Primary sequence of ionic self-assembling peptide gels affects endothelial cell adhesion and capillary morphogenesis. J Biomed Mater Res, Part A 87A:494–504. https://doi.org/10.1002/jbm.a.31785 58. Yokoi H, Kinoshita T, Zhang S (2005) Dynamic reassembly of peptide RADA16 nanofiber scaffold. Proc Natl Acad Sci USA 102:8414–8419. https://doi.org/10.1073/pnas.0407843102 59. Cho H, Balaji S, Sheikh AQ, Hurley JR, Tian YF, Collier JH, Crombleholme TM, Narmoneva DA (2012) Regulation of endothelial cell activation and angiogenesis by injectable peptide nanofibers. Acta Biomater 8:154–164. https://doi.org/10.1016/j.actbio. 2011.08.029

20

Y. Manabe et al.

60. Gandhi A, Balmer P, York LJ (2016) Characteristics of a new meningococcal serogroup B vaccine, bivalent rLP2086 (MenB-FHbp; Trumenba®). Postgrad Med 128:548–556. https:// doi.org/10.1080/00325481.2016.1203238 61. Fujimoto Y, Hashimoto M, Furuyashiki M, Katsumoto M, Seya T, Suda Y, Fukase K (2009) Lipopeptides from Staphylococcus aureus as Tlr2 Ligands: prediction with mrna expression, chemical synthesis, and immunostimulatory activities. ChemBioChem 10:2311–2315. https:// doi.org/10.1002/cbic.200900242 62. Azuma M, Sawahata R, Akao Y, Ebihara T, Yamazaki S, Matsumoto M, Hashimoto M, Fukase K, Fujimoto Y, Seya T (2010) The peptide sequence of diacyl lipopeptides determines dendritic cell TLR2-mediated NK activation. PLoS ONE 5: https://doi.org/10.1371/journal. pone.0012550 63. Holmberg LA, Sandmaier BM (2004) Vaccination with Theratope (STn-KLH) as treatment for breast cancer. Expert Rev Vaccines 3:655–663. https://doi.org/10.1586/14760584.3.6.655 64. Sahabuddin SK, Chang TC, Lin CC, Jan FD, Hsiao HY, Huang KT, Chen JH, Horng JC, Ho JAA, Lin CC (2010) Synthesis of N-modified sTn analogs and evaluation of their immunogenicities by microarray-based immunoassay. Tetrahedron 66:7510–7519. https://doi. org/10.1016/j.tet.2010.07.050 65. Salmasan RM, Manabe Y, Kitawaki Y, Chang TC, Fukase K (2014) Efficient glycosylation using In(OTf)3 as a lewis acid: activation of N-phenyltrifluoroacetimidate or thioglycosides with halogenated reagents or PhIO. Chem Lett 43:956–958. https://doi.org/10.1246/cl.140167 66. Fujimoto Y, Kimura E, Murata S, Kusumoto S, Fukase K (2006) Synthesis and bioactivity of fluorescence- and biotin-labeled lipid a analogues for investigation of recognition mechanism in innate immunity. Tetrahedron Lett 47:539–543. https://doi.org/10.1016/j.tetlet.2005.11.041

Chapter 2

Total Synthesis, Biological Evaluation, and 3D Structural Analysis of a Cyclodepsipeptide Natural Product Takayuki Doi

Abstract Cyclodepsipeptide natural products have been widely used in the production of anticancer drugs because of their high potency and structural diversity. Therefore, it is important to obtain a comprehensive understanding of their structure and function. As such, this study presents the total synthesis of apratoxin A, which is a middle-sized cyclodepsipeptide natural product. It also describes the solid-phase synthesis of apratoxin A derivatives, chemical probe synthesis, and its protein-network analysis. Moreover, the design and synthesis of apratoxin A mimetics, their structure-activity relationships, and 3D structural analysis are presented. It is worth acknowledging that obtaining a predictable method for 3D analysis of macrocycles will facilitate the design and synthesis of biologically active macrocycles in the future.













Keywords Apratoxin Cyclic peptide Cytotoxicity Macrocycle Total synthesis Combinatorial synthesis Solid-phase synthesis Conformational analysis




2.1







Introduction

The synthesis of combinatorial libraries for obtaining appropriate drugs has focused on heterocyclic compounds. It can be a powerful method if applied to a structurally complex molecule such as biologically active natural products. As numerous bioactive compounds isolated from natural secondary metabolites have unique structures with extreme potency, their unique skeletons are also of significant interest for the expansion of a unique compound library. For example, a compound library can be constructed in which synthetic blocks are combined to synthesize T. Doi (&) Graduate School of Pharmaceutical Sciences, Tohoku University, Sendai, Japan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_2

21

22

T. Doi

analogs based on the skeletal synthesis of natural products; thus, the diversity of compounds can be increased. The author is particularly interested in cyclic peptide compounds. The reason is that it has excellent structural diversity, as such, many compounds with unique biological activities have been isolated and structurally determined from natural sources as middle-size molecules [1]. The macrocyclic structure has numerous features, for example, the conformation of macrocyclic compounds is fairly restricted compared to linear compounds but still somewhat flexible. This is in contrast to rigid compounds, such as small-ring or fused-ring-containing molecules. Thereby, the macrocyclic compounds can take a well-fit conformation for binding to a drug target without restriction [2]. Structural optimization can be achieved easily by altering the amino acid side chain components. Among the cyclic peptides, bioactive cyclodepsipeptides are of special interest because they have hydroxy-carboxylic acids functionalized. In addition, they contain proteinogenic and unusual amino acids containing D- and/or N-methyl amino acids as components, potentially covering a substantial area of the chemical space. It is therefore important to understand the function and shape of cyclodepsipeptide natural products and their potent biological activities. Herein this research describes apratoxin A.

2.2 2.2.1

Total Synthesis of Apratoxin A Apratoxin A

Apratoxin A (1) is a 25-membered ring depsipeptide isolated from marine cyanobacteria Lyngbya majuscula and structurally determined by Luesch et al. in 2001 [3] (Fig. 2.1). 1 exhibits potent cytotoxicity against cancer cells and in vivo anti-tumor activity in mice. 1 consists of three methylated amino acid residues (N-

Fig. 2.1 Structures of apratoxins A, C, D, and E

2 Total Synthesis, Biological Evaluation, and 3D Structural …

23

methylisoleucine, N-methylalanine, and O-methyltyrosine) and a,b-unsaturated cysteine derivatives, fatty acid with four asymmetric centers, and proline. In 2003, Chen and Forsyth reported the first total synthesis of 1 [4]. The authors and Ma et al. reported the total synthesis and its derivatives [5, 6]. Other related compounds, namely apratoxins C, D, and E were also studied [7–10] (Fig. 2.1). The mechanism of action has also been studied by several researchers. It was proposed that the potent cytotoxicity is based on the inhibition of the secretory pathway by preventing cotranslational translocation of newly synthesized proteins [11], and further study indicated that 1 can bind a Sec61 protein complex that is the translocation channel localized in the ER membrane [12, 13]. 1 exhibited potent cytotoxicity against a wide range of cancer cell lines and in vivo anti-tumor activity for HCT-116 xenograft models with in vivo toxicity [3, 14] that would be caused by pancreas atrophy by high 1 exposure [12]. In contrast, Luesch et al. created apratoxins S8 and S10 that have a saturated amide instead of the a,b-unsaturated amide in the modified Cys moiety in 1. Those mimetics presented tumor suppression in HCT-116 xenograft mouse models without weight loss [15, 16]. Therefore, apratoxin mimetics would be useful in the development of anticancer drugs.

2.2.2

Our Fatty Acid Synthesis

Enzymatic desymmetrization of 3-methylglutaric anhydride (2) with 1-propanol in the presence of lipase PS afforded 3 in quantitative yield with high enantioselectivity (93% ee) (Scheme 2.1). After conversion to t-butyl ketone 4, asymmetric CBS reduction provided alcohol 5 at a ratio of 99:1. The 0.3 equivalent of Me-CBS and five equivalents of catecholborane are necessary to induce this high stereoselection [17]. Conversion to aldehyde 6 and asymmetric crotylation, followed by Troc-protection afforded anti-product 7 with >95% diastereoselectivity. Treatment of 7 with DDQ generated secondary alcohol, which underwent Yamaguchi esterification, led to proline ester. The terminal alkene moiety was oxidatively cleaved using cat. OsO4/oxone, and the partially produced hydroxy ketone was further oxidatively cleaved using NaIO4 to provide acid 8 [18].

2.2.3

Optimized Synthetic Route for 1

The optimized synthetic route involved coupling of acid 8 with cysteine-modified unit 9, formation of thiazoline using Kelly’s method (Tf2O/Ph3P = O), and removal of the Troc group and allyl ester to furnish acid 10 (Scheme 2.1) [5]. Elimination of the Troc group was observed during column chromatography after formation of the thiazoline ring. Therefore, immediate removal of the Troc group for the crude

24

Scheme 2.1 Optimized synthetic route for 1

T. Doi

2 Total Synthesis, Biological Evaluation, and 3D Structural …

25

product was essential. Coupling of the acid 10 with tripeptide 11 was performed using HATU/DIEA and sequential removal of the ally ester and the Fmoc group at both the C and N termini provided linear precursor for macrocyclization. Macrolactamization was successfully performed using HATU/DIEA under 1 mM high dilution conditions to obtain 1 in 72% yield. Thus, the optimized synthetic route still requires 21 steps from commercially available 3-methylglutaric anhydride, however, 1 was obtained in sufficient overall yield (16%).

2.3 2.3.1

Structure-Activity Relationships of 1 and the Synthesis of Its Chemical Probe Solid-Phase Synthesis of Apratoxin A Analogs and Their Biological Evaluation

To prepare various analogs of 1, solid-phase synthesis was performed (Scheme 2.2) [19]. After chlorination of trityl alcohol SynPhaseTM lanterns, Fmoc-AA3-OH was loaded. Then, Fmoc-AA2-OH and Fmoc-AA1-OH were sequentially coupled using PyBrop/DIEA. As the amino acids, AA2 and AA1 were N-methylamino acids, this study investigated other condensation agents (HBTU, HATU, and TFFH). PyBrop produced the best results and double coupling was performed. Thereafter, coupling of polymer-supported tripeptide with acid 10 was performed using PyAOP/DIEA. Although acid 10 is a relatively large coupling component and it appeared to be less reactive, the reaction was completed in 24 h. Following the removal of the Fmoc group, the linear precursor was cleaved from lanterns with 30% hexafluoroisopropyl alcohol (HFIPA) in CH2Cl2. Macrolactamization was conducted in solution, as stated, to furnish the desired apratoxin A analogs 12, which were diastereomeric mixtures (9:1 * 7:3) at the 34 position possibly because the epimerization would occur during the reaction of the Fmoc cleavage on a polymer-support. The cytotoxicities of 1, its 34-epimer, and azido-functionalized analogs 12a–12e were evaluated against HeLa cells and the IC50 values are depicted in Table 2.1. Interestingly, 34-epi-1 is as potent as 1 (entry 1 vs. entry 2). Therefore, the configuration of the methyl group at the 34 position does not affect the biological activity. The alternation of the s-Bu group in MeIle to a 4-azidobutyl group slightly decreased the potency (entry 3). In contrast, the cytotoxicity of 12b and 12c decreased tenfold where the a-methyl group in MeAla was substituted to azidomethyl and 4-azidobutyl groups (entries 4 and 5). Remarkably, 12d and 12e retained the cytotoxicity. It is worth noting that a hydrophobic 7-azidoheptyl linker was found to be efficient (entry 6 vs. entry 7).

26

T. Doi

Scheme 2.2 Solid-phase synthesis of apratoxin A analogs 12

Table 2.1 Cytotoxicity of 1 and its analogs 12 against HeLa cells Entry

Compound

R1

R2

R3

IC50/lM

1 2 3 4 5 6 7

1 34-epi-1 12a 12b 12c 12d 12e

Me Me Me Me Me (CH2)7–N3 (CH2CH2O)2–CH2CH2N3

Me Me Me CH2N3 (CH2)4–N3 Me Me

s-Bu s-Bu (CH2)4–N3 s-Bu s-Bu s-Bu s-Bu

0.19 0.19 0.45 2.1 1.7 0.08 0.35

2.3.2

Chemical Probe Synthesis and Binding Protein Analysis

According to the results in Table 2.1, a chemical probe for a target search was prepared using 12d. A FLAG tag was selected in the preparation of the chemical probe. The FLAG-tagged apratoxin A derivative 14 was prepared by the Huisgen 1,3-dipolar cycloaddition of 12d and acetylene-containing FLAG-tagged 13 under mild reaction conditions, such as CuSO4/sodium ascorbate/Na2HPO4 in t-BuOH/ H2O [19] (Scheme 2.3). In the Natsume group, 14 was applied to HEK293 cell

2 Total Synthesis, Biological Evaluation, and 3D Structural …

27

lysate to capture interacting proteins. The FLAG-antibody beads purified the binding proteins by immunoprecipitation, which were identified by trypsin digestion, followed by sequencing analysis using a highly sensitive direct nanoflow LC/ MS/MS system [20–22]. This approach was followed four times, and the 57 proteins detected three times or more are listed in Table 2.2 [23]. Among them, 38 proteins were present in mitochondria, therefore, HeLa cells were treated with 1 in the presence of Rodamin 123 and it was discovered that the electron transport system in mitochondria was inhibited in a dose-dependent manner. In contrast, as described in the introduction, a protein complex of Sec61 was reported to be a target protein of 1 [12], the gamma subunit of Sec61 had been on the list.

Scheme 2.3 Preparation of FLAG-tagged apratoxin A derivative 14

28

T. Doi

Table 2.2 List of binding proteins detected by direct nanoflow LC/MS/MS analysis Gene symbol ABCD3 ACBD5 AGK ATAD3A or ATAD3B ATP5A1 ATP5B ATP5O CALM2 or CALM3 CBARA1 CKAP4

2.4

EFHA1 ERLIN2 GK GPSN2

MLSTD1 MLSTD2 NDUFA10 NDUFA12

NDUFB3 NDUFB6 NDUFC2 NDUFS1

PTPLAD1 QPCTL REEP4 SCCPDH

SOAT1 STOML2 UGT8 UQCRC2

HADHA HADHB HSD17B12 IMMT LETM1 LETMD1

NDUFA5 NDUFA7 NDUFA8 NDUFA9 NDUFB1 NDUFB10

NDUFS3 NDUFS4 NDUFV1 NDUFV2 PHB PHB2

SCD SEC61G SGPL1 SLC25A3 SLC25A5 SLC25A6

VDAC1 VDAC2 VDAC3

Solid-Phase Synthesis of Apratoxin A Mimetics

As 1 demonstrates a narrow therapeutic window, it was presumed to be because of the presence of the MoCys moiety that includes a Michael acceptor and a thiazoline ring. Therefore, the research intended to replace the MoCys moiety (Unit A) with simpler amino acids, such as 3-(aminomethyl)benzoic acid (A1), 4-aminbutanoic acid (A2), and piperidine-4-carboxylic acid (A3) as the mimetics may retain their conformations similar to that of 1. Meanwhile, alternation of the MeAla–MeIle moiety (B1) (Unit B) with 3-(aminomethyl)benzoic acid (B2), trans4-aminocyclohexanecarboxylic acid (B3), and cis-4-aminocyclohexanecarboxylic acid (B4) were attempted. According to the aforementioned synthetic method, the combinatorial synthesis of apratoxin mimetics 15 was performed using the SynPhase lanterns which incorporated a split and mix method with colored cogs for decoding (Scheme 2.4) [24]. All compounds attempted were provided in a ca 5– 10 mg scale from two double lanterns (24.5 lmol/lantern). Notably, the cytotoxicity of 15{A, B} against HCT-116 cells indicated that the Unit A can be replaced by A1 and A3 with an IC50 value of less than 1 lM, whereas Unit B cannot be changed (Table 2.3). Encouraged by these results, the Tyr(Me)–MeAla–MeIle moiety was further optimized based on 15{A3, B1} which is referred to as apratoxin M7 [25]. Through modification of the three amino acids by changing the size of the side chain of apratoxin M7, nine mimetics 16a–16i were further synthesized and their cytotoxicities against HCT-116 cells are depicted in Table 2.4. For MeAla and MeIle, the activity decreased when the size of the side chain substituents was modified. Remarkably, the activity significantly increased by replacing the 4-methoxyphenyl group of Tyr(Me) with a 4-chlorophenyl or biphenyl group. Particularly in the latter case, the cytotoxicity of apratoxin M16 (16i) was found to be as potent as that of 1. Subsequently, the growth inhibition activities

2 Total Synthesis, Biological Evaluation, and 3D Structural …

29

Scheme 2.4 Solid-phase combinatorial synthesis of apratoxin mimetics 15{A1–A3, B1–B4}

were evaluated against various cancer cell lines. The results are depicted in Table 2.5. The activity of 16i is primarily comparable to that of 1. Therefore, the target molecule of 16i should be identical to that of 1 [25].

2.5

3D Analysis of Apratoxin A and Apratoxin M16

Conformational analysis of apratoxin A (1) and apratoxin M16 (16i) was performed using a distance geometry method. The ROEs and 3JH,H coupling constants of 1 and 16i were measured in CD3CN to obtain the information of the corresponding 1H–1H distances and dihedral angles. 20,000 conformations generated by Monte Carlo-based torsional sampling in each compound were energy minimized by constraining the aforementioned 1H–1H distances and dihedral angles on the MacroModel [26]. The typical conformers observed in solution are illustrated in Fig. 2.2. The top half moieties (Tyr(Me)–MeAla–MeIle–Pro and tBu group) are well-fitted between 1 and 16i. The figures are in agreement with the results of the structure-activity relationships discussed in 2.3.1 and 2.4. In the 1H NMR spectra,

30

T. Doi

Table 2.3 Overall yields of apratoxin mimetics 15{A1–A3, B1–B4} and their cytotoxicitiesagainst HCT-116 cells (IC50/lM) HN

N

NH

NH

N

O O

O

O

HN

O

O

B1

B2

B3

25%

24%

16%

33%

0.82

> 10

> 10

> 10

27%

26%

7%

29%

2.6

> 10

5.7

> 10

17%

20%

12%

24%

0.12

4.4

> 10

> 10

B4

A1

O

NH

A2

O N A3

(apratoxin M7)

both trans- and cis-amide bonds between Tyr(Me) and MeAla were obtained. Essentially, two signals corresponding to the N-methyl groups of MeAla were observed in 1H NMR spectra. The N-methyl signal in a cis-amide conformer is high-field shifted compared to that in a trans-amide one because of shielding by the magnetic anisotropic effect of an amide bond. Interestingly, the ratios of the transand cis-amide conformers do not correlate with the cytotoxicity against HCT-116 cells (>9:1 in 1; 1:2 in 16i) [25]. Therefore, elucidation of the binding conformation of apratoxins will be an additional issue, which can be solved by complexation of apratoxins and a human Sec61 complex, followed by X-ray crystallography or cryo-electron microscopy.

2 Total Synthesis, Biological Evaluation, and 3D Structural …

31

Table 2.4 Cytotoxicities of apratoxin mimetics 16a–16i designed based on apratoxin M7 against HCT-116 cells R1

R2

N

HN O

O

N

O

R3

O O

N

N

O HO

O

apratoxin M7 and 16a–16i

Apratoxin M7 (15{A3, B1}) 16a 16b 16c 16d 16e 16f 16g 16h Apratoxin M16 (16i)

Table 2.5 Growth inhibition activities of 1 and 16i against various cancer cell lines

R1

R2

R3

IC50/lM

OMe OMe OMe OMe OMe OMe OMe H Cl Ph

Me Me Me Me H i-Bu PhCH2 Me Me Me

s-Bu Me i-Pr PhCH2 s-Bu s-Bu s-Bu s-Bu s-Bu s-Bu

0.12 7.5 0.13 0.71 0.85 0.36 1.6 0.49 0.069 0.0011

Cancer cell line Tissue

GI50 (lM)a 1

16i

HCT-116 Colon BxPC-3 Pancreas A549 Lung HuH-7 Liver MKN74

0.011

0.011

0.0049

0.0040

0.0064

0.0063

0.0072

0.0068

0.0097

0.0080 (continued)

32 Table 2.5 (continued)

T. Doi Cancer cell line Tissue

GI50 (lM)a 1

16i

Stomach U-87 MG 0.018 0.012 Brain SK-OV-3 0.031 0.018 Ovary HEC-6 0.039 0.023 Uterus 786-O 0.041 0.12 Kidney MCF7 >1 >1 Breast a GI50 = compound concentration at which a growth inhibition of 50% is achieved

Fig. 2.2 3D-structures of 1 and 16i (trans-amide conformer) in CD3CN based on a distance geometry method

2 Total Synthesis, Biological Evaluation, and 3D Structural …

2.6

33

Conclusion

Cyclic peptide natural products are of significant biological interest because of their large chemical diversity, potent activities, and unique structural properties. Currently, cyclopeptides and cyclodepsipeptides can be synthesized in solution and solid phase. Macrocycles are highly flexible and have a variety of conformers dependent on solvents with relatively high energy barriers compared to that of linear molecules, making them difficult to predict. It is necessary to obtain a predictable method for 3D analysis of macrocycles that will allow the designing and synthesis of biologically active macrocycles in the future. Acknowledgements The author thanks all the coworkers listed in the references. This work was supported by grants from the MEXT (Nos. 23,310,145, and 26,282,208), JSPS (JP15H05837 in Middle Molecular Strategy), the Naito Foundation (2009), and Astellas Foundation for Research on Metabolic Disorders (2011).

References 1. Marsault E, Peterson ML (eds) (2017) Practical medicinal chemistry with macrocycles. John Wiley & Sons Inc, NJ 2. Giordanetto F, Kihlberg J (2014) Macrocyclic drugs and clinical candidates: What can medicinal chemists learn from their properties? J Med Chem 57:278–295. https://doi.org/10. 1021/jm400887j 3. Luesch H, Yoshida WY, Moore RE, Paul VJ, Corbett TH (2001) Total structure determination of apratoxin A, a potent novel cytotoxin from the marine cyanobacterium Lyngbya majuscule. J Am Chem Soc 123:5418–5423. https://doi.org/10.1021/ja010453j 4. Chen J, Forsyth CJ (2003) Total synthesis of apratoxin A. J Am Chem Soc 125:8734–8735. https://doi.org/10.1021/ja036050w 5. Doi T, Numajiri Y, Munakata A, Takahashi T (2006) Total synthesis of apratoxin A. Org Lett 8:531–534. https://doi.org/10.1021/ol052907d 6. Ma D, Zou B, Cai G, Hu X, Liu JO (2006) Total synthesis of the cyclodepsipeptide apratoxin A and its analogues and assessment of their biological activities. Chem Eur J 12:7615–7626. https://doi.org/10.1002/chem.200600599 7. Masuda Y, Suzuki J, Onda Y, Fujino Y, Yoshida M, Doi T (2014) Total synthesis and conformational analysis of apratoxin C. J Org Chem 79:8000–8009. https://doi.org/10.1021/ jo501130b 8. Robertson BD, Wengryniuk SE, Coltart DM (2012) Asymmetric total synthesis of apratoxin D. Org Lett 14:5192–5195. https://doi.org/10.1021/ol302309c 9. Wu P, Cai W, Chen QY, Xu S, Yin R, Li Y, Zhang W, Luesch H (2016) Total synthesis and biological evaluation of apratoxin E and its C30 epimer: configurational reassignment of the natural product. Org Lett 18:5400–5403. https://doi.org/10.1021/acs.orglett.6b02780 10. Rastelli EJ, Coltart DM (2018) Synthesis and biological activity of apratoxin derivatives. Tetrahedron 74:2269–2290. https://doi.org/10.1016/j.tet.2017.11.004 11. Liu Y, Law BK, Luesch H (2009) Apratoxin a reversibly inhibits the secretory pathway by preventing cotranslational translocation. Mol Pharmacol 76:91–104. https://doi.org/10.1124/ mol.109.056085 12. Huang KC, Chen Z, Jiang Y, Akare S, Kolber-Simonds D, Condon K, Agoulnik S, Tendyke K, Shen Y, Wu KM, Mathieu S, Hw Choi, Zhu X, Shimizu H, Kotake Y,

34

13.

14.

15.

16.

17.

18. 19.

20.

21.

22.

23. 24.

25.

26.

T. Doi Gerwick WH, Uenaka T, Woodall-Jappe M, Nomoto K (2016) Apratoxin A shows novel pancreas-targeting activity through the binding of Sec 61. Mol Cancer Ther 15:1208–1216. https://doi.org/10.1158/1535-7163.MCT-15-0648 Puyenbroeck VV, Vermeire K (2018) Inhibitors of protein translocation across membranes of the secretory pathway: novel antimicrobial and anticancer agents. Cell Mol Life Sci 75:1541– 1558. https://doi.org/10.1007/s00018-017-2743-2 Tidgewell K, Engene N, Byrum T, Media J, Doi T, Valeriote FA, Gerwick WH (2010) Evolved diversification of a modular natural product pathway: Apratoxins F and G, two cytotoxic cyclic depsipeptides from a Palmyra collection of Lyngbya bouillonii. ChemBioChem 11:1458–1466. https://doi.org/10.1002/cbic.201000070 Chen QY, Liu Y, Cai W, Luesch H (2014) Improved total synthesis and biological evaluation of potent apratoxin S4 based anticancer agents with differential stability and further enhanced activity. J Med Chem 57:3011–3029. https://doi.org/10.1021/jm4019965 Cai W, Ratnayake R, Gerber MH, Chen QY, Yu Y, Derendorf H, Trevino JG, Luesch H (2019) Development of apratoxin S10 (Apra S10) as an anti-pancreatic cancer agent and its preliminary evaluation in an orthotopic patient-derived xenograft (PDX) model. Invest New Drugs 37:364–374. https://doi.org/10.1007/s10637-018-0647-0 Onda Y, Fukushi K, Ohsawa K, Yoshida M, Masuda Y, Doi T (2020) Synthesis of a biphenylalanine analogue of apratoxin A displaying substantially enhanced cytotoxicity. Heterocycles 101:679–691. https://doi.org/10.3987/COM-19-S(F)35 Numajiri Y, Takahashi T, Doi T (2009) Total synthesis of (−)-apratoxin A, 34-epimer, and its oxazoline analogue. Chem Asian J 4:111–125. https://doi.org/10.1002/asia.200800365 Doi T, Numajiri Y, Takahashi T, Takagi M, Shin-ya K (2011) Solid-phase total synthesis of (−)-apratoxin A and its analogues and their biological evaluation. Chem Asian J 6:180–188. https://doi.org/10.1002/asia.201000549 Natsume T, Yamauchi Y, Nakayama H, Shinkawa T, Yanagida M, Takahashi N, Isobe T (2002) A direct nanoflow liquid chromatography–tandem mass spectrometry system for interaction proteomics. Anal Chem 74:4725–4733. https://doi.org/10.1021/ac020018n Shinya K, Natsume T, Doi T (2005) Labeling substance and chimera substance, process for preparing these substances, and method of biosubstance trapping, structural analysis or/and identification with use of the labeling substance. PCT Int Appl WO 2005094187:A2 Hayakawa N, Noguchi M, Takeshita S, Eviryanti A, Seki Y, Nishio H, Yokoyama R, Noguchi M, Shuto M, Shima Y, Kuribayashi K, Kageyama S, Eda H, Suzuki M, Hatta T, Iemura S, Natsume T, Tanabe I, Nakagawa R, Shiozaki M, Sakurai K, Shoji M, Andou A, Yamamoto T (2014) Structure–activity relationship study, target identification, and pharmacological characterization of a small molecular IL-12/23 inhibitor, APY0201. Bioorg Med Chem 22:3021–3029. https://doi.org/10.1016/j.bmc.2014.03.036 Numajiri Y (2009) Ph.D. thesis, Tokyo Institute of Technology, TT00009918 Parsons JG, Sheehan CS, WuIan Z, James IW, Bray AM (2003) A review of solid-phase organic synthesis on SynPhaseTM lanterns and SynPhaseTM crowns. Method Enzymol 369:39–74. https://doi.org/10.1016/S0076-6879(03)69003-8 Onda Y, Masuda Y, Yoshida M, Doi T (2016) Conformation-based design and synthesis of apratoxin A mimetics modified at the a, b-unsaturated thiazoline moiety. J Med Chem 60:6751–6765. https://doi.org/10.1021/acs.jmedchem.7b00833 MacroModel, version 9.9, Schrödinger, Inc., New York

Chapter 3

Development of Middle-Size Molecules for Alkylation to Higher-Order Structures of Nucleic Acids Fumi Nagatsugi and Kazumitsu Onizuka

Abstract DNA and RNA can adopt many varieties of stable higher-order structure motifs, such as G-quadruplexes (G4s), mismatch, and bulge, in addition to the canonical duplex DNA. Many of these secondary structures were found to be closely related to the control of the gene expression. Therefore, the higher-order structure of nucleic acids is one of the candidates for therapeutic targets. Due to the therapeutic potentials, efforts for the development of small molecule binders to specifically target the higher-order structures of nucleic acids have been intensely carried out. With the aim to augment the stabilization effect, selective alkylation using small molecules targeting the higher-order structures of nucleic acids have also been pursued. In this review, we describe the development of middle-size molecules for alkylation to higher-order structures of nucleic acids. We have designed molecules for the selective alkylation to the higher-order structures of nucleic acids and these molecules consist of a binding group to target the nucleic acids structure and the alkylating moiety. These synthesized molecules exhibited an efficient reactivity to thymine in the target higher-order structures of the nucleic acids. Keywords Higher-order structures of nucleic acids structure Abasic site G-quadruplex








Alkylation
Mismatch

F. Nagatsugi (&)
K. Onizuka Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, 2-1-1 Katahira, Aoba-Ku, Sendai, Miyagi 980-8577, Japan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_3

35

36

F. Nagatsugi and K. Onizuka

3.1

Introduction

Genetic information is stored in the primary sequences of DNA and transferred through replication, transcription, and translation to provide the protein synthesis. Recently, in addition to the primary sequences, the higher-order structures of nucleic acids play critical roles in the regulation of transcription and translation. Figure 3.1 shows the representative higher order structures for DNA and RNA. Duplex DNA typically forms a right-handed double helix structure (B type DNA) supported by adenine (A)–thymine (T) and guanine (G)–cytosine (C) base pairing (Fig. 3.1a). In the duplex DNA, the loss of a nucleobase by hydrolysis generates an abasic (AP) site, which can lead to mutations and strand breaks [1, 2]. Mismatch base pairs occur in the replication of DNA to induce errors and mutations [3]. RNA is known to possess various structural properties, and functional RNA structures are generally created by the three-dimensional organization of small structural motifs, mismatch base pair, bulge, hairpin loop, and internal loop structures [4]. These diverse structures of RNA form binding sites for proteins and are important for the different cellular functions of RNA [5–7]. Guanine (G)-rich sequences in the single-strand DNA and RNA can form non-canonical four-stranded structures known as the G-quadruplexes (G4s) [8]. G4s forming sequences are mostly in functional regions, including the telomere, located at the end of the eukaryotic chromosome, the promoter regions of oncogenes (e.g., k-RAS, c-MYC, c-KIT) [9– 12]. Thus, these higher-order structures of nucleic acids have numerous functions in the regulation of gene expressions, and the selective binding molecules to the higher-order structures would become useful chemical tools for elucidating these functions. In this review, we describe the development of the middle-size molecules for alkylation to the higher-order structures of nucleic acids to stabilize these structures. We focused on the alkylation targeted to the two higher-order structures, the T-T mismatch structure and G4 structure, related to diseases.

(A) Typical Duplex DNA

Fig. 3.1 Structure of DNA and RNA

(B) Representative higher-order structures

Duplex DNA

Mismatch

Abasic (AP) site

Hairpin loop

Internal loop

G-quadruplex (G4)

3 Development of Middle-Size Molecules for Alkylation …

3.2 3.2.1

37

Alkylation to T-T Mismatch Site T-T/U-U Mismatch Structure and Recognition Molecules

The T-T/U-U mismatch can be formed in the abnormally expanded trinucleotide repeat sequences [13]. Myotonic dystrophy type 1 (DM1) is one of the trinucleotide repeat disorders caused by an abnormal expansion of the CTG trinucleotide repeats in the 3’-UTR of the DMPK gene. While the normal expansion of the CTG is 5–37 repeats, the abnormal one reaches 50–2000 repeats [14]. The double-stranded DNA fragments that contain expandable repeats promote the formation of the ‘slipped-stranded’ DNA conformation. In this case, an out-of-register realignment of the complementary repetitive strands gives rise to ‘slip-outs’ that are folded into hairpin-like structures (Fig. 3.2). These abnormal structures might induce the abnormal expansion of the repeat sequences, but the mechanisms that are responsible for the repeat expansions in DNA are not generally understood. This repeat DNA produces the toxic CUG repeat RNA which forms a complex with the alternative-splicing regulator muscleblind-like protein (MBNL), leading to splicing defects and disease symptoms [15–18]. Thus, the abnormal extended CTG or CUG repeats are anticipated to be one of the important targets for DM1 and several promising compounds have been developed [19–21]. Disney’s group developments lead to the identification strategy called Inforna that integrates advances in RNA structure determination and prediction, identification of RNA motif–small molecule interactions, and the scoring of those interactions via the statistical analysis of small molecules [22]. They designed the target r(CUG)exp using an Inforna database, which identified the Hoechst derivative as a ligand for the UU internal loops. The bis-Hoechst derivative (1) can recognize two adjacent UU loops and improved the MBNL1-dependent MBNL1 exon 5 pre-mRNA splicing defect (Fig. 3.3). (a) Replication of (CTG)n repeat sequences

CTGCTGCTGCTGCTGCTGCTGCTG GACGACGACGACGACGACGACGAC slippage C T G C T G CTGCTG GACGAC G A C Abnormal G expansion of DNA A C

CTG repeat

G T C G T C

(b) Transcription (CTG)n repeat sequences 3’ 5’

5’ 3’

CTGCTGCTGCTGCTGCTGCTGCTG GACGACGACGACGACGACGACGAC transcription 5’ CUGCUGCUGCUGCUGCUGCUGCUG

3’ 5’ 3’

MBNL1 (Splicing factor)

C A G C A G

C G U C G U C

G C U G C U G

Production of abnormal protein

Fig. 3.2 Replication and transcription of (CTG)n repeat sequences, *Reproduced from CSJ Current Review (ISBN978-4-7598-13906, p 91)

38

F. Nagatsugi and K. Onizuka

Fig. 3.3 Binding molecules to U-U mismatch structure by Disney’s group

They also developed bis-Hoechst derivatives conjugated with chlorambucil for the covalent bond formation (2), and bleomycin for the RNA cleavage (3) [23]. The bis-Hoechst conjugated bleomycin derivatives were applied to cells [24] and a myotonic dystrophy mouse model [25] that resulted in improvement of the pre-mRNA splicing defect by cleaving the target RNA. Berglund’s group reported that actinomycin D (a general transcription inhibitor) bound to the CTG repeats in DNA decreased the CUG transcript levels in a dose-dependent manner in the DM1 cell and mouse models at significantly lower concentrations (nanomolar) [26]. They found that the pentamidine derivative (4) bound to the CTG repeats reduced the abnormal extended CUG RNA levels [27] (Fig. 3.4). Furamidine (5) showed the efficient improvement of the pre-mRNA splicing defect by reducing the toxicity compared to pentamidine (4) [28]. Recently, they studied the mechanism of action of furamidine (5), which affected multiple pathways in the DM1 mechanism to rescue the mis-splicing [29]. The Nakatani group designed 2,9-dialkylamino substituted phenanthroline (DAP: 6) to bind to the unpaired uracil with two hydrogen bonds (Fig. 3.5). They demonstrated that DAP (6) was bound to r(CUG) by an Isothermal Titration Calorimetry (ITC) measurement and inhibited the translation during the in vitro dual luciferase assay by binding to the r(CUG) repeats [30]. The dimeric form of DAP (7) exhibited a binding affinity to r(CUG) repeats and was partially effective in the recovery of the pre-mRNA splicing defects in the DM1 cell model and DM1 mouse model [31]. The mechanism of the recovery was attributed to the interference with the binding of MBNL1 to the r(CUG) repeats by 7 as indicated by the in vitro competitive binding assay. Zimmerman’s group designed the triaminotriazine-aminoacridine conjugates (8) as a T-T/U-U mismatch binding molecule which stabilizes the complex by the intercalation of the acridine and hydrogen bonds (Fig. 3.6). The binding molecule (8a) binds to the CTG or CUG repeats and strongly inhibited the formation of the complex with MBNL and the r(CUG) repeat [32]. The bis-amidinium ligand (9a) with groove-binding parts carrying two triaminotriazine units were designed and interacted with r(CUG)12 to disrupt the MBNL1-r(CUG)12 interaction in vitro. In

3 Development of Middle-Size Molecules for Alkylation …

39

Fig. 3.4 The structures by Berglund group studies on the DM1 models

Pentamidine: 4

Furamidine: 5

6 N N H2 N

N H O

N H N

H O

H2 N

N

N

N

H

H

N NH O

NH2

2

HN

N RNA

N H2 N

N

N

H

H

N

O

2 NH

7

Fig. 3.5 The structures by Nakatani group studies on the DM1 models

addition, this ligand partially improved the pre-mRNA splicing defect in the DM1 cell model and achieved reversal of the phenotypic defects observed in the DM1 infected Drosophila [33]. They rationally designed the ligands 8b and 9b as multi-target agents for the DM1. These agents operated in the following three distinct ways: (1) bind CTGexp and inhibit formation of the CUGexp transcript, (2) bind CUGexp and inhibit sequestration of MBNL1, and (3) cleave CUGexp in an RNase-like manner [34]. The ligand (9b) improved the neurodegenerative phenotype in the DM1 Drosophila model.

8a: R=

9a: R=H 9b: R=(CH2CH2NH2)2

8b: R= 8 10 (a=1~3, d=1~3) 9

Fig. 3.6 The structures by Zimmerman group studies on the DM1 models

40

F. Nagatsugi and K. Onizuka

The oligomer of 9a exhibited an excellent multitargeting activity binding to both d(CTG)exp and r(CUG)exp, and showed important activities in the DM1 cells and in a DM1 liver mouse model with a high cell permeability [35]. The acridine-triaminotriazine macrocycles (10) bound to the CTG trinucleotide repeats in DNA with a minimal nonspecific binding and showed an almost 10-fold lower cytotoxicity in HeLa cells and up to a fourfold higher transcription inhibition of d(CTG
CAG)74 compared to a noncyclic analog [36]. Thus, many binding molecules to the T-T/U-U mismatch site are developed for DM1 treatments but their mechanism of action is very complicated. New chemical tools for selective alkylation to the target site are necessary for studying the action mechanism.

3.2.2

Alkylation to T-T Mismatched Structure [37]

We designed the vinyldiaminotriazine (VDAT)-acridine conjugate (11) for the selective alkylation of the T-T mismatched DNA or U-U mismatched RNA. Based on the paper reported by Zimmerman [32], we expected that the VDAT-acridine conjugate (11) would recognize the T-T or U-U by the stable complex formation with hydrogen bonds as shown in Fig. 3.7 and selectively react with the T or U base by the proximity effect. The VDAT-acridine conjugate (11) was synthesized and the alkylation to the T-T mismatched DNA or U-U mismatched RNA evaluated. The alkylation to the T-T mismatch DNA efficiently proceeded with 11, but the reaction efficiency to the U-U mismatched RNA was much lower than that to the T-T mismatched DNA as shown in Fig. 3.8. The higher reactivity to the T-T mismatched DNA with 11 was partly due to more tightly binding to the T-T mismatched DNA than the U-U mismatched RNA. The purified alkylated oligonucleotides were enzymatically hydrolyzed for determination of the alkylated nucleoside structure. The NMR studies using the purified alkylated nucleoside showed that the alkylation selectively proceeded at the N3 position of thymine (Fig. 3.9). The thermal stability of the alkylated duplex DNA containing the alkylated thymine (dT*) and 4 bases (A, G, T, C) at the complementary position was investigated by measuring the melting temperature (Tm). Interestingly, all the Tm values were comparable to the native full match duplex, suggesting that the alkylated oligonucleotide would stabilize the duplex by the intercalation of the acridine part and flipping out the complementary base (Fig. 3.10). To confirm the base flipping, the fluorescence emission assay using 2-amino purine (2-AP) as a fluorescent probe was performed. Since the AP fluorescence is highly quenched in a duplex due to the stacking interactions, we can confirm whether the 2-AP base is flipped out of the DNA helix by checking the fluorescence increase [38, 39]. The fluorescence of the duplex containing the alkylated thymine (dT*) and 2-AP at the complementary site was significantly high, indicating that alkylated thymine can induce the flipping of the 2-AP base. Based on

3 Development of Middle-Size Molecules for Alkylation …

41

=R vinyldiaminotriazine (VDAT)-acridine (11)

T-T mismatch in DNA

U-U mismatch in RNA

Fig. 3.7 Design of the reactive molecule to T-T/U-U mismatch structure

Reaction to DNA duplex target 5í 3í

C T G G C X G C G C 3í 5í G A C C G Y C G C G X = A, G, C, T; Y = A, G, C, T

Reaction to RNA duplex target 3í

C T G G C U G C G C 3í 5í G A C C G Y C G C G Y = A, U

T-T

C-C A-A, G-G Full match

U-U ssU Full match

The reaction was carried out with the duplex DNA (5 μM) and 7 (100 μM) in MES buffer (50 mM, pH 7.0) containing NaCl (100 mM) and 2% DMSO at 37 o C. Fig. 3.8 Time course for alkylation to mismatch structure with VDAT-acridine

42

F. Nagatsugi and K. Onizuka

Fig. 3.9 Chemical structure of adduct

3

dT* this alkylated structure, we also developed new artificial bases with the flipping-out property [40]. Next, we studied the alkylation to the (CTG)12 oligonucleotide as the model for the alkylation to the CTG repeats DNA and the reaction rate to (CTG)12 was 6 times higher than that of (CTG)2. These results suggested that the efficiency of one alkylation would increase in proportion to the number of the T-T mismatches, and the alkylation efficiency to the abnormal expansion repeats target can be expected to increase in proportion to the number of CTGs. In addition, the alkylation of the CTG repeat DNA led to strongly inhibiting the primer extension reaction and transcription. Hence, the VDAT-acridine conjugate (11) would be a new biochemical tool for a CTG repeats study and may provide a new strategy for the molecular therapy of DM1.

3.3 3.3.1

Alkylation to G4s Structure G4 Structure and Alkylation Molecules to G4

Guanine (G)-rich sequences can form non-canonical four-stranded structures known as G4s stabilized by Hoogsteen hydrogen bonding of the G-tracts in the presence of monovalent cations, i.e., potassium or sodium [41] (Fig. 3.11). There is a variety of G4s topologies depending on the sequences and folding conditions. For example, in the human telomeric sequence (TTAGGG), the intramolecular G4s show three kinds of topologies, i.e., hybrid in potassium (K+), anti-parallel in sodium (Na+) and parallel in potassium (K+) under the crowding conditions [42–44] (Fig. 3.12). The G4 structure has been found in biologically significant regions of the genomic DNA, including the telomere and promoter, and is known to play an important role in a number of biological processes. Thus, the G4s have been significant therapeutic targets in oncology [45, 46], and a variety of small molecules to stabilize these secondary structures have been developed [47, 48]. The covalent modification of G4 is one of the strategies for augmenting the binding affinity. Several chemically reactive moieties conjugated with the G4 ligand were developed as G4 alkylating agents. The nitrogen mustard Chlorambucil

3 Development of Middle-Size Molecules for Alkylation …

5´ -GCGC 3´ -CGCG 60

43

dT*GCCAG-3´

X

CGGTC-5´

Non-modified T

X = A, G, C, T

dT*

55

Tm / ºC

50 adduct

45

Flipped-out base

40 35 30

dA dG dC dT The Tm values were measured using duplex (2.0 μM) in MES buffer (20 mM, pH 7.0) containing NaCl (100 mM).

Fig. 3.10 Tm values of dT*-containing DNA–DNA and the base flipping structure of the alkylated duplex DNA, *Reproduced from CSJ Current Review (ISBN978-4-7598-13906, p 91)

Fig. 3.11 Chemical structure of a G-tetrad

(Chl) was conjugated with pyridostatin (PDS) as a selective G4 ligand and PDS-Chl (12) alkylated the G4 structure without external stimulation [49]. The naphthalene diimides (NDIs) are very potent G4 binding ligands, which conjugated with a variety of reactive entities and the NDI-oxirane derivative (13) alkylated G4 structure [50]. These ligands have an intrinsic reactivity and cause considerable instability under physiological conditions (Fig. 3.13). The quinone-methides (QMs) are reactive intermediates and the stable precursors of the QMs are conjugated with G4 binding molecules as shown Fig. 3.14. These derivatives produced a reactive QM by heating or photo-irradiation and alkylated to the G4s [51, 52].

44

F. Nagatsugi and K. Onizuka

Human telomeric sequence: d[TA(GGGTTA)3GGG]

K+ (3+1) hybrid

Na+ antiparallel

K++PEG200 parallel

Fig. 3.12 Three representative G4 topologies in the human telomeric sequence

Fig. 3.13 G4 alkylating molecules without external stimulation

12

13

Recently, the Freccero group developed a photo-reactive NDI derivative (14) activated by green light. The NDI derivative (14) generated the reactive phenoxide radical by an intramolecular electron transfer process and alkylated the G4 structure [53]. The photo induced-alkylating groups (benzophenone (15) and an aryl azide

Fig. 3.14 Quinone methide precursors conjugated NDI for G4 alkylating molecules

Δ or h

X=OH, N+(Me)3

DNA

3 Development of Middle-Size Molecules for Alkylation …

14

15

45

16

Fig. 3.15 G4 alkylating molecules activated by photo-irradiation

(16)) tethered to PDS reacted with the G4 structure by photo-irradiation [54] (Fig. 3.15). Thus, several G4 ligands containing chemically reactive moieties have been developed for the G4 alkylation. However, in many cases, the efficiency and selectivity of the alkylation for the G4 structures require further optimization to allow such molecules to serve as biological tools or therapeutic agents.

3.3.2

Alkylation to G4 Structure by Activated Proximity Effects

We have reported that the 2-amino-6-vinylpurine (AVP) derivatives conjugated with DNA binding molecules induce the selective alkylation to thymine at the complementary site of an abasic site in the duplex DNA [55]. A variety of acridine derivatives are reported as G4 ligands with a high affinity, and AVP was conjugated with acridine as a G4 binding part. We evaluated the reactivity to G4 (G4-23: d [TAG3TT AG3TTAG3TT]) with two kinds of acridine-AVP derivatives (17 and 18). The alkylation yields of aminoacridine-AVP (18) appeared higher than that of the amidoacridine-AVP derivative (17) (Fig. 3.16). The lower alkylation yields with the amidoacridine-AVP derivative (17) would be attributed to the lower affinity to the DNA because of the lack of an electrostatic interaction by the low pKa (lower than 4.0) for the amidoacridine derivatives (17). These results suggested that the reaction of the AVP derivatives would be accelerated by the proximity to the target G4 structure. The lower yields with 18 were observed in the reactions to the single-strand and duplex DNAs, indicating that the aminoacridine-AVP (18) exhibited a comparably higher selectivity to the G-4 structure than the single strand DNA or double helix DNA. The alkylation yields to G4-23 with 18 were significantly higher in the K+ buffer than in the Na+ buffer, indicating that the alkylation with 18 occurred with a higher efficiency to the [3 + 1] hybrid structure than to the antiparallel one (Fig. 3.17). The analysis of the 3’ exonuclease digestion with purified G4 adducts suggested that the alkylation site might be T7, T12, and T13 in the loop structure of G4. The alkylation of G4 DNA

46

F. Nagatsugi and K. Onizuka

N H N

O

X

O

O

N

N

H N

N

N

T

amidoacridine-AVP derivative (17) N

X=

O

O

HN

O

R=

O

N H

Acridine derivatives

N

N O

NH2

O

H

N

N N

O

O

N

H N

N N

NH2

O

N H

aminoacridine-AVP derivative (18)

Fig. 3.16 Design of the alkylating probes for the G4 DNA

with 18 effectively caused the inhibition of the primer extension and replication in the in vitro experiments (Fig. 3.18). The alkylation with aminoacridine-AVP (18) was highly selective to thymine [56], but the reaction rate was slow. The increasing of the reaction rate might promote the instability of the alkylating agent under physiological conditions. To overcome this dilemma, it is necessary to develop the OFF-ON type of alkylating agents. Namely, the reactivity of the alkylating agents can be in the OFF state until approaching the target site and its alkylating reactivity is in the ON state

Fig. 3.17 Time course for alkylation yields with 18, *Reproduced from ref. [56] with permission from the Royal Society of Chemistry

3 Development of Middle-Size Molecules for Alkylation …

47

Fig. 3.18 Schematic structures of G4 in K+ buffers and major alkylation sites, *Reproduced from ref. [56] with permission from the Royal Society of Chemistry

triggered by binding to the target site to promote the alkylation. The reactive OFF-ON type alkylating agents to the G4 structure activated by external stimuli, such as photo-irradiation, were described in the preceding section. However, the ideal alkylating agent is activated just by binding with the target nucleic acid with no requirement of external stimulation. We designed the OFF-ON type alkylating agent based on our previous vinyl chemistry [57, 58]. The vinyl group of the quinazoline derivative (VQ: 20) was expected to be highly reactive due to its electron-withdrawing carbonyl group conjugated to the vinyl group. The highly-reactive vinyl group in VQ was protected with several sulfide groups as shown in Fig. 3.19 and expected to regenerate by means of an E1cB-type elimination only when the ligand reaches and interacts with the target nucleic acids. The generated reactive VQ moiety will then promote the efficient alkylation with the target base. The synthesized VQ precursors (19) were evaluated for their alkylation reactivity to the telomere G4 (G4-23), double-strand DNA (ds), and single-strand DNA (ss). The comparison of the reactivity to G4 (G4-23) for the VQ precursors (19) showed that the order of the reactivity was S(O)Ph > S(O) Me > SPh > SMe. The alkylation yields with the sulfoxide derivatives (19b) and (19d) reached 70% and 90% after 8 h, respectively. The sulfide derivatives (19a)

Target nucleic acids

X: SMe

19 OFF Stable

19a: X=SMe b: X=S(O)Me c: X=SPh d: X=S(O)Ph

Fig. 3.19 Design of the OFF-ON type alkylating agents

20 ON reactive

R:

S(O)Me

SPh

S(O)Ph

48

F. Nagatsugi and K. Onizuka

and (19c) produced the alkylation adducts with 20% and 60% yields after 16 h respectively. The reaction rates of 19b–d to G4 were considerably higher than that of the AVP derivatives (18). All the precursors (19) did not show any significant reactivity to the double-strand (ds) and single-strand DNA (ss), and exhibited a high selectivity to G4 over ds and ss. The VQ-S(O)Me derivative (19b) was treated with excess glutathione (GSH) to produce the GSH adduct and the half-life t1/2 of the ligand (19b) was 7.2 h. On the other hand, the VQ-SMe derivative (19a) did not form an adduct with GSH nor decomposed even after a long 120 h of incubation but gradually alkylated to G4. These results suggested that the VQ-S(O)Me precursor (19b) undergoes a spontaneous vinyl generation and the VQ-SMe precursor (SMe: 19a) generates the vinyl site-specifically or only when being triggered by the presence of the target DNA site (Fig. 3.20). The evaluation of the alkylation to G4 using aromatic substituted thiopenol precursors (SPh: 19c) indicated that vinyl generation of the VQ-thiophenol precursors took place in an equilibrium state. The thiophenyl (SPh) and sulfide (SMe) precursors are more stable in water than chlorambucil, which undergoes hydrolysis after 1 h [47]. Thus, the VQ precursors are some of the candidates for the OFF-ON alkylation to the G4 structure, which are activated only by binding to the target higher-order structures of nucleic acids without any external stimuli. With the aim to determine the adduct structure, the purified alkylated G4 with 19c (SPh) was subjected to enzymatic hydrolysis. The monomer adduct was purified and the mass spectrum determined as the thymidine (dT) adduct with one molecule of the aminoacridine-VQ ligand. The precise analysis of the adduct structure indicated that the position of the alkylation of VQ was the N3 position of thymine (21) (Fig. 3.21). For an explanation of the selective alkylation mechanism, we attempted the comparison of the alkylation to five kinds of bases with an opposite AP site using the VQ-SPh precursor (19c). The alkylation proceeded to thymidine (T) with 90% yields and deoxy inosine (dI) with 60% yield after 24 h,

Fig. 3.20 Plausible alkylation mechanism of VQ-precursors

3 Development of Middle-Size Molecules for Alkylation …

49

O

Fig. 3.21 Adduct structure with VQ alkylation in G4

N

O

N

N N

HO

R

O

O OH

but the others were in low yields. Based on these results and the VQ alkylation of the thymidine at the N3 position, the 4-pyrimidinone structure is essential to efficiently alkylate the target by the VQ. We proposed a concerted mechanism in which the VQ reacts with the enol form of the thymidine base, indicating a similar reactivity to inosine at the complementary site of the AP site. The proton transfer from the target base to the VQ during base pairing may be a key factor for efficient alkylation. The guanosine (dG) would not react due to the steric hindrance of the amino group or less reactivity (Fig. 3.22). In our experiments, 19c reacted with the U base at an AP-site in RNA, G4 RNA, and T-T mismatch structures in good yields [59]. Since the VQ-precursors have an effective reactivity, their conjugates with the selective DNA or RNA binder to a higher-order structure are expected to induce the efficient alkylation for various targets.

3.4

Summary and Perspective

The higher-order structures of nucleic acids play critical roles in the gene expression regulation and it is anticipated that selective binding molecules to higher-order structures could be promising tools for elucidating these functions. In this paper, we focused on each binding molecule to the higher-ordered structure, T-T/U-U

DNA strand containing AP

Target DNA strand

O

O N

N

X

N

R

N

S

X=

O

N=dA, dT, dG, dC, dI

N

X

T or I

N O

R

DNA

Thymidine (T)

N O

O

N

19c

O O

H

H

N

O N

N N

O N N R

DNA

N R

N

H O

Inosine (dI)

Fig. 3.22 Alkylation with 19c to AP site and plausible alkylation mechanism of VQ

N

N

N

DNA

H2N

Guanosine (dG)

50

F. Nagatsugi and K. Onizuka

mismatch, and G4. We have developed middle-size molecules, which consist of a binding group to the target nucleic acids and the alkylating moiety, for alkylation to the higher-order structures of nucleic acids. These molecules exhibited a selective alkylation to thymine in the target higher order structure by a proximity effect. Unfortunately, we could not create a probe for alkylation to the U-U mismatch RNA. Many kinds of higher-order structures are contained in the functional RNA and play a crucial role in the interaction with the RNA binding protein. Thus, the development of binding molecules to the RNA higher-order structure is desired, but the diverse RNA structure makes it difficult to develop such binding molecules. The novel strategies of exploration for the binding molecules to the RNA higher-order structure are essential to developing an effective alkylation probe to the RNA higher-order structure.

References 1. McNeill DR, Whitaker AM, Stark WJ, Illuzzi JL, McKinnon PJ, Freudenthal BD, Wilson DM (2020) Functions of the major abasic endonuclease (APE1) in cell viability and genotoxin resistance. Mutagenesis 35:27–38 2. Yang ZY, Nejad MI, Varela JG, Price NE, Wang YS, Gates KS (2017) A role for the base excision repair enzyme NEIL3 in replication-dependent repair of interstrand DNA cross-links derived from psoralen and abasic sites. DNA Repair 52:1–11 3. Kingsland A, Maibaum L (2018) DNA base pair mismatches induce structural changes and alter the free-energy landscape of base flip. J Phys Chem B 122:12251–12259 4. Halder S, Bhattacharyya D (2013) RNA structure and dynamics: a base pairing perspective. Prog Biophy Mol Bio 113:264–283 5. Nam Y, Chen C, Gregory RI, Chou JJ, Sliz P (2011) Molecular basis for interaction of let-7 MicroRNAs with Lin28. Cell 147:1080–1091 6. Mayr F, Heinemann U (2013) Mechanisms of Lin28-mediated miRNA and mRNA regulation —a structural and functional perspective. Int J Mol Sci 14:16532–16553 7. Sugimoto Y, Vigilante A, Darbo E, Zirra A, Militti C, D’Ambrogio A, Luscombe NM, Ule J (2015) hiCLIP reveals the in vivo atlas of mRNA secondary structures recognized by Staufen 1. Nature 519:491–494 8. Varshney D, Spiegel J, Zyner K, Tannahill D, Balasubramanian S (2020) The regulation and functions of DNA and RNA G-quadruplexes. Nature Rev Mol Cell Biol 21:459–474 9. Hsu STD, Varnai P, Bugaut A, Reszka AP, Neidle S, Balasubramanian S (2009) A G-Rich sequence within the c-kit oncogene promoter forms a parallel G-Quadruplex having asymmetric G-Tetrad dynamics. J Am Chem Soc 131:13399–13409 10. Phan AT, Kuryavyi V, Burge S, Neidle S, Patel DJ (2007) Structure of an unprecedented G-quadruplex scaffold in the human c-kit promoter. J Am Chem Soc 129:4386–4392 11. Cogoi S, Xodo LE (2006) G-quadruplex formation within the promoter of the KRAS proto-oncogene and its effect on transcription. Nucleic Acids Res 34:2536–2549 12. Siddiqui-Jain A, Grand CL, Bearss DJ, Hurley LH (2002) Direct evidence for a G-quadruplex in a promoter region and its targeting with a small molecule to repress c-MYC transcription. Proc Natl Acad Sci USA 99:11593–11598 13. Mirkin SM (2007) Expandable DNA repeats and human disease. Nature 447:932–940 14. Pettersson OJ, Aagaard L, Jensen TG, Damgaard CK (2015) Molecular mechanisms in DM1 —a focus on foci. Nucleic Acids Res 43:2433–2441

3 Development of Middle-Size Molecules for Alkylation …

51

15. Miller JW, Urbinati CR, Teng-Umnuay P, Stenberg MG, Byrne BJ, Thornton CA, Swanson MS (2000) Recruitment of human muscleblind proteins to (CUG)n expansions associated with myotonic dystrophy. EMBO J 19:4439–4448 16. Fardaei M, Rogers MT, Thorpe HM, Larkin K, Hamshere MG, Harper PS, Brook JD (2002) Three proteins, MBNL, MBLL and MBXL, co-localize in vivo with nuclear foci of expanded-repeat transcripts in DM1 and DM2 cells. Hum Mol Genet 11:805–814 17. Konieczny P, Stepniak-Konieczna E, Sobczak K (2014) MBNL proteins and their target RNAs, interaction and splicing regulation. Nucleic Acids Res 42:10873–10887 18. Timchenko L (2020) Correction of RNA-Binding Protein CUGBP1 and GSK3 beta signaling as therapeutic approach for congenital and adult myotonic dystrophy type 1. Int J Mol Sci 21:94–112 19. Thornton CA, Wang E, Carrell EM (2017) Myotonic dystrophy: approach to therapy. Curr Opin Genet Dev 44:135–140 20. Konieczny P, Selma-Soriano E, Rapisarda AS, Fernandez-Costa JM, Perez-Alonso M, Artero R (2017) Myotonic dystrophy: candidate small molecule therapeutics. Drug Discov Today 22:1740–1748 21. Lopez-Morato M, Brook JD, Wojciechowska M (2018) Small molecules which improve pathogenesis of myotonic dystrophy type 1. Front Neurol 9:1664–2295 22. Velagapudi SP, Gallo SM, Disney MD (2014) Sequence-based design of bioactive small molecules that target precursor microRNAs. Nature Chem Biol 10:291–297 23. Rzuczek SG, Colgan LA, Nakai Y, Cameron MD, Furling D, Yasuda R, Disney MD (2017) Precise small-molecule recognition of a toxic CUG RNA repeat expansion. Nature Chem Biol 13:188–193 24. Angelbello AJ, Defeo ME, Glinkerman CM, Boger DL, Disney MD (2020) Precise targeted cleavage of a r(CUG) repeat expansion in cells by using a small-molecule-deglycobleomycin conjugate. ACS Chem Biol 15:849–855 25. Angelbello AJ, Rzuczek SG, McKee KK, Chen JL, Olafson H, Cameron MD, Moss WN, Wang ET, Disney MD (2019) Precise small-molecule cleavage of an r(CUG) repeat expansion in a myotonic dystrophy mouse model. Proc Natl Acad Sci USA 116:7799–7804 26. Siboni RB, Nakamori M, Wagner SD, Struck AJ, Coonrod LA, Harriott SA, Cass DM, Tanner MK, Berglund JA (2015) Actinomycin D specifically reduces expanded CUG repeat RNA in myotonic dystrophy models. Cell Reports 13:2386–2394 27. Coonrod LA, Nakamori M, Wang W, Carrell S, Hilton CL, Bodner MJ, Sibori RB, Docter AG, Haley MM, Thornton CA, Berglund JA (2013) Reducing levels of toxic RNA with small molecules. ACS Chem Biol 8:2528–2537 28. Siboni RB, Bodner MJ, Khalifa MM, Docter AG, Choi JY, Nakamori M, Haley MM, Berglund JA (2015) Biological efficacy and toxicity of diamidines in myotonic dystrophy type 1 models. J Med Chem 58:5770–5780 29. Jenquin JR, Coonrod LA, Silverglate QA, Pelletier NA, Hale MA, Xia GB, Nakamori M, Berglund JA (2018) Furamidine rescues myotonic dystrophy type I associated mis-splicing through multiple mechanisms. ACS Chem Biol 13:2708–2718 30. Li J, Matsumoto J, Bai LP, Murata A, Dohno C, Nakatani K (2016) A ligand that targets CUG trinucleotide repeats. Chem Eur J 22:14881–14889 31. Li JX, Nakamori M, Matsumoto J, Murata A, Dohno C, Kiliszek A, Taylor K, Sobczak K, Nakatani K (2018) A dimeric 2,9-diamino-1,10-phenanthroline derivative improves alternative splicing in myotonic dystrophy type 1 cell and mouse models. Chem Eur J 24:18115– 18122 32. Arambula JF, Ramisetty SR, Baranger AM, Zimmerman SC (2009) A simple ligand that selectively targets CUG trinucleotide repeats and inhibits MBNL protein binding. Proc Natl Acad Sci USA 106:16068–16073 33. Wong CH, Nguyen L, Peh J, Luu LM, Sanchez JS, Richardson SL, Tuccinardi T, Tsoi H, Chan WY, Chan HYE, Baranger AM, Hergenrother PJ, Zimmerman SC (2014) Targeting toxic RNAs that cause myotonic dystrophy type 1 (DM1) with a bisamidinium inhibitor. J Am Chem Soc 136:6355–6361

52

F. Nagatsugi and K. Onizuka

34. Nguyen L, Luu LM, Peng SH, Serrano JF, Chan HYE, Zimmerman SC (2015) Rationally designed small molecules that target both the DNA and RNA causing myotonic dystrophy type 1. J Am Chem Soc 137:14180–14189 35. Lee J, Bai YG, Chembazhi UV, Peng SH, Yum K, Luu LM, Hagler LD, Serrano JF, Chan HYE, Kalsotra A, Zimmerman SC (2019) Intrinsically cell-penetrating multivalent and multitargeting ligands for myotonic dystrophy type 1. Proc Natl Acad Sci USA 116:8709– 8714 36. Serrano JF, Lee J, Curet LD, Hagler LD, Bonson SE, Schuster EJ, Zimmerman SC (2019) Development of novel macrocyclic small molecules that target CTG trinucleotide repeats. Bioorg Med Chem 27:2978–2984 37. Onizuka K, Usami A, Yamaoki Y, Kobayashi T, Hazemi ME, Chikuni T, Sato N, Sasaki K, Katahira M, Nagatsugi F (2018) Selective alkylation of T-T mismatched DNA using vinyldiaminotriazine-acridine conjugate. Nucleic Acids Res 46:1059–1068 38. Holz B, Klimasauskas S, Serva S, Weinhold E (1998) 2-Aminopurine as a fluorescent probe for DNA base flipping by methyltransferases. Nucleic Acids Res 26:1076–1083 39. Nakano S, Uotani Y, Uenishi K, Fujii M, Sugimoto N (2005) DNA base flipping by a base pair-mimic nucleoside. Nucleic Acids Res 33:7111–7119 40. Onizuka K, Ishida K, Mano E, Nagatsugi F (2019) Alkyne-Alkyne photo-cross-linking on the flipping-out field. Org Lett 21:2833–2837 41. Burge S, Parkinson GN, Hazel P, Todd AK, Neidle S (2006) Quadruplex DNA: sequence, topology and structure. Nucleic Acids Res 34:5402–5415 42. Dai J, Carver M, Yang D (2008) Polymorphism of human telomeric quadruplex structures. Biochimie 90:1172–1183 43. You H, Zeng X, Xu Y, Lim CJ, Efremov AK, Phan AT, Yan J (2014) Dynamics and stability of polymorphic human telomeric G-quadruplex under tension. Nucleic Acids Res 42:8789– 8795 44. Dailey MM, Miller MC, Bates PJ, Lane AN, Trent JO (2010) Resolution and characterization of the structural polymorphism of a single quadruplex-forming sequence. Nucleic Acids Res 38:4877–4888 45. Asamitsu S, Obata S, Yu ZT, Bando T, Sugiyama H (2019) Recent progress of targeted G-quadruplex-preferred ligands toward cancer therapy. Molecules 24:429–458 46. Hänsel-Hertsch R, Antonio MD, Balasubramanian S (2017) DNA G-quadruplexes in the human genome: detection, functions and therapeutic potential. Nature Rev Mol Cell Biol 18:279–284 47. Asamitsu S, Bando T, Sugiyama H (2019) Ligand design to acquire specificity to intended G-quadruplex structures. Chem Eur J 25:417–430 48. Sun ZY, Wang XN, Cheng SQ, Su XX, Ou TM (2019) Developing novel G-quadruplex ligands: from interaction with nucleic acids to interfering with nucleic acid-protein interaction. Molecules 24:396–425 49. Antonio MD, McLuckie KIE, Balasubramanian S (2014) Reprogramming the mechanism of action of chlorambucil by coupling to a G-quadruplex ligand. J Am Chem Soc 136:5860– 5863 50. Doria F, Nadai M, Folini M, Scalabrin M, Germani L, Sattin G, Mella M, Palumbo M, Zaffaroni N, Fabris D, Freccero M, Richter SN (2013) Targeting loop adenines in G-quadruplex by a selective oxirane. Chem Eur J 19:78–81 51. Nadai M, Doria F, Antonio MD, Sattin G, Germani L, Percivalle C, Palumbo M, Richter SN, Freccero M (2011) Naphthalene diimide scaffolds with dual reversible and covalent interaction properties towards G-quadruplex. Biochimie 93:1328–1340 52. Antonio MD, Doria F, Richter SN, Bertipaglia C, Mella M, Sissi C, Palumbo M, Freccero M (2009) Quinone methides tethered to naphthalene diimides as selective G-quadruplex alkylating agents. J Am Chem Soc 131:13132–13141 53. Nadai M, Doria F, Germani L, Richter SN, Freccero M (2015) A photoreactive G-quadruplex ligand triggered by green light. Chem Eur J 21:2330–2334

3 Development of Middle-Size Molecules for Alkylation …

53

54. Verga D, Hamon F, Poyer F, Bombard S, Teulade-Fichou MP (2014) Photo-cross-linking probes for trapping G-quadruplex DNA. Angew Chem Int Ed 53:994–998 55. Sato N, Tsuji G, Sasaki Y, Usami A, Moki T, Onizuka K, Yamada K, Nagatsugi F (2015) A new strategy for site-specific alkylation of DNA using oligonucleotides containing an abasic site and alkylating probes. Chem Commun 51:14885–14888 56. Sato N, Takahashi S, Tateishi-Karimata H, Hazemi ME, Chikuni T, Onizuka K, Sugimoto N, Nagatsugi F (2018) Alkylating probes for the G-quadruplex structure and evaluation of the properties of the alkylated G-quadruplex DNA. Org Biomol Chem 16:1436–1441 57. Nagatsugi F, Kawasaki T, Usui D, Maeda M, Sasaki S (1999) Highly efficient and selective cross-linking to cytidine based on a new strategy for auto-activation within a duplex. J Am Chem Soc 121:6753–6754 58. Kawasaki T, Nagatsugi F, Ali MM, Maeda M, Sugiyama K, Hori K, Sasaki S (2005) Hybridization-promoted and cytidine-selective activation for cross-linking with the use of 2-amino-6-vinylpurine derivatives. J Org Chem 70:14–23 59. Onizuka K, Hazemi ME, Sato N, Tsuji G, Ishikawa S, Ozawa M, Tanno K, Yamada K, Nagatsugi F (2019) Reactive OFF-ON type alkylating agents for higher-ordered structures of nucleic acids. Nucleic Acids Res 47:6578–6589

Chapter 4

In Situ Synthesis of Glycoconjugates on the Cell Surface: Selective Cell Imaging Using Low-Affinity Glycan Ligands Shogo Nomura, Misako Taichi, and Katsunori Tanaka

Abstract In the field of molecular imaging, selectivity for target cells is a key determinant of the clarity and accuracy of imaging contrast. Recently, a novel pre-targeted imaging method has been developed whereby target cells can be selectively imaged using a labeled N-glycan that has been ligated with an integrin-targeted cyclic arginylglycylaspartic acid (RGD) peptide on the cell surface in situ. The conceptual basis of this method, its development, and its application to cancerous and noncancerous cells is described in this chapter. This method has the powerful advantage of having the ability to discriminate between various cancerous and non-cancerous cells that cannot be distinguished using conventional RGD ligands. Using this method, various N-glycan molecules, even those with millimolar affinities for their cognate lectins, could be used for selective cancer cell differentiation. Keywords Cell targeting imaging


In vivo synthesis
Glycoconjugate
Molecular

S. Nomura Department of Chemical Science and Engineering, School of Materials and Chemical Technology, Tokyo Institute of Technology, 2-12-1 O-Okayama, Meguro-Ku, Tokyo 152-8552, Japan M. Taichi Biofunctional Synthetic Chemistry Laboratory, RIKEN Cluster for Pioneering Research, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan K. Tanaka (&) GlycoTargeting Research Laboratory, RIKEN Baton Zone Program, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan e-mail: [email protected]; [email protected] K. Tanaka Biofunctional Chemistry Laboratory, a. Butlerov Institute of Chemistry, Kazan Federal University, 18 Kremlyovskaya Street, Kazan 420008, Russian Federation © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_4

55

56

4.1

S. Nomura et al.

Introduction

Molecular imaging research has conventionally used the noninvasive analysis of molecular kinetics in small animals when developing diagnostic applications [1–7]. This method allows for in vivo information about biologically active small molecules and biomolecules to be easily imaged and detected using fluorescence or radionuclides. Specific organs or tumors can be targeted using many promising tracers, which can identify the localization or expression level of target receptors [8–13]. Improvement is still needed in the selectivity and specificity of these tracers toward their intended target cells to achieve maximal diagnostic accuracy. For example, arginylglycylaspartic acid (RGD) peptides are high-affinity ligands of aVb3 integrins [9, 14], which are cell adhesion molecules that are highly expressed in the tumor vasculature [11]. RGD peptides are therefore used to image a wide variety of cancers including breast [15], brain [16], and lung [17] in small animal models. However, the use of RGD peptides often results in poor imaging contrast as the aVb3 integrins are also expressed on other endothelial cells. This interferes with the ability to clearly differentiate tumor cells from surrounding cells. Because RGD peptides bind with a high-affinity but low selectivity to these various integrins, the tumor to background signal ratio is decreased as the peptides are rapidly captured by the cells via receptor-mediated endocytosis. Tumor imaging research has generally focused on optimizing both the affinity and selectivity of ligands to cell surface targets. Recently, a new approach to imaging has been developed which utilizes both high- and low-affinity ligands targeted to independent receptors on the target cell surface. This approach was inspired by the pre-targeted method utilizing biorthogonal and click chemistry, and is used frequently in the molecular imaging field [18–23]. Figure 4.1 illustrates this approach and compares it to conventional imaging methods. This chapter describes the development of this novel approach and its power over conventional methods in discriminating between several types of cancerous and noncancerous cells.

4.2

Development of the Pre-targeted Novel Imaging Approach

The development of this new imaging approach, involving both high- and low-affinity ligands, utilized a simplified model in which various surface receptors are expressed on Cells A and B. Using conventional imaging methods, the application of the fluorescently labeled high-affinity ligand allows for the selective visualization of Cell A, as the ligand has high affinity toward the receptor overexpressed on the surface of A, i.e., at a KD in the nanomolar range (Fig. 4.1A). Cell B can also express the same receptor since RGD/integrin interactions occur on the surfaces of both tumor and common endothelial cells. Because Cells A and B can be visualized simultaneously using the same ligand, the degree of image

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

57

Fig. 4.1 Schematic diagram comparing conventional imaging methods to the novel cell imaging method involving ligands with high and low affinity to cell surface receptors. A Conventional labeling method using a high-affinity ligand (red circle). B Labeling method using a low-affinity ligand (blue triangle). C Novel labeling method using both high- and low-affinity ligands and a bioorthogonal reaction on the cell surface

contrast and its diagnostic ability is reduced, even though a difference exists in the quantities of high-affinity ligands attached to these cells. This lack of differentiation between cell types is a common issue in the application of ligands that strongly interact with a target cell. One alternative to this problem is to use a low-affinity ligand that interacts specifically with target Cell A; however, the low-affinity ligand-receptor binding is very weak (i.e., a KD in the millimolar range), and therefore is not able to bind sufficiently to enable visualization of Cell A (Fig. 4.1B). In order to address these issues, an approach involving a pre-targeted combination of strong and weak ligand/receptor interactions (Fig. 4.1C) formed the basis for the novel pre-targeted imaging method. This method uses a high-affinity ligand

58

S. Nomura et al.

prepared with a reactive tag rather than an imaging label, which is pre-targeted to Cells A and B. Subsequently, a low-affinity ligand, which has an imaging label and bioorthogonal functional groups that could react with the pre-targeted tag, was prepared and applied to the cells. The low-affinity ligand bound to the pre-targeted cell was selectively anchored through the biorthogonal reaction, whereas it was washed away immediately from the cell surface due to very weak binding to the receptor. This interaction was facilitated by the proximal effects between the two surface receptors via click conjugation. The combination of strong and weak ligand/ receptor interactions and this in situ click conjugation on the cell surface achieved both high selectivity and enhanced imaging contrast between Cells A and B. It is important to note that the labeled low-affinity ligand indeed has low affinity to avoid the problems encountered with high-affinity ligands, as described above. Using weak interactions of the ligand to the cell surface receptor is vital to the success of the pre-targeted approach.

4.2.1

Use of Peptide and Glycan Ligands

This synergistic imaging approach using high- and low-affinity ligands were applied to peptide and glycan ligands, which have different affinities to cell surface receptors (Fig. 4.1C). This in situ ligation concept is the key to taking advantage of both the strong and weak ligand interactions to produce clear imaging contrast of target cells. The development of this novel imaging method also involved synthesizing the desired glycoconjugates, which could properly distinguish the targeted cells directly through the bioorthogonal reaction produced by two proximal surface receptors. This achieves highly selective cell recognition under appropriate conditions both in vitro and in vivo. Note that cell labeling using pre-reacted peptide–glycan conjugates never produces sufficient imaging contrast because the interactions between the pre-linked molecules and the cell surface are not specific to tumor cells. These conjugates also suffer from similar issues to the strong peptide/receptor interactions described earlier in this section. Most previous studies of peptide- or antibody-conjugates that utilized these two interactions have failed for this reason.

4.3

Proof-of-Concept Study: Selective Imaging of HUVECs Using High- and Low-Affinity Ligands

The concept of the novel pre-targeted imaging method has been demonstrated using a model system in which human umbilical vein endothelial cells (HUVECs) were selectively imaged. HUVECs were chosen as the cells of study as they express two characteristic receptors which are of interest in tumor imaging—aVb3 integrins and

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

59

platelet endothelial cell adhesion molecule (PECAM) (Fig. 4.2). aVb3 integrins are cis-associated with PECAM on the same endothelial cell surface; therefore, imaging with a conventional method could potentially be confounded [24, 25]. Several dendrimer-type N-glycoclusters have been developed with an enhanced affinity toward lectins which are mediated by multivalency effects. It has also been shown that PECAM is a sialic acid-binding lectin [25] and exhibits a high-affinity toward the a(2,6)Sia-terminated N-glycan (Fig. 4.2). This model confirms that the affinity for the single glycan/lectin interaction is low, and that the a(2,6)Sia-terminated N-glycan is a low-affinity ligand for PECAM. The high-affinity ligand used for aVb3 integrins are the cyclic RGDyK peptide. For the in situ chemoselective and bioorthogonal reaction between the ligands, the strain-promoted azide-alkyne cycloaddition (SPAAC) reaction was used. This reaction is activated upon binding to surface receptors [26, 27]. The strained acetylene used in this method was dibenzocyclooctyne (DIBO), developed by Ning et al. [28]. DIBO was chosen as it displays a relatively high reactivity and is easily prepared from simple starting materials. To investigate the dependence of the SPAAC reaction on distance, the azide function was introduced on to the cyclic RGDyK peptide (Fig. 4.2, 1a–1d) using various short ethylene glycol linker lengths. These linkers dictate the distance between the ligand/receptor complexes that are tolerated in the context of the labeling reaction. A suitable linker was used to attach DIBO and tetramethylrhodamine (TAMRA), a fluorescent label, to a(2,6)-disialo-N-glycan (2a). The asialoglycan derivative (2b) was prepared as a control to confirm the sialic acid-dependent interaction with PECAM. This method demonstrated the in situ achievement of simultaneously targeting two cell surface receptors through direct linkage of their high- and low-affinity ligands, facilitated by the functionalized RGD peptide and N-glycan ligands (Fig. 4.3).

4.3.1

Laboratory Investigation of HUVEC Imaging in Proof-of-Concept Study

Method The laboratory investigation of this model utilized ideal cell incubation conditions to optimize the pre-targeting and SPAAC processes to ensure that sufficiently high clickable reactivity was displayed by the azide group on the pre-targeted cyclic RGDyK before integrin-mediated endocytosis. The HUVECs were treated with the cyclic RGDyK peptides 1a–1d (50 µm) for 15 min at room temperature (RT). Pre-targeting with 1a at 4 °C to inhibit endocytosis was attempted, but at this temperature, the affinity of 1a to integrin was reduced significantly. Identical conditions were used to treat TAMRA-labeled cyclic RGDyK peptide 3 (Fig. 4.2)

60

S. Nomura et al.

Fig. 4.2 Structures of the functionalized peptide and N-glycan ligands

in order to estimate the amounts of the available azide function on the cell surface prior to the reaction with the second glycan probe 2a. Anti-PECAM antibody was used as a representative cell surface marker to evaluate cell surface fluorescence. From the total fluorescence on an entire cell, 57% of cell surface TAMRA-fluorescence was detected. This indicates the amount of available azide-tagged 1a on the cell surface. Cells were then washed to remove excess amounts of cyclic RGDyK peptides and the pre-targeted cells were further incubated with the sialoglycan ligand 2a (50 µm) for 30 min at 4 °C. 4% paraformaldehyde was used to fix the cells and confocal microscopy analysis was carried out after staining the nuclei with

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

61

Fig. 4.3 Imaging of HUVECs using both high- and low-affinity ligands and the strain-promoted azide-alkyne cycloaddition (SPAAC) reaction. The following ligand combinations were used for HUVEC labeling: A the glycan ligand 2a alone (red); B the RGDyK peptide 1a followed by 2a; C the RGDyK peptide 1a followed by 2a in the presence of an excess amount of disialo-N-glycan; D 1a in the presence of an excess amount of the RGDyK peptide followed by 2a; and E 1a followed by the asialoglycan ligand 2b (red). After treatment with the ligands, cells were fixed and stained with DAPI (blue). Scale bars indicate 20 µm. F Comparison of the fluorescent intensities measured in (A)–(E). Data are presented as the means ± S.E. [n = 10 (10,000 cells  10), one way ANOVA post hoc Tukey–Kramer’s test, *p < 0.01, **p < 0.05]

4′,6-diamidino-2-phenylindole (DAPI). Sialoglycan ligand 2a incubated under the same conditions without using cyclic RGDyK ligands 1a–1d to pre-target the cells was used as a control. Upon comparison of the cells treated with sialoglycan ligand 2a alone (Fig. 4.3A) and with cyclic RGDyK peptides 1a (Fig. 4.3B), fluorescence intensity was three times higher in cells pre-targeted with 1a compared to 2a alone (Fig. 4.3F; see also Fig. 4.1B, C). The direct microscopic imaging of the live cells without fixing the cells, Microscopic imaging of live cells after treatment with RGDyK peptide 1a, followed by sialoglycan 2a, without fixation, gave results similar to those with fixation. This result demonstrates that the method can be applied to live-cell imaging as the fixation process did not affect the imaging outcome. The in situ synthesized glycoconjugates were successfully eluted from the cell surface by treatment with KCl-HCl buffer. Analysis of the glycoconjugates confirmed the cell surface SPAAC reaction, and liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) profiles of cell surface products were consistent with those performed with authentic samples. The calculated yield

62

S. Nomura et al.

of the SPAAC reaction on the cell was 47% and was based on the amount of pre-targeted RGDyK ligand bound to the cells. It was important to inhibit the endocytosis of the pre-targeted 1a during the SPAAC reaction. To ensure inhibition, the click reaction was performed after the cells had been treated with 1a and were fixed with paraformaldehyde. Upon completing this process, the pre-targeted azide function on RGDyK 1a could react at a higher amount with the acetylene in glycan 2a. This significantly improved the fluorescence contrast in comparison with that shown in Fig. 4.3A, B. Of note, the SPAAC reaction was accelerated on the cell surface after ligand binding to receptors. This was demonstrated by conducting the reaction in a flask under the same conditions applied to the cell-based experiments in Fig. 4.3B. Applying 50 µm of 1a and 2a in phosphate-buffered saline (PBS) at 4 °C for 30 min yielded only trace amounts of the clicked product. The most effective RGDyK peptide for visualizing cells using red fluorescence was the one that contained the shortest ethylene glycol linker. 1a may have the optimal distance between the two ligands, mediated by the linker, upon binding to surface receptors during the SPAAC reaction [29]. Ligand saturation effect on strong and weak receptor interactions Ligand saturation experiments were conducted using HUVECs to delineate the importance of strong and weak ligand/receptor interactions when visualizing cells. Cell saturation was achieved by treating them with excess amounts of the unfunctionalized cyclic RGDyK or the sialoglycan ligands prior to performing the pre-targeting procedure. Saturation with unfunctionalized cyclic RGDyK or the sialoglycan ligands decreased the fluorescence intensities dramatically by 80% and 75%, respectively (Fig. 4.3C, D). For the SPAAC reaction on pre-targeted cells, the asialoglycan ligand 2b was used as it is not a ligand for PECAM. This resulted in a similar 78% decrease in fluorescence intensity (Fig. 4.3E). These results clearly show that each ligand/receptor interaction contributed to the SPAAC reaction on the cell, as shown in Fig. 4.1C. This allows HUVECs that simultaneously express two target receptors on the cell surface to be recognized with high sensitivity. Selective visualization Figure 4.4 illustrates the selective visualization of target cells. HUVECs with siRNA-attenuated PECAM expression levels (to about 50% of normal) were prepared for this demonstration. Good imaging contrast was not achieved in control HUVEC and PECAM-knockdown cells upon treatment with TAMRA-labeled RGDyK 3 (Fig. 4.4A). On the other hand, the pre-targeted method resulted in five times stronger fluorescence intensity in control HUVECs compared to PECAM-attenuated cells, despite an overall decrease in fluorescence intensity (Fig. 4.4B). Selective imaging with HeLa cells HUVECs were selectively imaged together with HeLa cells to further investigate the novel pre-targeted imaging method. Conventional imaging methods using the

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

63

Fig. 4.4 Selective imaging of HUVECs expressing both aVb3 integrin and the platelet endothelial cell adhesion molecule (PECAM). A HUVECs transfected with siRNA against PECAM (siPECAM) or nontargeted siRNA (siControl), were treated with TAMRA-labeled RGDyK ligand 3 (red). B HUVECs, transfected with siPECAM or siControl, were treated with 1a followed by 2a (red). C HeLa cells and HUVECs were treated with 3 (red) or with the pre-clicked product between 1a and 2a (red). D HeLa cells and HUVECs were treated with 1a followed by 2a (red). After treatment with the ligands, the cells were fixed and stained with DAPI (blue). Scale bars indicate 20 µm. All data are presented as the means ± S.E. [n = 10 (10,000 cells  10), Student’s t-test, *p < 0.001]

TAMRA-labeled RGDyK 3 are not able to distinguish between HUVEC and HeLa cells (Fig. 4.4C, left and middle). As HeLa cells overexpress integrins but not PECAM, the two receptors were targeted using the novel pre-targeted method, and ligands were applied to both the integrins and PECAM. This allowed for successful imaging of the HUVECs with a fluorescence intensity contrast that was four times greater relative to the HeLa cell background (Fig. 4.4D). Interestingly, better contrast was not obtained in the image collected after adding the TAMRA-labeled peptide 3 (Fig. 4.4C, left and middle) than when the pre-conjugated product of a reaction between peptide 1a and the glycan 2a (Fig. 4.4C, right) was used. As well, strong peptide interactions were found to override the effects of the weak glycan interaction

64

S. Nomura et al.

on the pre-conjugated molecule. The strongly interacting peptide components were the only mechanism by which interactions between the molecules could be controlled. Cell-selective targeting was therefore achieved through the combination of the strong and the weak ligand/receptor interactions and the in situ ligation on the cell surface to synthesize the desired glycoconjugate (see Fig. 4.4B, D).

4.4

Advantages of Low-Affinity Ligands

Selectivity is an important issue for ensuring high image contrast and much of the research has focused on the development of ligands that have a high affinity to the surface receptors of interest. Using a low-affinity ligand such as the glycan used in the conceptual study, provides the advantage of enabling excess labeled ligand to be washed away from the cells, while the ligand bound to the cell surface, even at a KD in the millimolar range, can be strongly attached via reaction with a pre-targeted high-affinity ligand on the cell surface. This allows the nonspecific background fluorescence to be minimized. The experiments related to the glycan-saturation experiments, discussed earlier, strongly implicate the importance of applying a weak glycan/lectin interaction to the development and utility of the novel method. This also highlights the criticality of in situ ligation of weakly interacting ligands to a pre-targeted, strongly interacting ligand on the target cell, in order to achieve high labeling affinity and specificity. The pre-linked glycan/peptide conjugate is not as advantageous as the weak glycan interaction, as the binding characteristics of the conjugate are dominated by the strongly interacting peptide component. The interactions of the whole molecule were then controlled only by the strong interaction. Because the affinity of a single molecule of glycan toward lectin is quite low, multivalency effects can be used both in vitro and in vivo to detect lectins. This is achieved by introducing glycoclusters onto proteins using dendrimers, liposome templates, or even microchips [30–37]. The novel pre-targeted method uses the weak interaction of the monovalent glycan to enhance labeling selectivity.

4.4.1

Antibody-Based Methods Versus Small Ligand Methods

Other methods that provide good sensitivity in detecting the interactions of different proteins are antibody-based methods such as homogeneous time-resolved fluorescence (HTRF) immunoassay [38] or proximity ligation assay [39]. These methods can provide high throughput screening and imaging of surface protein interactions. The newly developed pre-targeted method discussed in this chapter confers further

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

65

advantages by using the strong and weak combination of two small surface ligands. By working with a combination of ligands, this method is compatible with both in vitro (cell-based experiments) and in vivo applications (molecular imaging, to be discussed later in the chapter). In vivo applications are achievable through the favorable pharmacokinetics of low molecular weight ligands. In addition, near-infrared dyes, radioactive labels, and various other fluorescent labels can easily be optimized. These advantages allow for the novel method to use small ligands for the selective cell surface recognition processes in a wide range of applications beyond the conventional antibody-based strategy. The novel pre-targeted method also allows for efficient bridging of integrin to PECAM on a cell surface. The method can then be used to dynamically analyze the relative spatial arrangements of two and possibly more receptors on a cell surface by using linkers of various lengths. This is achieved due to the dynamic cell surface orientation and conformation of two receptors, i.e., PECAM and integrin [24, 40], allowing their ligands to react with each other when receptors are close enough. Cell surface dynamics, as indicated by the spatial arrangements of the target proteins and hence, the ligand-directed signaling pathways, could be monitored through imaging or even controlled simply by treating live cells with chemical reagents.

4.5

Discrimination of Various Cancerous and Non-cancerous Cells

Distinguishing cancerous and non-cancerous cells is a distinct advantage of the novel pre-targeted method. In the development of this method, investigation was carried out to differentiate between cancerous and non-cancerous cells that express avb3 integrins on their cell surface. The conventional cyclic RGDyK peptide, known as a “strongly interacting ligand,” cannot selectively image targeted cells, but it has been demonstrated that the pre-targeted method can distinguish cancer cells both in vitro and in vivo. The effectiveness of the pre-targeted method, in this case, depends on the N-glycan structures used as the “weakly interacting ligands” with specific lectins, and on the spatial arrangement between the ligand/receptor complexes on the target cell surface [41].

4.5.1

Selective Targeting in Vitro

To carry out this portion of the development of the pre-targeted method, six cell lines were selected as representative cancer cells: HeLaS3 (human cervical cancer cells) [42], A549 (adenocarcinomic human alveolar basal epithelial cells) [43], BxPC3 (human pancreatic cancer cells) [44], PC3 (human prostate cancer cells)

66

S. Nomura et al.

[45], and SW620 (human colon cancer cells) [46]; TIG3 (human fibroblast cells) [47] was selected as the non-cancerous cell. The goal in using these cell lines was to determine if the pre-targeted method could be generalized to multiple different types of cancer cells as all of these cell lines express avb3 integrins on their surface. For this investigation, we used the strongly interacting ligand, RGD peptide, and the weakly interacting ligand, N-glycan. Both ligands were designed and prepared as shown in Fig. 4.5 [29]. Various lengths of polyethylene glycol (PEG) linkers (i.e., PEG3, PEG5, PEG7, or PEG9) were used to introduce the azide function onto the cyclic RGDyK peptide ligands (Fig. 4.2, 1a–d) in order to investigate the effect of distance on the click reaction. This, in turn, dictates the degree of tolerance of the spatial arrangement between the ligand/receptor complexes in the context of the labeling reaction. Five biantennary N-glycans terminated by either N-acetylneuraminic acids 2a,c [either a(2,6)- or a(2,3)-linked to galactose], galactose 2b, Nacetylglucosamine 2d, or mannose 2e were applied to differentiate cancer cells on the basis of specifically expressed lectins. The fluorescent labels of tetramethylrhodamine (TAMRA) and DBCO were attached to the reducing ends of the glycan molecules through a short PEG linker, in a fashion similar to the proof-of-concept study described in Sect. 4.3. The optimal combination of both RGD and N-glycan ligands for imaging of the target cancer cells (Fig. 4.6) were identified by simultaneously screening four RGD peptides and the five glycan ligands in a library of cells in 96-well plates. Previously optimized cell incubation conditions for pre-targeting and the strain-promoted click reaction were applied to ensure that the azide group on the integrin-targeted cyclic RGDyK had a sufficiently high clickable reactivity on the

Fig. 4.5 Structures of the functionalized N-glycan ligands. Fluorescent labels and dibenzocyclooctyne both react selectively with the azide, and were attached to the reducing ends of the glycan molecules. Glycans are shown as symbols

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

67

cell surface [29]. The cells (5  104) in each well were initially treated with the cyclic RGDyK peptides 1a–d (50 µM) for 15 min at RT. After the cells were washed with medium to remove excess peptide, the pre-targeted cells were further

Fig. 4.6 Fingerprints providing the optimal cyclic RGDyK and N-glycan ligands for specific cell labeling. A Imaging with simple TAMRA-cyclic RGDyK ligand 3. Specific cell types [HeLaS3 (B), A549 (C), BxPC3 (D), PC3 (E), SW620 (F), and TIG3 (G)] were seeded on 96-well plates (5  104 cells/well) and treated with one of four cyclic RGDyK ligands (50 µM, RT, 15 min) or one of five fluorescently labeled N-glycan ligands (100 µM, 4 °C, 30 min). Fluorescence intensities were normalized against cell number (n = 4 plates). A–G Scale bars indicate 100 µm. Data presented as mean ± SEM

68

S. Nomura et al.

incubated with N-glycan ligands 2a–e (100 µM) for 30 min at 4 °C. The cells were then washed with medium and fixed with 4% paraformaldehyde prior to fluorescent microscopy analysis. Cells treated with (1) cyclic RGDyK ligand labeled with TAMRA or (2) only glycan ligands 2a–e were used as controls. The cyclic RGDyK ligand labeled with TAMRA fluorescence (ligand 3) has been widely used to visualize avb3 integrins on tumor cells. It was able to very clearly image all of the cell lines used in the experiment, but as expected, could not distinguish between the different cell lines (Fig. 4.6A). The N-glycan ligands 2a– e without the pre-targeting procedure did not label any of the cell lines (Fig. 4.6B– G, left vertical lines). Specific cell labeling, which was dependent on N-glycan structures, was achieved through the combined use of cyclic RGDyK peptides and N-glycan ligands. GlcNAc- and Man-terminated glycans (2d, e) on HeLaS3 cells (Fig. 4.6B), a(2,3)Sia-terminated glycan (2c) on A549 cells (Fig. 4.6C), Gal-, GlcNAc-, and Man-terminated glycans (2b, d, e) on BxPC3 cells (Fig. 4.6D), a(2,6)Sia-, a(2,3) Sia-, and GlcNAc-terminated glycans (2a, c, d) on SW620 cells (Fig. 4.6F), and a(2,6)Sia-, a(2,3)Sia-, GlcNAc-, and Man-terminated glycans (2a, c, d, e) on TIG3 cells (Fig. 4.6G) were all specifically labeled. PC3 cancer cells (Fig. 4.6E) were not targeted by any combination of the N-glycans used in this study. However, fingerprints of “pre-targeted” cell labeling profiles could discriminate among these six cell lines (five cancerous and one non-cancerous cell line), which are very difficult to distinguish using commonly applied, strongly interacting RGD peptides.

4.5.2

Effect of Spatial Arrangement on Cell Labeling

Cell labeling profiles were noted to be affected by both the structures of N-glycans and the linker lengths of the RGD peptide connected to the azide function. BxPC3 cells, for example, were only significantly labeled by the Man-terminated glycan (2e) when pre-targeted with the RGD derivative 1c (Fig. 4.6D) and non-cancerous TIG3 cells were only stained by the GlcNAc- or Man-terminated glycans (2d and e) when combined with RGD 1b (Fig. 4.6G). An optimal spatial arrangement is required between the cyclic RGDyK peptides and specific N-glycans in order to be “clicked” onto the target cell surface, and this may therefore be the ideal arrangement between the avb3 integrins and target lectins. Therefore, Fig. 4.6 may indicate not only the information derived from two cell-surface interactions of highand low-affinity ligands but also that of the spatial arrangement between two receptors (or receptor/ligand complexes). This may allow for differentiation, based on cell surface arrangement, of cells that express the same two receptors on their surface and could further be used to identify individual cell lines.

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

4.5.3

69

Glycan Specificities for Lectins

Table 4.1 provides a summary of the previously reported lectins expressed in these six cell lines. These lectins represent potential candidates for the N-glycans identified in Fig. 4.6 (vimentin and SP-D for GlcNAc- and Man-terminated glycans [2d, e]) for targeting HeLaS3 cells. Table 4.1 illustrates the multifactorial nature of imaging using these lectins and glycans. While A549 expresses the SP-D, the Man-terminated glycan 2e cannot image the cells. Alternatively, the galactose-terminated glycan 2b slightly stains PC3 cells regardless of the linker length of RGD derivatives 1a–d, while no representative galactose-binding lectins are expressed on the PC3 cell surface. It is likely that the spatial arrangement and distance between the integrin and lectins on the cell surface affect these interactions. In addition, cell surface binding is caused by both the glycan/lectin interaction and also by the hydrophobic or H-bonding interactions with various biomolecules on the cell surface of glycans. This occurs especially when the interaction is very weak. The analysis is complicated by multiple interactions of the glycan on the cell surface and the result cannot be explained simply by lectins expressed on the cell surface.

4.5.4

Selective Cancer Targeting in Vivo

In vivo investigation of the pre-targeted method based on the cell labeling profiles (i.e., cell-identifying fingerprints) obtained in Fig. 4.6, was also conducted to discriminate between cancers in mice (Fig. 4.7). Treatment with cells pre-targeted by

Table 4.1 Literature survey of lectins expressed on cancer and normal cells, as well as their glycan specificities Cell lines

Lectin expression and their known ligands 2,6-Sia 2,3-Sia Gal

HeLa

BxPC3

Siglec-3 [48] Siglec-10 [52] –

Galectin-1 [49] Galectin-1 [53] –

PC3

–

SW620

Siglec-7 [59] –

Galectin-1 [58] Galectin-8 [58] –

A549

TIG3

GlcNAc

Man SP-D [51]

–

Vimentin [50] –

Galectin-3 [55] –

Vimentin [56] –

DC-SIGN [57] –

–

Vimentin [60] Vimentin [61]

–

–

–

SP-D [54]

–

70

S. Nomura et al.

cyclic RGDyK peptide 1d resulted in HeLaS3 and A549 cancer cells that were selectively targeted by GlcNAc-terminated glycan (2d, Fig. 4.6B) and a(2,3) Sia-terminated glycan (2c, Fig. 4.6C), respectively. Near-infrared fluorescence dye Cy7.5 (Fig. 4.5) was used to clearly analyze the fluorescence signals derived from the pre-targeting protocol among the various biologically relevant background fluorescence from mouse tissues. Assuming 2 mL of blood volume, the amount of the RGD and N-glycan ligands administered to each mouse were adjusted to the amounts used in the cell experiments shown in Fig. 4.6, so that the pre-targeting and receptor-mediated bioorthogonal click reaction could proceed in the tumor regions. BALB/c nude mice were used for the experiment and were implanted with either HeLaS3 or A549 cells in the left shoulder regions. The mice were initially treated with cyclic RGDyK peptide 1d (intravenous injection, 150 nmol). At 30 min intervals, N-glycan ligands 4c and b (15 nmol) were administered intravenously. After 3 h, the mice were dissected, and the fluorescence of the cancerous regions was analyzed on an IVIS Spectrum in vivo imaging system (Fig. 4.7A). Non-specific accumulation of the fluorescently labeled glycans occurred, largely due to the primary and hydrophobic fluorophore structures, and caused an increase in background fluorescence. Importantly, both HeLaS3 and A549 tumors were fluorescently surface-labeled to a greater extent, more than 2.5 times, by injection of both functionalized cyclic RGDyK and the “matched” N-glycan ligands than by injection with RGDyK or N-glycan alone (i.e., without the pre-targeting procedure) or by non-targeted glycan combinations (Fig. 4.7B, C). These data clearly validate that the pre-targeted strategy can distinguish between cancerous cells in vivo.

4.6

Summary

In conclusion, a selective cell targeting strategy based on strong and weak ligand/ receptor interactions and in situ click ligation on a cell surface was developed in order to improve imaging accuracy and diagnostic utility, despite the presence of other background cells. Using this method, cells can be screened and identified efficiently on 96-well plates, and the optimal RGD peptide and glycan ligands can be imaged as “fingerprints.” Importantly, the development of this method profiled two cell surface receptors/ligands interactions, as well as their spatial orientation, both key factors in the method’s efficacy. Targeted cancers in xenografted mice could then be selectively imaged using these fingerprints. This novel pre-targeted method is unique in that it can be used both in vitro and in vivo. Future research could see this method be applied to any weak cell surface interactions, which are interactions that have been neglected and hence underutilized in imaging. The principles of this new method open the flood gates for other low-affinity ligands to be used to identify and selectively image cells.

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

71

Fig. 4.7 Selective labeling of cancer tissues in a mouse xenograft model. A Mice (n = 4) were treated intravenously with RGD peptide ligand 1d (150 nmol in 100 µL saline), and then after 30-min intervals, with N-glycan ligands 4c and 4d labeled with the near-infrared fluorescent dye Cy7.5 (15 nmol in 100 µL saline). After dissection, tumor regions [HeLaS3 (B) and A549 (C)] were imaged on a PerkinElmer IVIS Spectrum in vivo imaging system (ex: 640 nm; em: 710 nm). Data presented as mean ± SEM, *p < 0.05

72

S. Nomura et al.

References 1. James ML, Gambhir SS (2012) A molecular imaging primer: modalities, imaging agents, and applications. Physiol Rev 92(2):897–965. https://doi.org/10.1152/physrev.00049.2010 2. Glunde K, Artemov D, Penet M-F, Jacobs AM, Bhujwalla ZM (2010) Magnetic resonance spectroscopy in metabolic and molecular imaging and diagnosis of cancer. Chem Rev 110:3043–3059. https://doi.org/10.1021/cr9004007 3. Cai W, Chen X (2008) Multimodality molecular imaging of tumor angiogenesis. J Nucl Med 49:113S–128S. https://doi.org/10.2967/jnumed.107.045922 4. Ametamey SM, Honer M, Schubiger PA (2008) Molecular imaging with PET. Chem Rev 108:1501–1516 5. Weissleder R, Ntziachristos V (2003) Shedding light onto live molecular targets. Nat Med 9 (1):123–128 6. Massoud TF, Gambhir SS (2003) Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Genes Dev 17(5):545–580. https://doi.org/10.1101/gad. 1047403 7. Weissleder R, Mahmood U (2001) Molecular imaging. Radiology 219(2):316–333 8. Urano Y (2012) Novel live imaging techniques of cellular functions and in vivo tumors based on precise design of small molecule-based ‘activatable’ fluorescence probes. Curr Opin Chem Biol 16(5–6):602–608. https://doi.org/10.1016/j.cbpa.2012.10.023 9. Zhou Y, Chakraborty S, Liu S (2011) Radiolabeled cyclic RGD peptides as radiotracers for imaging tumors and thrombosis by SPECT. Theranostics 1:58–82 10. Alford R, Ogawa M, Choyke PL, Kobayashi H (2009) Molecular probes for the in vivo imaging of cancer. Mol BioSyst 5(11):1279–1291. https://doi.org/10.1039/b911307j 11. Tanaka K, Fukase K (2008) PET (positron emission tomography) imaging of biomolecules using metal-DOTA complexes: a new collaborative challenge by chemists, biologists, and physicians for future diagnostics and exploration of in vivo dynamics. Org Biomol Chem 6 (5):815–828. https://doi.org/10.1039/b718157b 12. Mankoff DA, Link JM, Linden HM, Sundararajan L, Krohn KA (2008) Tumor receptor imaging. J Nucl Med 49(Suppl 2):149S–163S. https://doi.org/10.2967/jnumed.107.045963 13. Gambhir SS (2002) Molecular imaging of cancer with positron emission tomography. Nat Rev Cancer 2(9):683–693. https://doi.org/10.1038/nrc882 14. Pierschbacher MD, Ruoslahti E (1984) Cell attachment activity of fibronectin can be duplicated by small synthetic fragments of the molecule. Nature 309:30–34 15. Liu Z, Yan Y, Liu S, Wang F, Chen X (2009) 18F, 64Cu, and 68 Ga labeled RGD-bombesin heterodimeric peptides for PET imaging of breast cancer. Bioconjugate Chem 20(5):1016– 1025 16. Chen X, Park R, Hou Y, Khankaldyyan V, Gonzales-Gomez I, Tohme M, Bading JR, Laug WE, Conti PS (2004) MicroPET imaging of brain tumor angiogenesis with 18F-labeled PEGylated RGD peptide. Eur J Nucl Med Mol Imaging 31(8):1081–1089. https://doi.org/10. 1007/s00259-003-1452-2 17. Liu Z, Huang J, Dong C, Cui L, Jin X, Jia B, Zhu Z, Li F, Wang F (2012) 99mTc-labeled RGD-BBN peptide for small-animal SPECT/CT of lung carcinoma. Mol Pharm 9(5):1409– 1417. https://doi.org/10.1021/mp200661t 18. Reardan DT, Meares CF, Goodwin DA, McTigue M, David GS, Stone MR, Leung JP, Bartholomew RM, Frincke JM (1985) Antibodies against metal chelates. Nature 316:265–268 19. Goldenberg DM, Chang CH, Rossi EA, McBride JW, Sharkey RM (2012) Pretargeted molecular imaging and radioimmunotherapy. Theranostics 2(5):523–540. https://doi.org/10. 7150/thno.3582 20. Carroll L, Evans HL, Aboagye EO, Spivey AC (2013) Bioorthogonal chemistry for pre-targeted molecular imaging–progress and prospects. Org Biomol Chem 11(35):5772– 5781. https://doi.org/10.1039/c3ob40897c

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

73

21. Lee SB, Kim HL, Jeong HJ, Lim ST, Sohn MH, Kim DW (2013) Mesoporous silica nanoparticle pretargeting for PET imaging based on a rapid bioorthogonal reaction in a living body. Angew Chem Int Ed 52(40):10549–10552. https://doi.org/10.1002/anie.201304026 22. Rossin R, Verkerk PR, van den Bosch SM, Vulders RC, Verel I, Lub J, Robillard MS (2010) In vivo chemistry for pretargeted tumor imaging in live mice. Angew Chem Int Ed 49 (19):3375–3378. https://doi.org/10.1002/anie.200906294 23. Devaraj NK, Upadhyay R, Haun JB, Hilderbrand SA, Weissleder R (2009) Fast and sensitive pretargeted labeling of cancer cells through a tetrazine/trans-cyclooctene cycloaddition. Angew Chem Int Ed 48(38):7013–7016. https://doi.org/10.1002/anie.200903233 24. Wong CWY, Wiedle G, Ballestrem C, Wehrle-Haller B, Etteldorf S, Bruckner M, Engelhardt B, Gisler TH, Imhof BA (2000) PECAM-1/CD31 trans-homophilic binding at the intercellular junctions is independent of its cytoplasmic domain; evidence for heterophilic interaction with integrin AVB3 in cis. Mol Biol Cell 11:3109–3121 25. Kitazume S, Imamaki R, Kurimoto A, Ogawa K, Kato M, Yamaguchi Y, Tanaka K, Ishida H, Ando H, Kiso M, Hashii N, Kawasaki N, Taniguchi N (2014) Interaction of platelet endothelial cell adhesion molecule (PECAM) with alpha 2,6-sialylated glycan regulates its cell surface residency and anti-apoptotic role. J Biol Chem 289(40):27604–27613. https://doi. org/10.1074/jbc.M114.563585 26. Agard NJ, Prescher JA, Bertozzi CRA (2004) A Strain-promoted [3 + 2] azide–alkyne cycloaddition for covalent modification of biomolecules in living systems. J Am Chem Soc 126(46):15046–15047 27. Sletten EM, Bertozzi CR (2011) From mechanism to mouse: a tale of two bioorthogonal reactions. Acc Chem Res 44(9):666–676 28. Ning X, Guo J, Wolfert MA, Boons GJ (2008) Visualizing metabolically labeled glycoconjugates of living cells by copper-free and fast huisgen cycloadditions. Angew Chem Int Ed 47(12):2253–2255. https://doi.org/10.1002/anie.200705456 29. Taichi M, Nomura S, Nakase I, Imamaki R, Kizuka Y, Ota F, Dohmae N, Kitazume S, Taniguchi N, Tanaka K (2017) In situ ligation of high- and low-affinity ligands to cell surface receptors enables highly selective recognition. Adv Sci 1700147. https://doi.org/10.1002/ advs.201700147 30. Tanaka K, Siwu ER, Minami K, Hasegawa K, Nozaki S, Kanayama Y, Koyama K, Chen WC, Paulson JC, Watanabe Y, Fukase K (2010) Noninvasive imaging of dendrimer-type N-glycan clusters: in vivo dynamics dependence on oligosaccharide structure. Angew Chem Int Ed 49 (44):8195–8200. https://doi.org/10.1002/anie.201000892 31. Roy R, Beak MG (2002) Glycodendrimers: novel glycotope isosteres unmasking sugar coding. Case study with T-antigen markers from breast cancer MUC1 glycoprotein. Rev Mol Biotechnol 90:291–309 32. Leeuwenburgh M, Marel GA, Overkleeft HS (2003) Olefin metathesis in glycobiology: new routes towards diverse neoglycoconjugates. Curr Opin Chem Biol 7(6):757–765. https://doi. org/10.1016/j.cbpa.2003.10.009 33. Renaudet O (2008) Recent advances on cyclopeptide-based glycoclusters. Mini-Rev Org Chem 5:274–286 34. Chabre YM, Roy R (2008) Recent trends in glycodendrimer syntheses and applications. Curr Top Med Chem 8(14):1237–1285 35. Chabre YM, Roy R (2010) Advances in carbohydrate chemistry and biochemistry. Adv Carbohydr Chem Biochem 63 36. Lee RT, Lee YC (2000) Affinity enhancement by multivalent lectin-carbohydrate interaction. Glycoconjugate J 17(7–9):543–551 37. Lundquist JJ, Toone EJ (2002) The cluster glycoside effect. Chem Rev 102(2):555–578 38. Degorce F, Card A, Soh S, Trinquet E, Knapik GP, Xie B (2009) HTRF: a technology tailored for drug discovery—a review of theoretical aspects and recent applications. Curr Chem Genom 3:22–32 39. Soderberg O, Gullberg M, Jarvius M, Ridderstrale K, Leuchowius KJ, Jarvius J, Wester K, Hydbring P, Bahram F, Larsson LG, Landegren U (2006) Direct observation of individual

74

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

S. Nomura et al. endogenous protein complexes in situ by proximity ligation. Nat Methods 3(12):995–1000. https://doi.org/10.1038/nmeth947 Tzima E, Irani-Tehrani M, Kiosses WB, Dejana E, Schultz DA, Engelhardt B, Cao G, DeLisser H, Schwartz MA (2005) A mechanosensory complex that mediates the endothelial cell response to fluid shear stress. Nature 437(7057):426–431. https://doi.org/10.1038/ nature03952 Nomura S, Egawa Y, Urano S, Tahara T, Watanabe Y, Tanaka K (2020) Cancer discrimination by on-cell N-glycan ligation. Communi Chem 3(1):26. https://doi.org/10. 1038/s42004-020-0270-9 Oba M, Fukushima S, Kanayama N, Aoyagi K, Nishiyama N, Koyama H, Kataoka K (2007) Cyclic RGD peptide-conjugated polyplex micelles as a targetable gene delivery system directed to cells possessing avb3 and avb5 integrins. Bioconjugate Chem 18(5):1415–1423. https://doi.org/10.1021/bc0700133 Hilderbrand SA, Weissleder R (2010) Near-infrared fluorescence: application to in vivo molecular imaging. Curr Opin Chem Biol 14(1):71–79. https://doi.org/10.1016/j.cbpa.2009. 09.029 Liu Z, Liu H, Ma T, Sun X, Shi J, Jia B, Sun Y, Zhan J, Zhang H, Zhu Z, Wang F (2014) Integrin alphavbeta(6)-targeted SPECT imaging for pancreatic cancer detection. J Nucl Med 55(6):989–994. https://doi.org/10.2967/jnumed.113.132969 Zheng DQ, Woodard AS, Tallini G, Languino LR (2000) Substrate specificity of alpha(v)beta (3) integrin-mediated cell migration and phosphatidylinositol 3-kinase/AKT pathway activation. J Biol Chem 275(32):24565–24574. https://doi.org/10.1074/jbc.M002646200 Lei Y, Huang K, Gao C, Lau QC, Pan H, Xie K, Li J, Liu R, Zhang T, Xie N, Nai HS, Wu H, Dong Q, Zhao X, Nice EC, Huang C, Wei Y (2011) Proteomics identification of ITGB3 as a key regulator in reactive oxygen species-induced migration and invasion of colorectal cancer cells. Mol Cell Proteomics 10(10):M110 005397. https://doi.org/10.1074/mcp.m110.005397 Asakura S, Niwa K, Tomozawa T, Jin Y, Madoiwa S, Sakata Y, Sakai T, Funayama H, Soe G, Forgerty F, Hirata H, Matsuda M (1997) Fibroblasts spread on immobilized fibrin monomer by mobilizing a a1-class integrin, together with a vitronectin receptor avb3 on their surface. J Biol Chem 272(13):8824–8829 Shimoda A, Tahara Y, Sawada SI, Sasaki Y, Akiyoshi K (2017) Glycan profiling analysis using evanescent-field fluorescence-assisted lectin array: importance of sugar recognition for cellular uptake of exosomes from mesenchymal stem cells. Biochem Biophys Res Commun 491(3):701–707. https://doi.org/10.1016/j.bbrc.2017.07.126 Vyakarnam A, Lenneman AJ, Lakkides KM, Patterson RJ, Wang JL (1998) A comparative nuclear localization study of galectin-1 with other splicing components. Exp Cell Res 242 (2):419–428. https://doi.org/10.1006/excr.1998.4111 Ise H, Kobayashi S, Goto M, Sato T, Kawakubo M, Takahashi M, Ikeda U, Akaike T (2010) Vimentin and desmin possess GlcNAc-binding lectin-like properties on cell surfaces. Glycobiology 20(7):843–864. https://doi.org/10.1093/glycob/cwq039 Oberley RE, Goss KL, Ault KA, Crouch EC, Snyder JM (2004) Surfactant protein D is present in the human female reproductive tract and inhibits Chlamydia trachomatis infection. Mol Hum Reprod 10(12):861–870. https://doi.org/10.1093/molehr/gah117 Li N, Zhang W, Wan T, Zhang J, Chen T, Yu Y, Wang J, Cao X (2001) Cloning and characterization of Siglec-10, a novel sialic acid binding member of the Ig superfamily, from human dendritic cells. J Biol Chem 276(30):28106–28112. https://doi.org/10.1074/jbc. M100467200 Nita-Lazar M, Banerjee A, Feng C, Amin MN, Frieman MB, Chen WH, Cross AS, Wang LX, Vasta GR (2015) Desialylation of airway epithelial cells during influenza virus infection enhances pneumococcal adhesion via galectin binding. Mol Immunol 65(1):1–16. https://doi. org/10.1016/j.molimm.2014.12.010 Takeuchi T, Misaki A, Fujita J, Sonobe H, Ohtsuki Y (2001) T-cadherin (CDH13, H-cadherin) expression downregulated surfactant protein D in bronchioloalveolar cells. Virchows Arch 438(4):370–375. https://doi.org/10.1007/s004280000385

4 In Situ Synthesis of Glycoconjugates on the Cell Surface …

75

55. Zhang L, Wang P, Qin Y, Cong Q, Shao C, Du Z, Ni X, Li P, Ding K (2017) RN1, a novel galectin-3 inhibitor, inhibits pancreatic cancer cell growth in vitro and in vivo via blocking galectin-3 associated signaling pathways. Oncogene 36(9):1297–1308. https://doi.org/10. 1038/onc.2016.306 56. Zhang Y, Zhou M, Wei H, Zhou H, He J, Lu Y, Wang D, Chen B, Zeng J, Peng W, Du F, Gong A, Xu M (2017) Furin promotes epithelial-mesenchymal transition in pancreatic cancer cells via Hippo-YAP pathway. Int J Oncol 50(4):1352–1362. https://doi.org/10.3892/ijo.2017. 3896 57. Bharadwaj U, Li M, Zhang R, Chen C, Yao Q (2007) Elevated interleukin-6 and G-CSF in human pancreatic cancer cell conditioned medium suppress dendritic cell differentiation and activation. Cancer Res 67(11):5479–5488. https://doi.org/10.1158/0008-5472.CAN-06-3963 58. Lahm H, André S, Hoeflich A, Fischer JR, Sordat B, Kaltner H, Wolf E, Gabius H-J (2001) Comprehensive galectin fingerprinting in a panel of 61 human tumor cell lines by RT-PCR and its implications for diagnostic and therapeutic procedures. J Cancer Res Clin Oncol 127 (6):375–386. https://doi.org/10.1007/s004320000207 59. Huang HC, Chao CC, Wu PH, Chung HY, Lee HY, Suen CS, Hwang MJ, Cai BH (1862) Kannagi R (2019) Epigenetic silencing of the synthesis of immunosuppressive Siglec ligand glycans by NF-kappaB/EZH2/YY1 axis in early-stage colon cancers. BBA-Gene Regul Mech 2:173–183. https://doi.org/10.1016/j.bbagrm.2019.01.002 60. Maamer-Azzabi A, Ndozangue-Touriguine O, Breard J (2013) Metastatic SW620 colon cancer cells are primed for death when detached and can be sensitized to anoikis by the BH3-mimetic ABT-737. Cell Death Dis 4:e801. https://doi.org/10.1038/cddis.2013.328 61. Shiga T, Shirasawa H, Shimizu K, Dezawa M, Masuda Y, Simizu B (2013) Normal human fibroblasts immortalized by introduction of human papillomavirus type 16 (HPV-16) E6–E7 genes. Microbiol Immunol 41:313–319

Chapter 5

Chemical Approach Toward Controlling of Transient Protein Interactions Junko Ohkanda

Abstract Intrinsically disordered proteins and regions (IDPs and IDRs) have emerged as a central class of regulatory factors in diverse cellular processes, and their dysregulation has been implicated in many diseases. Development of synthetic molecules capable of controlling of their transient and weak interactions is therefore desirable. This chapter describes our recent work on structure-based design of inhibitors for enzymes and adaptor proteins that are responsible for the posttranslational modification and the cellular translocation of IDPs. Identification of synthetic inhibitors through a chemical library screening exploiting the fluorescent spectroscopic binding assay for IDP proteins is also discussed.








Keywords Inhibitors Protein–protein interactions Intrinsically disordered proteins K-Ras Prenylation Fusicoccin 14-3-3 Circadian clock transcription factors




5.1










Introduction

In humans, approximately 20,687 protein-coding genes have been identified [1], and 311,962 protein-protein interactions (PPIs) are predicted [2]. The PPI network forms molecular basis of diverse signaling pathways that mediate cellular response to environmental and genetic signals. While recent proteome analyses have remarkably advanced the identification of human PPIs, the essential dynamic regulatory mechanisms, including their time-dependent switching and the directions of the interactions, still largely remain unknown [3]. Intrinsically disordered proteins and regions (IDPs and IDRs) are unfolded proteins that lack well-defined tertiary structures and are abundant in human proteomes [4]. In the PPI network, IDPs play major roles in controlling transient J. Ohkanda (&) Institute of Agriculture, Shinshu University, 8304 Minami-Minawa, Kami-Ina, Nagano 399-4598, Japan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_5

77

78

J. Ohkanda

interactions comprising about 15–45% of all interactions [5] and are associated with allosteric regulation [6], posttranslational modification (PTM) [7, 8], splicing [9], and liquid-liquid phase separation [10, 11]. The weak binding of IDPs and IDRs to their partner proteins with dissociation constants in the micromolar range and lifetimes of seconds triggers dynamic assembly of reversible and fast signaling transmission [12]. Because dysregulation of IDPs and IDRs have been implicated in many diseases, such as cancers and neurodegenerative diseases [13], IDPs and IDRs have emerged as a new class of potential drug targets [4, 13, 14]. Several low-molecular weight or peptidic inhibitors for IDPs including c-Myc, EWS-FLI1, p27, NUPR1, and a-synuclein have been successfully developed by yeast two-hybrid assay [15], surface plasmon resonance [16], NMR [17], thermal shift assay [18], and virtual screening [19]. However, their structural flexibility and complex regulatory mechanisms complicate the rational approach to chemical controlling of IDP function. We envisioned that chemical regulation of transient PPIs of IDPs with their PTM enzymes or with rigid adaptor proteins may be achievable by rational design of synthetic agents based on the molecular mechanism of PPIs. We also conceived that library screening methods based on a spectroscopic methodology would be useful for discovery of IDP inhibitors. For example, when structural information for association between a PTM enzyme and a IDP is available, designing of a bivalent compound that simultaneously binds to the PPI interface and the active site of the enzyme would be possible (Fig. 5.1a). In addition, since substantial number of IDPs tend to interact with relatively well-folded adaptor proteins, it would also be possible to design an inhibitor that binds to the PPI interface of the adaptor (Fig. 5.1b). Furthermore, high-throughput-formatted screening platform using recombinant IDP proteins would be useful for identification of synthetic compounds that bind to a flexible region and induce structural alteration (Fig. 5.1c). This chapter describes our recent efforts based on the strategies, which led us to obtain synthetic compounds capable of controlling IDP-mediated PPIs in cells.

5.2 5.2.1

Bivalent Inhibitors for Posttranslational Lipidation of K-Ras K-Ras Prenylation

The Ras proteins (H-Ras, N-Ras, K-Ras4A, K-Ras4B) are 21 kDa-GTPases encoded by three genes [20]. In normal cells, Ras cycles between its GTP (active)and its GDP (inactive)-bound states and plays a critical function by transmitting intracellular signals from growth factor receptors to signaling pathways, including the PI3K/Akt and the Raf/MEK/Erk cascades. Single amino acid mutations at

5 Chemical Approach Toward Controlling …

79

Fig. 5.1 Schematic presentation of synthetic molecule-based strategies of rational design and exploration of IDP-modulators

GTP-binding site render Ras GTPase deficient and therefore lock it in its GTP-bound state resulting in uncontrolled cell growth. Mutation in the Ras gene was first reported in cancer in 1982, and numerous studies on development of Ras-directed therapeutics have been reported [21]. Among them is K-Ras4B which has become an intense focus of therapeutics because K-Ras4B is the most frequently mutated Ras isoform in human cancers. Mutation in K-Ras4B promotes aggressive tumorigenesis, leading to drug resistance and poor clinical outcomes [22–24]. Recent efforts exploiting library screening have successfully identified synthetic K-Ras inhibitors [25–27], whereas rational approaches for designing K-Ras inhibitors remain challenging due to the lack of apparent binding cavities for small molecules. To execute its switching role for the growth and survival signals, Ras must localize near the plasma membrane. This translocation is promoted by the posttranslational lipid modification at the intrinsically disordered C-terminus of the protein. Two protein prenyltransferases are responsible for the lipid modification of the Ras super family. The structurally related farnesyltransferase (FTase) and type-I geranylgeranyl transferase (GGTase I) have received considerable attention for their potential in treating cancers and other diseases over the past few decades [28], and a number of synthetic inhibitors have been developed [29, 30]. FTase and GGTase I are heterodimeric zinc metalloenzymes consisting of identical 48 kDa a-subunits that are product of the same gene. These enzymes attach a C15 farnesyl or a C20 geranylgeranyl group to a thiol group on the C-terminal CAAX tetrapeptide (C = cysteine, AA = aliphatic dipeptide, X = Met, Ser, Gln for FTase, Leu, Phe for GGTase I) of corresponding substrate proteins

80

J. Ohkanda

Fig. 5.2 Schematic presentation of alternative prenylation of K-Ras and the dual inhibition strategy for farnesylation and geranylgeranylation of K-Ras using a bivalent compound mimicking the structure of the K-Ras C-terminus

[31]. As with other Ras isoforms, K-Ras4B is normally farnesylated by FTase at the thiol group of the C-terminal CVIM sequence (Fig. 5.2). However, disruption of K-Ras4B farnesylation by FTase inhibitors causes an alternative geranylgeranylation by GGTase I, which enables K-Ras4B to retain full biological activity [32]. This obstacle is explained by the electrostatic interaction between the acidic surface of the a-subunit of FTase and GGTase I and the polylysine domain near the C-terminus of K-Ras4B.

5.2.2

Dual Inhibitors for Transient Interaction of K-Ras with Protein Prenyltransferases

We hypothesized that a compound mimicking the C-terminal structure of K-Ras4B would simultaneously bind to the active site and the acidic surface of both FTase and GGTase I, resulting in inhibition of both farnesylation and geranylgeranylation of K-Ras4B by disrupting transient PPIs (Fig. 5.2). We anticipated that this anchoring strategy would deliver the surface-binding moiety to the featureless PPI interface, allowing for minimal design of protein surface-directed agents, which would help for cell permeation. To that end, a series of the bivalent compounds were first designed by linking of CVIM tetrapeptide to a gallate moiety through an alkyl spacer (1, Fig. 5.3). A significant improvement in inhibitory activity against farnesylation of K-Ras4B C-terminal peptide was observed comparing to CVIM tetrapeptide (Ki = 0.005 lM for 1, 1.10 lM for CVIM, respectively) [33]. Furthermore, the bivalent compound also exhibited submicromolar activity against geranylgeranylation, validating that the anchoring strategy is useful for controlling

5 Chemical Approach Toward Controlling …

81

Fig. 5.3 Chemical structures of tetrapeptide CVIM, FTI-277, and bivalent compounds (1-2)

transient PPIs of IDRs and enzymes. This study also showed that common enzyme surface structures can be targeted by dual enzyme inhibitors [33, 34]. Nevertheless, the poor cell permeability hampered further evaluation in cells. Thus, we next modified the chemical structure of 1 to improve cell-based activity. Peptidomimetic approaches for developing FTase inhibitors have been intensively studied over the past decades and have made a major contribution to cancer therapy [35]. The CVIM peptidomimetics were designed based on an extended conformation of the tetrapeptide bound to the active site of FTase. An earlier example is FTI-249 developed by Sebti and Hamilton [36], in which 4-amino benzoic acid was introduced to replace hydrophobic VI dipeptide moiety. The study showed that its methyl ester pro-drug form was found to be active in whole cells at 200 lM, although CVIM itself was inactive. Thus, we prepared a bivalent compound in which the CVIM moiety in 1 was replaced with FTI-249. Cell-based evaluation demonstrated that the resulting compound moderately inhibited HDJ-2 processing at a concentration of 100 lM in cells, indicating that this peptidomimetic modification partially improves cell permeability [37]. To further improve cell-based activity, we designed a new series of bivalent compounds, in which guanidyl groups were introduced to replace amino groups and conducted the structure-activity-relationship study. As a result, a biphenyl-based peptidomimetic-containing compound 2 was found to exhibit significant dual inhibition activity (Ki = 0.6 nM and 0.71 lM for FTase and GGTase I,

82

J. Ohkanda

Fig. 5.4 Effect of FTI-277 and 2 on processing K-Ras and Rap-1A and on induction of p21 [38]

respectively) [38]. Cell-based evaluation using NIH-3T3 cells stably expressing G12V-K-Ras demonstrated that 2 exhibited substantial inhibition of K-Ras processing, whereas only the biphenyl peptidomimetic FTI-277 was inactive (Fig. 5.4, upper row). Surprisingly, 2 did not interfere the native GGTase I processing of Rap-1A (Fig. 5.5, middle row). Rap-1A is a Ras-related protein with a C-terminal CLLL sequence and exclusively receives geranylgeranylation. The western blot analysis using anti-geranylgeranylated Rap-1A antibody showed that FTI-277 partially inhibited Rap-1A processing as previously reported [39], whereas 2 did not show apparent activity (Fig. 5.4, middle row). This result suggests that covalent linking of the gallate moiety to FTI-277 reduced the affinity of 2 for GGTase I, therefore 2 failed to inhibit geranylgeranylation of the native substrate, which comprises the C-terminal leucine residue that has higher affinity to the active site of GGTase I than methionine residue does. It has been shown that inhibition of FTase and GGTase I causes induction of cyclin-dependent kinas inhibitor p21 protein and G1 arrest [40]. As expected, both FTI-277 and 2 clearly induced p21 protein (Fig. 5.4, lower row), supporting their inhibitory potency against FTase and GGTase I in cells. A cell imaging experiment using the fluorescence visualization of PPI-visualization system (Fluoppi) [41] was performed to evaluate whether 2 inhibited K-Ras translocation to the plasma membrane. K-Ras(G12C) fused to an

Fig. 5.5 Fluorescence images of cultured HEK293 cells stably expressing K-Ras (G12C) fused to Ash-tag and cRaf (RBD) fused to humanized Azami-Green (hAG). Scale bars, 50 mm. Arrows in magnified window indicate puncta accumulated at the plasma membrane [38]

5 Chemical Approach Toward Controlling …

83

assembly helper protein (Ash-K-Ras(G12C)) and the Ras-binding domain of c-Raf (c-Raf(RBD)) fused to fluorescent Azami-Green (hAG-cRaf(RBD)) were stably expressed in cells, where Ash protein forms an oligomer and hAG protein forms a tetramer. As prenylation of clustered Ash-K-Ras(G12C) occurs and the tagged K-Ras proteins are promoted to localize near cytoplasmic membranes, tetrameric cRaf(RBD) is recruited to the membranes through PPIs with the tagged K-Ras to form a large protein assembly. This causes the formation of fluorescent puncta composed of these tagged proteins near the cell periphery (Fig. 5.5, left). When cells were treated with FTI-277, puncta were clearly observed near membranes (Fig. 5.5, middle), indicating that K-Ras translocation was not disrupted. In contrast, 2 reduced the degree of puncta formation (Fig. 5.5, right), demonstrating that K-Ras translocation, and thereby K-Ras-c-Raf interaction were inhibited [38]. These results clearly demonstrated that peptidomimetic-containing mid-sized bivalent agent 2 penetrates into cells and disrupts PPIs between the C-terminal disordered region of K-Ras and FTase and/or GGTase I. The rationally designed dual inhibitors of FTase and GGTase I were shown to be effective for selective inhibition of K-Ras prenylation over native geranylgeranylation by GGTase I. This strategy may provide a solution for overcoming the obstacles of FTase inhibitors in K-Ras-directed anticancer clinical applications, and may be applicable for other enzyme-substrate interactions, for example, kinases and phosphatases.

5.3

Fusicoccins for Controlling 14-3-3 Interactions

Chemical control of IDP function may also be achievable by disrupting PPIs with adaptor proteins. In this case, designing inhibitors can be straightforward as the IDP adaptor proteins are generally well folded, and their three-dimensional structural information are available in many cases. This section describes our studies of structure-based design of 14-3-3 inhibitors comprising a diterpene natural product and a peptide fragment and their intracellular synthesis.

5.3.1

14-3-3 Proteins

The 14-3-3 proteins are a family of regulatory proteins expressed in all eukaryotic cells. These highly conserved dimeric proteins participate in a wide range of cellular processes through binding interactions with hundreds of structurally and functionally diverse proteins including kinases, phosphatases, transmembrane receptors, and transcription factors [42–44]. Each 14-3-3 monomer is composed of nine a-helices with a shallow amphiphilic binding groove in the middle of the protein (Fig. 5.6a, top) [45]. The conserved residues form the dimer interface and

84

J. Ohkanda

a)

C

C

N

b)

N

c)

Fig. 5.6 a Crystal structure of human 14-3-3f protein (1QID, Ref. [45]). Top: Ribbon representation. Middle: Surface representation. Residues that are conserved among the seven human isoforms are highlighted in red. Bottom: Top view. b Overview of the X-ray crystal structure of plant 14-3-3 (white surface) bound to FC-A (pink surface) and PMA2 phosphopeptide (blue surface, 1O9F, Ref. [60]). c A close-up view of FC-A and the phosphopeptide in the ternary complex (1O9F, Ref. [60])

inner walls of the groove, whereas the variable residues are distributed on the outer surface (Fig. 5.6a). 14-3-3 proteins function as a “hub” protein in the cellular signaling pathways. Their interaction with many partner proteins is facilitated by posttranslational modification (i.e., phosphorylation) occurred at a serine or threonine residue in the 14-3-3 binding consensus motifs, which are generally present in partially disordered region of the binding partners [46]. There are three types of consensus motifs: RSXpSXP (mode-1), RXY/FXpSXP (mode-2), and the carboxy terminal motif pS/ pTX1-2-CO2H (mode-3), where pS and pT are phosphoserine and phosphothreonine, and X is any amino acid residue [47–49]. Binding of a 14-3-3 protein upon phosphorylation promotes PPIs with other 14-3-3 partner proteins, structural alteration, and cellular trafficking [44]. Hundreds of intracellular protein ligands have been identified, including major oncogenic regulators, such as Ref. [50]. A number of studies have implicated dysregulation of 14-3-3 interactions in cancer [51] and neurologic diseases [52], suggesting 14-3-3 as a new therapeutic target [53]. Studies based on library screenings have successfully identified several 14-3-3 inhibitors, and among them is

5 Chemical Approach Toward Controlling …

85

a 20-mer peptide R18, which possesses nanomole affinity for 14-3-3 (Kd of *80 nM). [54] R18 occupies the amphipathic groove and thereby interferes with 14-3-3 and native phosphorylated IDPs interactions. Moreover, expression of a dimer form of R18 in cells induces apoptosis [55], thus validating the potential of 14-3-3 as a therapeutic target. Hydrophobic synthetic 14-3-3 inhibitors have been also developed [56]; however, these compounds show moderate activity in cells in the sub-millimolar range. The characteristic long and shallow amphiphilic binding groove makes 14-3-3 a difficult drug target. A potential approach for developing 14-3-3 inhibitors is to explore a reasonably sized compound in which multiple spatially oriented functional groups are appropriately placed, so that capable of recognizing the both hydrophilic and hydrophobic region in the amphiphilic binding groove. The X-ray crystal structure of 14-3-3 bound to R18 revealed that R18 binds via the amphipathic sequence WLDLE with the two hydrophilic acidic groups coordinating the basic residues (Lys49, Arg56, Arg60, and Arg127) that are involved in coordinating to a phospho group, and the hydrophobic tryptophan and leucine residues that bind to the hydrophobic cavity to which fusicoccin binds (Fig. 5.7) [57]. Inspired by the binding mode of R18, it was conceivable that covalent linking of an appropriately designed hydrophilic and a hydrophobic fragment may work to generate bivalent 14-3-3 inhibitors.

Fig. 5.7 Crystal structure of human 14-3-3 bound to pentapeptide WLDLE, a part of R18 (gray rods, 1A38)

86

5.3.2

J. Ohkanda

Fusicoccin A

Fusicoccin A (FC, Fig. 5.8) is a phytotoxin produced by the pathogenic fungus Phomopsis amygdali [58]. FC binds to the complex of 14-3-3 and plant plasma membrane H+-ATPase and enhances the PPI, [59] although FC itself is a weak ligand for 14-3-3 (Kd * 66 lM) [60]. The crystal structure of plant 14-3-3 bound to FC and the PMA2 phosphopeptide (QSYpTV; the 14-3-3-binding consensus motif found in the C-terminal of plant H+-ATPase) clearly revealed that FC binds in the hydrophobic cavity adjacent to the peptide and forms a ternary complex (Fig. 5.6b, c) [61]. Formation of this ternary complex is driven by van der Waals interactions between FC and the valine residue of the PMA2 peptide, resulting in an increase in the affinity of both ligands for 14-3-3 of nearly two orders of magnitude. These results suggested that FC would be a suitable fragment for development of the proposed bivalent 14-3-3 inhibitors.

5.3.3

Fusicoccin-Peptide Conjugates as 14-3-3 Inhibitors

We envisioned that 14-3-3-templated ligation between two fragments, a fusicoccin and a peptide fragment comprising a complementary reactive group, might allow the efficient generation of amphipathic macromolecules that can bind to the shallow binding groove of 14-3-3 (Fig. 5.8). To test this hypothesis, we first designed and synthesized a series of FC-derivatives in which an epoxide group was introduced at the C-19 position using spacers of various lengths and a pentapeptide QSYDC based on PMA2 peptide in which Val and phosphoThr were replaced by Cys and Asp, respectively. An HPLC study showed that the reaction between these fragments proceeded to afford the corresponding conjugate and was clearly accelerated in the presence of an equimolar amount of 14-3-3 [61]. Importantly, the relative yield of the conjugate varied depending on the length of the spacer, indicating that the proximity effect between the epoxide group of FC and the thiol group in the peptide upon binding to 14-3-3 together is the determining factor for the 14-3-3-templated reaction. The intracellular generation of synthetic macromolecular agents using cell-permeable reactive compounds would be highly beneficial for overcoming the cell penetration issues of PPI inhibitors. Since epoxide-containing fragments may react with other biomolecules, more bioorthogonal reactions are desirable. Oxime ligation is widely used in in vitro [62] and in vivo [63] studies, and this bioorthogonal condensation produces a stable oxime conjugate. Thus, formyl-containing FC 3 and peptide 4, in which an amino oxy group was introduced into the C-terminus, were designed (Fig. 5.9). A HPLC evaluation demonstrated that the reaction of 3 and 4 producing 5 was significantly accelerated in the presence of 14-3-3 more than 50 times that of the protein-free control [64]. Strikingly, cell-based experiments demonstrated that oxime ligation proceeds in

5 Chemical Approach Toward Controlling …

87

Fig. 5.8 Chemical structures of fusicoccin A and PMA2 peptide and schematic representation of 14-3-3-templated chemical ligation

cells, and as a consequence, cell growth was significantly inhibited. Importantly, the resulting 5 was shown to significantly disrupt the interaction between 14-3-3 and c-Raf, validating the inhibition activity against PPIs of 14-3-3 in cells. The study

88

J. Ohkanda

Fig. 5.9 Chemical structures of the fragments designed for the oxime ligation [64]

suggests that intracellular generation of large molecules may serve as a new path to the chemical modulation of intracellular PPIs [64]. This approach was further elaborated by using the copper-free Huisgen reaction and incorporating 4,8-diazacyclononyne (DACN) [65] in the molecular design (Fig. 5.10). Heteroatoms embedded in the cycloalkyne ring system cause the distortion angle of the alkyne moiety to be smaller than in conventional difluorinated cyclooctyne, and thus the reaction proceeds more slowly, making DACN more selective toward the azide group and stable against biomolecular nucleophiles such as thiols. Interestingly, the cycloaddition reaction between 6 and 7 proceeds regioselectively in the presence of 14-3-3, producing 8a,b [66]. A theoretical model of the transition state generated by molecular dynamics simulation suggests that 8b might be the dominant product. The results of this study demonstrate that 14-3-3-templated synthesis is valuable for the stereoselective functionalization of fusicoccins.

5.4

Spectroscopic Library Screening for Identification of IDP Inhibitors

Screening-based approaches would be feasible for efficient identification of compounds that bind to IDPs which lack structural information. A challenge relies on the practical possibility to discover a compound which binds to an IDP of interest in a selective manner and alters its function in cells. To that end, two approaches based

5 Chemical Approach Toward Controlling …

89

Fig. 5.10 Chemical structures of the fragments designed for copper-free Huisgen cycloaddition and the resulting two regioisomers [66]

on cell-based and cell-free binding assays are required and should be in use complimentary. This section describes our recent effort on setting up a spectroscopic in vitro binding assay using recombinant IDP proteins. The resulting assay platform was subjected to a high-throughput library screening, which led to identification of a potent IDP inhibitor [67].

90

5.4.1

J. Ohkanda

Circadian Clock Transcription Factors, BMAL1 and CLOCK

Cell-based screenings allow for identifying a cell-permeable compound, although the results do not assure that the compound directly associate with the protein of interest in cells. In contrast, cell-free binding assays using recombinant protein of interest allows for identification of a compound that is capable of binding to the protein, although the results do not assure that it would be active in cells. In order to advance medicinal chemistry studies for untapped class of drug targets such as IDPs, it is necessary that these two approaches are complimentary used for finding a suitable lead compound. Studies based on cell-based screenings have successfully identified several low-molecular-weight IDP inhibitors [68]. On the other hand, access to the cell-free screening platforms has been limited because the experimental handling of recombinant IDP proteins remains challenging [69]. An in vitro fluorescence polarization (FP) assay using recombinant IDPs might be useful for exploration and functional evaluation of inhibitors via high-throughput screening (HTS) of chemical libraries. To that end, we focused on the circadian clock transcription factors, brain and muscle ARNT-like 1 (BMAL1) [70], and circadian locomotor output cycles kaput (CLOCK) [71]. Heterodimerization of BMAL1 and CLOCK followed by binding to E-box DNA plays the central role of the regulation of the 24-hour circadian clock of mammals. More than 30% of BMAL1 and nearly 60% of CLOCK are predicted to be disordered [72] by computational simulation of disordered probability [73]. BMAL1 and CLOCK form a heterodimer and bind to E-box regulatory elements in Period (Per1 and Per2) and Cryptochrome (Cry1 and Cry2) genes, activating their transcription during daytime. At night, the protein products PER and CRY accumulate, dimerize, translocate to the nucleus, and bind to BMAL1/CLOCK to reduce its transcriptional activity [74, 75]. This transcriptional feedback loop is key to the 24-hour circadian clock of mammals. The 69-kDa BMAL1 and 95-kDa CLOCK proteins contain basic helix-loop-helix (bHLH) domains, Per-ARNT-Sim (PAS) domains A and B, and transactivation domains (TADs). The bHLH is an N-terminal DNA-binding domain, whereas the flexible PAS domains adopt various conformations depending on bound ligands or interacting partners and serve as a versatile sensor in signaling pathways [76]. Moreover, study based on the X-ray crystal structure of the heterodimer clearly indicates that the PAS domains play important role for the dimer formation [77].

5.4.2

High-Throughput Screening of Chemical Library

Truncated recombinant proteins of BMAL1 and CLOCK including both bHLH and PAS domains expressed in E. coli were obtained as an inclusion bodies, which were

5 Chemical Approach Toward Controlling …

91

dissolved in urea and purified, respectively. The proteins were confirmed to dimerize and bind to an E-box DNA fragment with an affinity which is consistent with the value reported in the literature [77]. Screening of a chemical library led to identification of 5,8-quinoxialinedione (9), which significantly inhibited binding of BMAL1 and CLOCK to E-box DNA at low micromolar concentration. The results of biochemical experiments suggest that 9 covalently reacts with protein(s) and may modulate the dimer formation. Although further study is required in order to elucidate the details of the mechanism of action and improve cellular activity to our knowledge, this is the first example of a synthetic compound that disrupts binding of BMAL1/CLOCK to an E-box DNA fragment. This simple low-molecular-weight heterocycle may serve as a useful scaffold for the development of BMAL1/CLCOK inhibitors. Most importantly, this study suggests that appropriately truncated recombinant IDPs enable simple spectroscopic binding assay, which is in turn readily applicable for IDP-directed library screenings.

5.5

Conclusions and Outlook

Intrinsically disordered proteins and regions have rapidly emerged as a central class of element involved in diverse cellular processes. Synthetic molecules that enable control of their structural alteration, PTMs, translocations, and assembling processes are highly desirable to advance our understanding of the details of functions and will hold a large promise for serving as a new class of potential therapeutics. Although development of IDP-directed agents still remains challenging, targeting their PTM enzymes and the adaptor proteins may provide a way toward chemical control of IDPs. In addition, continued effort on development of new methods and tools for in vitro evaluation using recombinant IDPs will be necessary, and it will certainly contribute to accelerate elucidation of the mechanism of IDP functions in-depth. I believe such chemical efforts are critical to unveil dynamic regulatory machinery of the living system. Acknowledgements I am grateful to my co-workers who contributed to the work described in this chapter. I would especially like to acknowledge R. Masuda and Y. Hosoya (Shinshu University), Profs. K. Tomooka, Y. Kawasaki, and K. Igawa (Kyushu University), Prof. T. Suzuki (Hokkaido University), and Prof. M. Imanishi (Kyoto University) for their dedication. This research was supported by the Japan Society for the Promotion of Science (18H02106, 20H04769) and the Ministry of Education, Culture, Sports, Science and Technology (18H04394, 20H04769), Astellas Foundation for Research on Metabolic Disorders, Naito Foundation, Uehara Foundation, and Koyanagi Foundation.

92

J. Ohkanda

References 1. The ENCODE Project Consortium (2012) An integrated encyclopedia of DNA elements in the human genome. Nature 489:57–74. https://doi.org/10.1038/nature11247 2. Stark C, Breitkreutz BJ, Reguly T, Boucher L, Breitkreuts A, Tyers M (2006) A general repository for interaction datasets. Nucleic Acids Res 34:D535–D539. https://doi.org/10. 1093/nar/gkj109 3. Silverbush D, Sharan R (2019) A systematic approach to orient the human protein–protein interaction network. Nature Commun 10:3015–3023. https://doi.org/10.1038/s41467-01910887-6 4. Uversky VN, Dunker AK (2010) Understanding protein non-folding. Biochim Biophys Acta 1804:1231–1264. https://doi.org/10.1016/j.bbapap.2010.01.017 5. Christoffer CW, Kihara D (2019) Modeling protein-protein interactions with intrinsically disordered proteins. In: Salvi N (ed) Intrinsically disordered proteins, 1st edn. Academic Press, Cambridge, pp 189–206 6. Berlow RB, Dyson HJ, Wright PE (2018) Expanding the paradigm: Intrinsically disordered proteins and allosteric regulation. J Mol Biol 430:2309–2320. https://doi.org/10.1016/j.jmb. 2018.04.003 7. Bah A, Vernon RM, Siddiqui Z, Krzeminski M, Muhandiran R, Zhao C, Sonenberg N, Kay LE, Forman-Kay JD (2015) Folding of an intrinsically disordered protein by phosphorylation as a regulatory switch. Nature 519:106–109. https://doi.org/10.1038/ nature13999 8. Tompa P, Davery NE, Gibson TJ, Babu MM (2014) A million peptide motifs for the molecular biologist. Mol Cell 55:161–169. https://doi.org/10.1016/j.molcel.2014.05.032 9. Hamann F, Schmitt A, Favretto F, Hofele R, Neumann P, Xiang S, Urlaub H, Zweckstetter M, Ficner R (2020) Structural analysis of the intrinsically disordered splicing factor Spp2 and its binding to the DEAH-box ATPase Prp2. Proc Natl Acad Sci USA 117:2948–2956. https:// doi.org/10.1073/pnas.1907960117 10. Hyman AA, Weber CA, Jülicher F (2014) Liquid-liquid phase separation in biology. Annu Rev Cell Dev Biol 30:39–58. https://doi.org/10.1146/annurev-cellbio-100913-013325 11. Shin Y, Brangwynne CP (2017) Liquid phase condensation in cell physiology and disease. Science 357:1253–1263. https://doi.org/10.1126/science.aaf4382 12. Nooren IMA, Thornton JM (2003) Structural characterisation and functional significance of transient protein-protein interactions. J Mol Biol 325:991–1018. https://doi.org/10.1016/ S0022-2836(02)01281-0 13. Iakoucheva LM, Brown CJ, Lawson JD, Obradovic Z, Dunker AK (2002) Intrinsic disorder in cell-signaling and cancer-associated proteins. J Mol Biol 323:573–584. https://doi.org/10. 1016/s0022-2836(02)00969-5 14. Theillet FX, Binolfi A, Frembgen-Kesner T, Hingorani, K. Sarkar,M. Kyne C, Li C, Crowley PB, Gierasch L, Pielak GJ, Elcock AH, Gershenson A, Selenko P (2014) Physicochemical properties of cells and their effects on intrinsically disordered proteins (IDPs). Chem Rev 114:6661–6714. https://doi.org/10.1021/cr400695p 15. Yin X, Giap C, Lazo JS, Prochownik EV (2003) Low molecular weight inhibitors of Myc-Max interaction and function. Oncogene 22:6151–6159. https://doi.org/10.1038/sj.onc. 1206641 16. Erkizan HV, Kong Y, Merchant M, Schlottmann S, Barber-Rotenberg JS, Yuan L, Abaan OD, Chou TH, Dakshanamurthy S. Brown ML, Uren A, Toretsky JA (2009) A small molecule blocking oncogenic protein EWS-FLI1 interaction with RNA helicase a inhibits growth of Ewing’s sarcoma. Nat Med 15:750–756. https://doi.org/10.1038/nm.1983 17. Iconaru LI, Ban D, Bharatham K, Ramanathan A, Zhang W, Shelat AA, Zuo J, Kriwacki RW (2015) Discovery of small molecules that inhibit the disordered protein, p 27Kip1 Sci Rep 5:15686–15701. https://doi.org/10.1038/srep15686

5 Chemical Approach Toward Controlling …

93

18. Neira JL, Bintz J, Arruebo M, Rizzuti B, Bonacci T, Vega S, Lanas A, Velázquez-Campoy A, Iovanna JL, Abián O (2017) Identification of a drug targeting an intrinsically disordered protein involved in pancreatic adenocarcinoma. Sci Rep 7:39732–39746. https://doi.org/10. 1038/srep39732 19. Tóth G, Gardai SJ, Zago W, Bertoncini CW, Cremades N, Roy SL, Tambe MA, Rochet J-C, Galvagnion C, Skibinski G, Finkbeiner S, Bova M, Regnstrom K, Chiou S-S, Johnston J, Callaway K, Anderson JP, Jobling MF, Buell AK, Yednock TA, Knowles TPJ, Vendruscolo M, Christodoulou J, Dobson CM, Schenk D, McConlogue L, McConlogue L (2014) Targeting the intrinsically disordered structural ensemble of a-synuclein by small molecules as a potential therapeutic strategy for Parkinson’s disease. PLoS One 9: https://doi. org/10.1371/journal.pone.0087133 20. Grand BA, Owen D (1991) The Biochemistry of ras p21. Biochem J 279:609–631. https:// doi.org/10.1042/bj2790609 21. Cox AD, Der CJ (2010) Ras history: the saga continues. Small GTPases 1:2–27. https://doi. org/10.4161/sgtp.1.1.12178 22. Bos JL (1989) ras oncogenes in human cancer: a review. Cancer Res 49:4682–4689 23. Lièvre A, Bachet J-B, Corre DL, Boige.V, Landi B, Emile J-F, Côté J-F, Tomasic G, Penna C, Ducreux M, Rougier P, Penault-Liorca F, Laurent-Puig P (2006) KRAS mutation status is predictive of response to cetuximab therapy in colorectal cancer. Cancer Res, 66:3992–3995. https://doi.org/10.1158/0008-5472 24. Cox AD, Fesik SW, Kimmelman AC, Luo J, Der CJ (2014) Drugging the undruggable Ras: mission possible? Nat Rev Drug Discov 13:828–851. https://doi.org/10.1038/nrd4389 25. Shima F, Yoshikawa Y, Ye M, Araki M, Matsumoto S, Liao J, Hu L, Sugimoto T, Ijiri Y, Takeda A, Nishiyama Y, Sato C, Muraoka S, Tamura A, Osoda T, Tsuda K, Miyakawa T, Fukunishi H, Shimada J, Kumasaka T, Yamamoto M, Kataoka T (2013) In silico discovery of small-molecule Ras inhibitors that display antitumor activity by blocking the Ras-effector interaction. Proc Natl Acad Sci USA 110:8182–8187. https://doi.org/10.1073/pnas. 1217730110 26. Ostrem JM, Peters U, Sos ML, Wells JA, Shokat KM (2013) K-Ras(G12C) inhibitors allosterically control GTP affinity and effector interactions. Nature 503:548–551. https://doi. org/10.1038/nature12796 27. Lim SM, Westover KD, Ficarro SB, Harrison RA, Choi HG, Pacold ME, Carrasco M, Hunter J, Kim ND, Xie T, Sim T, Jänne PA, Meyerson M, Marto JA, Engen JR, Gray NS (2014) Therapeutic targeting of oncogenic K-Ras by a covalent catalytic site inhibitor. Angew Chem Int Ed 53:199–204. https://doi.org/10.1002/anie.201307387 28. Gelb MH, Brunsveld L, Hrycyna CA, Michaelis S, Tamanoi F, Boorhis WCV, Waldmann H (2006) Therapeutic intervention based on protein prenylation and associated modifications. Nat Chem Biol 2:518–528. https://doi.org/10.1038/nchembio818 29. Palsuledesai CC, Distefano MD (2015) Protein prenylation: enzymes, therapeutics, and biotechnology applications. ACS Chem Biol 10:51–62. https://doi.org/10.1021/cb500791f 30. Berndt N, Hamilton AD, Sebti SM (2011) Targeting protein prenylation for cancer therapy. Nat Rev Cancer 11:775–791. https://doi.org/10.1038/nrc3151 31. Casey PJ, Seabra MC (1996) Protein prenyltransferases. J Biol Chem 271:5289–5292. https:// doi.org/10.1074/jbc.271.10.5289 32. Whyte DB, Kirschmeier P, Hockenberry TN, Nunez-Oliva I, James L, Catino JJ, Bishop WR, Pai JK (1997) K- and N-Ras are geranylgeranylated in cells treated with farnesyl protein transferase inhibitors. J Biol Chem 272:14459–14464. https://doi.org/10.1074/jbc.272.22. 14459 33. Machida S, Kato N, Harada K, Ohkanda J (2011) Bivalent inhibitors for disrupting protein surface-substrate interactions and for dual inhibition of protein prenyltransferases. J Am Chem Soc 133:958–963. https://doi.org/10.1021/ja1086112 34. Machida S, Usuba K, Blaskovich MA, Yano A, Harada K, Sebti SM, Kato N, Ohkanda J (2008) Module assembly for protein-surface recognition: geranylgeranyltransferase I bivalent

94

35.

36.

37.

38.

39.

40.

41.

42. 43. 44. 45.

46. 47.

48.

49.

50.

51.

J. Ohkanda inhibitors for simultaneous targeting of interior and exterior protein surfaces. Chem Eur J 14:1392–1401. https://doi.org/10.1002/chem.200701634 Knowles D, Sun J, Rosenberg S, Sebti SM, Hamilton AD (2001) Peptidomimetic-based inhibitors of farnesyltransferase. In: Sebti SM, Hamilton AD (eds) Farnesyltransferase inhibitors in cancer therapy. Humana Press, New Jersey, pp 49–64 Vogt A, Qian Y, Blaskovich MA, Fossum RD, Hamilton AD, Sebti SM (1995) A non-peptide mimetic of Ras-CAAX: selective inhibition of farnesyltransferase and Ras processing. J Biol Chem 270:660–664. https://doi.org/10.1074/jbc.270.2.660 Machida S, Tsubamoto M, Kato N, Harada K, Ohkanda J (2013) Peptidomimetic modification improves cell permeation of bivalent farnesyltransferase inhibitors. Bioorg Med Chem 21:4004–4010. https://doi.org/10.1016/j.bmc.2012.09.061 Tsubamoto M, Le TK, Li M, Watanabe T, Matsumi C, Parvatkar P, Fujii H, Kato N, Sun J, Ohkanda J (2019) A guanidyl-based bivalent peptidomimetic inhibits K-Ras prenylation and association with c-Raf. Chem Eur J 25:13531–13536. https://doi.org/10.1002/chem. 201903129 Sun J, Qian Y, Hamilton AD, Sebti SM (1998) Both farnesyltransferase and geranylgeranyltransferase I inhibitors are required for inhibition of oncogenic K-Ras prenylation but each alone is sufficient to suppress human tumor growth in nude mouse xenografts. Oncogene 16:1467–1473. https://doi.org/10.1038/sj.onc.1201656 Efuet ET, Keyomarsi K (2006) Farnesyl and geranylgeranyl transferase inhibitors induce G1 arrest by targeting the proteasome. Cancer Res 66:1040–1051. https://doi.org/10.1158/00085472.CAN-05-3416 Watanabe T, Seki T, Fukano T, Sakae-Sawano A, Karasawa S, Kubota M, Kurokawa H, Inoue K, Akatsuka J, Miyawaki A (2017) Genetic visualization of protein interactions harnessing liquid phase transitions. Sci Rep 7:46380–46392. https://doi.org/10.1038/ srep46380 Obsil T, Semin VO (2011) Structural basis of 14-3-3 protein functions Cell Dev. Biol 22:663– 672. https://doi.org/10.1016/j.semcdb.2011.09.001 Wiler E, Yaffe MB (2004) 14-3-3 Proteins—a focus on cancer and human disease. J Mol Cell 37:633–642. https://doi.org/10.1016/j.yjmcc.2004.04.015 Zhao J, Meyerkord CL, Du Y, Khuri FR, Fu HS (2011) 14-3-3 proteins as potential therapeutic targets. Cell Dev Biol 22:705–712. https://doi.org/10.1016/j.semcdb.2011.09.012 Rittinger K, Budman J, Xu J, Volinia S, Cantley LC, Smerdon SJ (1999) Structural analysis of 14-3-3 phosphopeptide complexes identifies a dual role for the nuclear export signal of 14-3-3 in ligand binding. Mol Cell 4:153–166. https://doi.org/10.1016/s1097-2765(00)80363-9 Sluchanko NN, Bustos DM (2019) In: Uversky VN (ed) Dancing protein clouds: intrinsically disordered proteins in health and diseases. Elsevier, San Diego, p 19 Muslin AJ, Tanner JW, Allen PM, Shaw AS (1996) Interaction of 14-3-3 with signaling proteins is mediated by the recognition of phosphoserine. Cell 84:889–897. https://doi.org/10. 1016/s0092-8674(00)81067-3 Yaffe MB, Rittinger K, Volinia S, Caron PR, Aitken A, Leffers H, Gamblin SJ, Smerdon SJ, Cantley LC (1997) The structural basis for 14-3-3: phosphopeptide binding specificity. Cell 91:961–971. https://doi.org/10.1016/s0092-8674(00)80487-0 Ganguly S, Weller JL, Ho A, Chemineau P, Malpaux B, Klein DC (2005) Melatonin synthesis: 14-3-3-dependent activation and inhibition of arylalkylamine N-acetyltransferase mediated by phosphoserine-205. Proc Natl Acad Sci USA 102:1222–1227. https://doi.org/10. 1073/pnas.0406871102 Freed E, Symons M; Macdonald SG, McCormick F, Ruggieri R (1994) Binding of 14-3-3 proteins to the protein kinase Raf and effects on its activation. Science 265:1713–1716. https://doi.org/10.1126/science.8085158 Neal CL, Xu J, Li P, Mori S, Yang J, Neal NN, Zhou X, Wyszomierski SL, Yu D (2012) Overexpression of 14-3-3f in cancer cells activates PI3K via binding the p85 regulatory subunit. Oncogene 31:897–906. https://doi.org/10.1038/onc.2011.284

5 Chemical Approach Toward Controlling …

95

52. Steinacker P, Aitken A, Otto M, Semin (2011) 14-3-3 proteins in neurodegeneration. Cell Dev Biol 22:696–704. https://doi.org/10.1016/j.semcdb.2011.08.005 53. Reinhardt HC, Yaffe MB (2013) Phospho-Ser/Thr-binding domains: navigating the cell cycle and DNA damage response. Nat Rev Mol Cell Biol 14:563–580. https://doi.org/10.1038/ nrm3640 and references therein 54. Wang B, Yang H, Liu YC, Jelinek T, Zhang L, Ruoslahti E, Fu H (1999) Isolation of high-affinity peptide antagonists of 14-3-3 proteins by phage display. Biochemistry 38:12499–12504. https://doi.org/10.1021/bi991353h 55. Masters SC, Fu H (2001) 14-3-3 proteins mediate an essential anti-apoptotic signal. J Biol Chem 276:45193–45200. https://doi.org/10.1074/jbc.M105971200 56. Stevers M, Sijbesma E, Botta M, MacKintosh C, Obsil T, Landrieu I, Cau Y, Wilson AJ, Karawajczyk A, Eickhoff J, Davis J, Hann M, O’Mahony G, Doveston RG, Brunsveld L, Ottmann C (2018) Modulators of 14-3-3 protein-protein interactions. J Med Chem 61:3755– 3778. https://doi.org/10.1021/acs.jmedchem.7b00574 57. Petosa C, Masters SC, Bankston LA, Pohl J, Wang B, Fu H, Liddington RC (1998) 14-3-3zeta binds a phosphorylated Raf peptide and an unphosphorylated peptide via its conserved amphipathic groove. J Biol Chem 273:16305–16310. https://doi.org/10.1074/jbc. 273.26.16305 58. Ballio A, Chain EB, Leo PD, Erlanger BF, Mauri M, Tonolo A (1964) Fusicoccin: a new wilting toxin produced by fusicoccum amygdali del. Nature 203:297. https://doi.org/10.1038/ 203297a0 59. Milroy LG, Brunsveld L, Ottmann C (2013) Stabilization and inhibition of protein-protein interactions: the 14–3-3 case study. ACS Chem Biol 8:27–35. https://doi.org/10.1021/ cb300599t 60. Würtele M, Ottmann CJ, Wittinghofer A, Oecking C (2003) Structural view of a fungal toxin acting on a 14-3-3 regulatory complex. EMBO J 22:987–994. https://doi.org/10.1093/emboj/ cdg104 61. Maki T, Kawamura A, Kato N, Ohkanda J (2013) Chemical ligation of epoxide-containing fusicoccins and peptide fragments guided by 14-3-3 protein. Mol BioSyst 9:940–943. https:// doi.org/10.1039/c2mb25388g 62. Rashidian M, Kumarapperuma SC, Gabrielse K, Fegan A, Wagner CR, Distefano MD (2013) Simultaneous dual protein labeling using a triorthogonal reagent. J Am Chem Soc 135:16388–16396. https://doi.org/10.1021/ja403813b 63. Tang L, Yin Q, Xu Y, Zhou Q, Cai K, Yen J, Dobrucki LW, Cheng J (2015) Bioorthogonal oxime ligation mediated in vivo cancer targeting. Chem Sci 6:2182–2186. https://doi.org/10. 1039/C5SC00063G 64. Parvatkar P, Kato N, Uesugi M, Sato S, Ohkanda J (2015) Intracellular generation of a diterpene-peptide conjugate that inhibits 14-3-3-mediated interactions. J Am Chem Soc 137:15624–15627. https://doi.org/10.1021/jacs.5b09817 65. Ni R, Mitsuda N, Kashiwagi T, Igawa K, Tomooka K (2015) Heteroatom-embedded medium-sized cycloalkynes: concise synthesis, structural analysis, and reactions. Angew Chem Int Ed 54:1190–1194. https://doi.org/10.1002/anie.201409910 66. Masuda R, Kawasaki Y, Igawa K, Manabe Y, Fujii H, Kato N, Tomooka K, Ohkanda J (2020) Copper-free Huisgen cycloaddition for the 14-3-3-templated synthesis of fusicoccin-peptide conjugates. Chem Asian J 15:742–747. http://dx.doi.org/0.1002/asia. 202000042 67. Hosoya Y, Nojo W, Kii I, Suzuki T, Imanishi M, Ohkanda J (2020) Identification of synthetic inhibitors for the DNA binding of intrinsically disordered circadian clock transcription factors. Chem Commun. https://doi.org/10.1039/d0cc04861e 68. Yu Y, Yu Q, Zhang X (2019) Allosteric inhibition of HIF-2a as a novel therapy for clear cell renal cell carcinoma. Drug Discov Today 24:2332–2440. https://doi.org/10.1016/j.drudis. 2019.09.008 69. Dunker AK, Lawson JD, Brown CJ, Williams RM, Romero P, Oh JS, Oldfield CJ, Campen AM, Ratliff CM, Hipps KW, Ausio J, Nissen MS, Reeves R, Kang C, Kissinger CR,

96

70.

71.

72. 73.

74.

75.

76. 77.

J. Ohkanda Bailey RW, Griswold MD, Chiu W, Obradovic Z (2001) Intrinsically disordered protein. J Mol Graph Model 19:26–59. https://doi.org/10.1016/s1093-3263(00)00138-8 Hogenesch JB, Gu Y-Z, Jain S, Bradfield CA (1998) The basic-helix–loop–helix-PAS orphan MOP3 forms transcriptionally active complexes with circadian and hypoxia factors. Proc Natl Acad Sci USA 95:5474–5479. https://doi.org/10.1073/pnas.95.10.5474 King DP, Zhao Y, Sangorma AM, Wilsbacher LD, Tanaka M, Antoch MP, Steeves TD, Vitaterna MH, Kornhauser JM, Lowrey PL, Turek FW, Takahashi JS (1997) Positional cloning of the mouse circadian clock gene. Cell 89:641–653. https://doi.org/10.1016/s00928674(00)80245-7 Tou WI, Chen CY-C (2014) May disordered protein cause serious drug side effect? Drug Discov Today 19:367–372. https://doi.org/10.1016/j.drudis.2013.10.020 Xue B, Dunbrack RL, Williams RW, Dunker AK, Uversky VN (2010) PONDR-FIT: a meta-predictor of intrinsically disordered amino acids. Biochim Biophys Acta 1804:996– 1010. https://doi.org/10.1016/j.bbapap.2010.01.011 Gekakis N, Staknis D, Nguyen HB, Davis FC, Wilsbacher LD, King DP, Takahashi JS, Weitz CJ (1998) Role of the CLOCK protein in the mammalian circadian mechanism. Science 280:1564–1569. https://doi.org/10.1126/science.280.5369.1564 Kume K, Zylka MJ, Sriram S, Shearman LP, Weaver DR, Jin X, Maywood ES, Hastings MH, Reppert SM (1999) mCRY1 and mCRY2 are essential components of the negative limb of the circadian clock feedback loop. Cell 98:193–205. https://doi.org/10.1016/s0092-8674(00) 81014-4 Möglich A, Ayers RA, Moffat K (2009) Structure and signaling mechanism of Per-ARNT-Sim domains. Structure 17:1282–1294. https://doi.org/10.1016/j.str.2009.08.011 Huang N, Chelliah Y, Shan Y, Taylor CA, Yoo S-H, Partch C, Green CB, Zhang H, Takahashi JS (2012) Crystal structure of the heterodimeric CLOCK:BMAL1 transcriptional activator complex. Science 337:189–194. https://doi.org/10.1126/science.1222804

Chapter 6

Mid-Sized Macrocyclic Peptides as a New Drug Modality Yuki Goto, Masanobu Nagano, and Hiroaki Suga

Abstract Even though peptides are a major class of therapeutics since the early twentieth century, they had been rather considered as an elusive modality due to their low proteolytic stability and poor membrane permeability. Recently, it has become increasingly evident that mid-sized macrocyclic peptides containing exotic building blocks could exhibit specific/strong binding to various protein targets. Particularly, recent advance in ribosomal construction of random peptide libraries and their selection-based screening has revolutionized the discovery process of such molecules. In this chapter, we comprehensively summarize the background of currently available technologies developed by our laboratory and their recent outcomes, and also disclose new emerging technologies for synthesis and screening of artificial macrocycles.







Keywords Macrocyclic peptide Genetic code reprogramming Posttranslational modification Peptidic drug In vitro selection




6.1




Introduction

Many peptides interact with physiologically relevant receptors and enzymes to exhibit various biological activities in organisms. Given their potential as bioactive molecules, various peptides/proteins have made significant contributions to drug development and therapeutic applications since the early twentieth century [1]. As a Y. Goto (&)
M. Nagano
H. Suga (&) Department of Chemistry, Graduate School of Science, The University of Tokyo, Bunkyo, Tokyo 113-0033, Japan e-mail: [email protected] H. Suga e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_6

97

98

Y. Goto et al.

representative, insulin, a hormone regulating the concentration of blood glucose, was marketed as the first peptidic drug in 1923 to treat diabetes. Since then, a number of peptides, such as the GLP-1 agonist liraglutide, the anabolic agent teriparatide, and the growth factor inhibitor octreotide, have been developed for the treatment of diverse diseases [2, 3]. However, peptidic molecules often suffer from low proteolytic stability and poor pharmacokinetics. Therefore, peptides had not been necessarily recognized as a promising class of therapeutic compounds. Indeed, most peptidic drugs developed in the twentieth century are limited to naturally occurring peptides (endogenous hormones and secondary metabolites) or their derivatives. Since the early 2000s, mid-sized macrocyclic peptides with non-proteinogenic building blocks have emerged as a new chemical modality, following conventional small molecules and biologics, such as antibodies [2–4]. This class of peptides generally affords a macrocyclic scaffold(s) rigidifying their global conformations, which provides an entropic gain and contributes to specific/strong binding to target proteins. Moreover, these peptides also have exotic amino acids, such as those with D-configuration and N-methylation, which possibly grants higher proteolytic stability and membrane permeability. For the discovery of such mid-sized macrocyclic peptides, our group and others have devised methodologies for construction of random peptide libraries and their screening to obtain de novo peptide ligands that can bind to targets of interest. In this chapter, we describe an overview of recent advances in our research focusing on selection-based discovery of macrocyclic peptide ligands with diverse chemical features and engineered in vitro biosynthesis systems allowing for ribosomal synthesis of peptides with expanded chemical diversity.

6.2

Discovery of Thioether-Closed Macrocyclic Peptide Ligands by the RaPID System

The contemporary methodologies for the discovery of de novo macrocyclic peptide ligands include the construction of chemically synthesized random peptide libraries (e.g., the one-bead-one-compound, OBOC, strategy), which can access broad chemical space (i.e., flexible choices of amino acids) but typically screen only *107 compounds [5] (Fig. 6.1a). Alternatively, in vitro display technologies such as phage display [6] and mRNA display [7, 8] construct genetically encoded combinatorial libraries with greater diversities (108–1013 members) that undergo efficient screening by in vitro selection, but the usable building blocks are generally restricted to the 20 proteinogenic amino acids. Thus, such common display methods require to utilize chemical linchpins [9, 10], which can posttranslationally crosslink between two or more sidechains of proteinogenic amino acids (e.g., Cys and Lys), to execute extrinsic macrocyclization of peptides (Fig. 6.1b, c). To overcome such a limitation, we have pursued a combination of the mRNA display method with the

6 Mid-Sized Macrocyclic Peptides as a New Drug Modality

99

Fig. 6.1 Comparison of contemporary methodologies for the discovery of de novo macrocyclic peptide ligands. a One-bead-one-compound libraries. b Phage display. c mRNA display. d RaPID system

genetic code reprogramming enabling us to utilize non-proteinogenic building blocks, which we refer to as the RaPID (random nonstandard peptides integrated discovery) system [11, 12] (Fig. 6.1d). This platform allows for incorporation of non-proteinogenic amino acids via genetic code reprogramming powered by the flexible in vitro translation (FIT) system [13], in which artificial tRNA aminoacylation ribozymes (so-called flexizymes) [14] are utilized as genetic code re-writing tools. In a typical RaPID screening (Fig. 6.2), a puromycin-ligated mRNA library bearing AUG-(NNK) n-UGC sequences is expressed in a Met-depleted FIT system supplemented with initiator tRNA aminoacylated with an N-chloroacetylated (ClAc) amino acid. In this engineered translation reaction, the AUG start codon is reprogrammed to encode the N-ClAc amino acid and the repeats of NNK mixed codons and the UGC codon encode random amino acid sequences and a Cys, respectively. The puromycin located at the 3′ end of each mRNA molecule attacks the C-terminal ester linkage of the nascent peptide chain upon translation termination, giving selective fusion of the translated precursor peptides with their cognate mRNA templates. The N-terminal ClAc group in the expressed precursors acts as an intrinsic chemical handle for spontaneous posttranslational cyclization with the thiol side chain of the designated downstream Cys residue to yield a thioether-closed macrocyclic peptide library [15, 16]. The resulting mRNA-displayed macrocyclic peptide library with over 1012 unique sequences is subjected to affinity selection against a target protein of interest, and the mRNAs connected to the recovered species are amplified by RT-PCR. Sequencing analysis of the cDNA library after 5–6 repetitions of this construction/selection/amplification cycle eventually identifies sequences of the objective macrocyclic peptide ligands specific for the target protein.

100

Y. Goto et al.

Fig. 6.2 In vitro selection of thioether-closed macrocyclic peptides by means of RaPID system

As a representative example, the RaPID selection against the histone demethylase KDM4A successfully identified a macrocyclic peptide, CP2, that showed tight binding (KD = 29.8 nM) to the target [17] (Fig. 6.3a). Importantly, CP2 exhibited potent inhibition of KDM4A (IC50 = 42 nM) with exceptional selectivity over not only enzymes in other KDMs/2OG oxygenase families (KDM1/ 2/3/5/6, PHD2, and FIH) but also intra-subfamily isoforms (KDM4D/E), which illustrates the potential of this peptide ligand to dissect the roles of opposing histone modifications and KDM4 isoforms in disease. Crystallographic analysis of CP2 in complex with KDM4A reveals that CP2 is compactly folded via several intramolecular hydrophobic and hydrogen-bonding interactions, adopting an appropriate conformation for tight binding to the active site of KDM4A. CP2 forms multiple interactions with KDM4A, including insertion of the Arg 6 side chain into the pocket usually occupied by the trimethyl-lysine side chain, thereby binding differently from but competing with histone substrates in the active site. Structure-guided derivatization of CP2 resulted in N-methylated cyclic peptides with improved proteolytic stability and membrane permeability, which indeed exhibits the intracellular activity altering the histone methylation status.

6 Mid-Sized Macrocyclic Peptides as a New Drug Modality

101

Fig. 6.3 Selected examples of macrocyclic peptide ligands obtained by the RaPID system. Non-proteinogenic building blocks in the peptides are highlighted in pink red

6.3

Development of Macrocyclic Peptide Ligands with Expanded Chemical Diversity by Genetic Code Reprogramming

The genetic code reprogramming technology allows for expanding the chemical diversity of the library to yield peptides bearing not only the thioether-closed macrocyclic scaffold but also various non-proteinogenic amino acids. For example, Passioura et al. intensively reprogrammed the genetic code to construct a library of macrocyclic peptides containing more than 10 distinct non-proteinogenic amino acids [18]. Although naturally occurring bioactive peptides often exhibit a strong bias for hydrophobic residues, which are presumably important for their potent bioactivities [19–21], normal in vitro display technologies have been restricted to the use of libraries composed of peptides containing mixtures of hydrophobic and hydrophilic/charged amino acids encoded by the standard genetic code. To circumvent this limitation, a RaPID library of macrocyclic peptides synthesized under a radically reprogrammed genetic code in which hydrophilic/charged proteinogenic

102

Y. Goto et al.

residues (Gln, Asn, Arg, Lys, Glu, and Asp) were replaced with 11 kinds of non-proteinogenic, relatively hydrophobic amino acids was constructed. Screening of this library against interleukin-6 receptor (IL6R) successfully identified a 10-mer macrocyclic peptide ligand containing 7 hydrophobic non-proteinogenic amino acids (Fig. 6.3b). This study has validated the feasibility of FIT-mediated genetic code reprogramming for the discovery of macrocyclic peptide ligands with elevated overall hydrophobicity and, in turn, improved pharmacokinetic profiles. The FIT-mediated genetic code reprogramming allows us to utilize not only non-proteinogenic a-L-amino acids but also more diverse exotic building blocks in the engineered translation reaction. Thus far, the FIT system has demonstrated ribosomal synthesis of peptides containing a variety of D-[22–24], b-[25–27], c-[28, 29], and N-methylated-amino acids [11]; amino carbothioic acids [30]; and even non-amino acids [31, 32]. As a noteworthy example, Katoh et al. recently demonstrated that the combinatorial use of an engineered suppressor tRNA (tRNAPro1E2) with elongation factor P (EF-P) drastically improves incorporation efficiencies of difficult substrates in translation [33, 34]. This engineered translation system enabled the construction and screening of a macrocyclic peptide library containing helix/turn-inducing cyclic b2,3-amino acids and identified human factor XIIa-inhibiting macrocycles that contained unique c-turn structures induced by two 2-aminocyclohexanecarboxylic acid (2-ACHC) residues [27] (Fig. 6.3c). Meanwhile, Okuma et al. took advantage of the inherent wide substrate tolerance of the translation initiation step to construct a peptide library with a structurally constrained cyclopropane-containing building block [35]. Utilization of an Nchloroacetylated cyclopropane-containing c-amino acid (ClAc-cCp) as the reprogrammed initiator altered the mode of posttranslational macrocyclization and yielded a macrocycle library with topologically different proportions from those of the conventional macrocyclic peptide libraries using ClAc-a-initiators (i.e., population of undesignated lariat-shaped peptides consisting of a small N-terminal cycle and a C-terminal tail was suppressed) because of the structural restriction of the cCp residue. The RaPID screening of this library against a phosphoglycerate mutase, for which we previously obtained a lariat-shaped inhibitor [36], demonstrated the development of macrocyclic peptides that inhibits the enzyme with a different mode of action from the previously reported peptide (Fig. 6.3d). As such, by appropriately designing engineered genetic codes with diverse exotic residues, we have demonstrated ribosomal construction and selection-based screening of macrocyclic peptide libraries composed of various building blocks beyond the proteinogenic repertoire. These works illustrate the potential of this platform technology to explore the previously inaccessible sequence space of mid-sized peptides bearing diverse structural features.

6 Mid-Sized Macrocyclic Peptides as a New Drug Modality

6.4

103

Upgrading In Vitro Biosynthesis System for Further Expansion of Accessible Modalities to Non-amide Backbones

Although the genetic code reprogramming by means of FIT system has enabled the use of diverse non-proteinogenic residues for library construction, the accessible backbone structures have still been limited to polycarbonyl structures such as amides, thioamides [30], and esters [31, 37]. However, hydrophilic polyamide backbones are inherently unfavorable for membrane permeability and not necessarily ideal for peptidic drug targeting intracellular proteins. In fact, naturally occurring peptides and bioactive peptidomimetics often possess non-amide backbone structures, such as azolines, azoles, and reduced amides. To further expand the range of accessible backbone variations by our in vitro biosynthesis strategy, we have devised new FIT systems by integrated with posttranslationally modifying enzymes and if necessary combined with chemical modification reactions (Fig. 6.4a). A representative example is the FIT-PatD system, in which the cyclodehydratase PatD, involved in the biosynthesis of azoline-containing cyanobactins, is supplemented to the FIT system [38]. In this system, a precursor peptide bearing Cys/Ser/ Thr residues is expressed from a synthetic DNA template and directly subjected to enzymatic posttranslational modification by PatD in a one-pot manner, where the Cys/Ser/Thr residues are converted to the corresponding azoline moieties to yield such a modified peptide (Fig. 6.4b). We have unveiled the unprecedented in vitro substrate tolerance of PatD and successfully produced a wide array of azoline-containing peptides [38, 39]. Remarkably, the FIT-PatD system also allows for PatD modification of non-proteinogenic amino acids to yield non-canonical heterocyclic backbones [40]. For instance, threonine analogs with arylated or alkylated b-carbons incorporated into precursors via genetic code reprogramming were successfully processed into substituted azolines by PatD. PatD also utilized an amino group as a nucleophile in the b-position to yield imidazoline, as well as amino and thiol groups at the c-carbon to obtain exotic six-membered backbone heterocycles (Fig. 6.4c). In addition, the FIT-PatD system has been used to generate the W[CH2NH] reduced amide backbone, which is a peptide isostere widely used in the development of bioactive pseudopeptides [41] (Fig. 6.4d). The thiazoline backbone produced in the FIT-PatD can be chemically converted to thiazolidine using a mild reducing agent NaCNBH3, and the resulting thiazolidine generates an imine intermediate in situ that can be further reduced to afford a W[CH2NH] structure. To the best of our knowledge, this work represents the first demonstration of chemoselective reduction of a designated amide bond in unprotected peptides in aqueous media. This chemoenzymatic modification is in sharp contrast to preceding methods for chemical reduction of amides, which generally require harsh reaction conditions and strong reductants. The wide substrate scope, site selectivity, and water-friendly mild reaction conditions of our chemoenzymatic modification

104

Y. Goto et al.

Fig. 6.4 Artificial in vitro biosynthesis systems for the synthesis of peptides with exotic backbone structures. a An overview of the FIT system integrated with enzymatic and chemical posttranslational modifications for generation of modified backbones. b Synthesis of azoline-containing peptides by the FIT-PatD system. c Non-canonical heterocyclic backbones accessible by the reprogrammed FIT-PatD system. d In vitro synthesis of W[CH2NH]-containing peptides by chemoenzymatic posttranslational modification. e An example of artificial thiopeptides synthesized by the reprogrammed FIT-Laz system. Non-proteinogenic residues and exotic backbones are highlighted in pink red and red, respectively

strategy offer template-dependent synthesis of diverse W[CH2NH]-containing peptides by means of in vitro translation. The FIT system has also been combined with multiple posttranslational modifying enzymes to achieve in vitro biosynthesis of peptidic natural products and their artificial analogs [42–44]. A prominent example is in vitro reconstitution of the whole biosynthetic pathway of a natural thiopeptide (lactazole A) by the FIT-Laz system [43]. The addition of five recombinantly produced lactazole biosynthetic enzymes (LazB/C/D/E/F) together with actinomycete GluRS and tRNAGlu into the FIT system enables a cascade of posttranslational modification reactions to produce an intensively modified peptide product bearing four azoles, two dehydroalanines, and a ring-closing pyridine moiety. Importantly, this system exhibits remarkable substrate tolerance and has been able to produce designer thiopeptide analogs with 10 consecutive mutations, 14- to 62-membered macrocycles, and 18 amino acid-long tail regions, as well as hybrid thiopeptides containing multiple non-proteinogenic amino acids (Fig. 6.4e). In principle, the combination of the FIT system with these chemical/enzymatic posttranslational modifications can be applied not only for the synthesis of rationally designed derivatives of natural products but also for the construction of randomized peptide libraries, which can be screened by an appropriate display technology to identify pseudo-natural peptides/products with the desired bioactivity. Such an approach can be coupled with various other posttranslational modification enzymes to expand the structural diversity accessible by our in vitro biosynthesis systems.

6 Mid-Sized Macrocyclic Peptides as a New Drug Modality

6.5

105

Summary and Outlook

In this chapter, we have comprehensively summarized showcases of selection-based discovery of artificial macrocyclic peptides and recent engineering of the in vitro translation system to produce peptides with non-amide backbones. The RaPID system has discovered many mid-sized macrocyclic peptides with high affinities and specificities for the protein targets of interest, and indeed contributed to application of this class of peptides as a new drug modality. However, the technology has not yet been perfect for the discovery of macrocyclic peptides with truly “drug-like” characteristics, i.e., cell membrane permeability and oral bioavailability. We envisage that further development of the related technologies, especially expansion of the structural diversity of peptides to those with more exotic or natural product-like entities, will be conducive to the establishment of a reliable platform to provide practical peptidic drugs.

References 1. Henninot A, Collins JC, Nuss JM (2018) The current state of peptide drug discovery: back to the future? J Med Chem 61:1382–1414. https://doi.org/10.1021/acs.jmedchem.7b00318 2. Santos GB, Ganesan A, Emery FS (2016) Oral administration of peptide-based drugs: beyond lipinski’s rule. ChemMedChem 11:2245–2251. https://doi.org/10.1002/cmdc.201600288 3. Lau JL, Dunn MK (2018) Therapeutic peptides: historical perspectives, current development trends, and future directions. Bioorg Med Chem 26:2700–2707. https://doi.org/10.1016/j. bmc.2017.06.052 4. Vinogradov AA, Yin YZ, Suga H (2019) Macrocyclic peptides as drug candidates: recent progress and remaining challenges. J Am Chem Soc 141:4167–4181. https://doi.org/10.1021/ jacs.8b13178 5. Qian Z, Upadhyaya P, Pei D (2015) Synthesis and screening of one-bead-one-compound cyclic peptide libraries. Methods Mol Biol 1248:39–53. https://doi.org/10.1007/978-1-49392020-4_3 6. Smith GP, Petrenko VA (1997) Phage display. Chem Rev 97:391–410. https://doi.org/10. 1021/cr960065d 7. Roberts RW, Szostak JW (1997) RNA-peptide fusions for the in vitro selection of peptides and proteins. Proc Natl Acad Sci USA 94:12297–12302. https://doi.org/10.1073/pnas.94.23. 12297 8. Nemoto N, Miyamoto-Sato E, Husimi Y, Yanagawa H (1997) In vitro virus: bonding of mRNA bearing puromycin at the 3’-terminal end to the C-terminal end of its encoded protein on the ribosome in vitro. FEBS Lett 414:405–408. https://doi.org/10.1016/s0014-5793(97) 01026-0 9. Heinis C, Winter G (2015) Encoded libraries of chemically modified peptides. Curr Opin Chem Biol 26:89–98. https://doi.org/10.1016/j.cbpa.2015.02.008 10. Derda R, Jafari MR (2018) Synthetic cross-linking of peptides: molecular linchpins for peptide cyclization. Protein Pept Lett 25:1051–1075. https://doi.org/10.2174/092986652566 6181120090650 11. Yamagishi Y, Shoji I, Miyagawa S, Kawakami T, Katoh T, Goto Y, Suga H (2011) Natural product-like macrocyclic N-methyl-peptide inhibitors against a ubiquitin ligase uncovered

106

12.

13. 14.

15.

16.

17.

18.

19.

20.

21. 22. 23.

24. 25. 26. 27.

28.

29.

Y. Goto et al. from a ribosome-expressed de novo library. Chem Biol 18:1562–1570. https://doi.org/10. 1016/j.chembiol.2011.09.013 Passioura T, Katoh T, Goto Y, Suga H (2014) Selection-based discovery of drug like macrocyclic peptides. Annu Rev Biochem 83:727–752. https://doi.org/10.1146/annurevbiochem-060713-035456 Goto Y, Katoh T, Suga H (2011) Flexizymes for genetic code reprogramming. Nat Protoc 6:779–790. https://doi.org/10.1038/nprot.2011.331 Murakami H, Ohta A, Ashigai H, Suga H (2006) A highly flexible tRNA acylation method for non-natural polypeptide synthesis. Nat Methods 3:357–359. https://doi.org/10.1038/ nmeth877 Goto Y, Ohta A, Sako Y, Yamagishi Y, Murakami H, Suga H (2008) Reprogramming the translation initiation for the synthesis of physiologically stable cyclic peptides. ACS Chem Biol 3:120–129. https://doi.org/10.1021/cb700233t Iwasaki K, Goto Y, Katoh T, Suga H (2012) Selective thioether macrocyclization of peptides having the N-terminal 2-chloroacetyl group and competing two or three cysteine residues in translation. Org Biomol Chem 10:5783–5786. https://doi.org/10.1039/c2ob25306b Kawamura A, Munzel M, Kojima T, Yapp C, Bhushan B, Goto Y, Tumber A, Katoh T, King ON, Passioura T, Walport LJ, Hatch SB, Madden S, Muller S, Brennan PE, Chowdhury R, Hopkinson RJ, Suga H, Schofield CJ (2017) Highly selective inhibition of histone demethylases by de novo macrocyclic peptides. Nat Commun 8:14773. https://doi. org/10.1038/ncomms14773 Passioura T, Liu W, Dunkelmann D, Higuchi T, Suga H (2018) Display selection of exotic macrocyclic peptides expressed under a radically reprogrammed 23 amino acid genetic code. J Am Chem Soc 140:11551–11555. https://doi.org/10.1021/jacs.8b03367 Ahlbach CL, Lexa KW, Bockus AT, Chen V, Crews P, Jacobson MP, Lokey RS (2015) Beyond cyclosporine A: conformation-dependent passive membrane permeabilities of cyclic peptide natural products. Future Med Chem 7:2121–2130. https://doi.org/10.4155/fmc.15.78 Bockus AT, Schwochert JA, Pye CR, Townsend CE, Sok V, Bednarek MA, Lokey RS (2015) Going out on a limb: delineating the effects of beta-branching, N-methylation, and side chain size on the passive permeability, solubility, and flexibility of sanguinamide A analogues. J Med Chem 58:7409–7418. https://doi.org/10.1021/acs.jmedchem.5b00919 Nielsen DS, Shepherd NE, Xu W, Lucke AJ, Stoermer MJ, Fairlie DP (2017) Orally absorbed cyclic peptides. Chem Rev 117:8094–8128. https://doi.org/10.1021/acs.chemrev.6b00838 Goto Y, Murakami H, Suga H (2008) Initiating translation with D-amino acids. RNA 14:1390–1398. https://doi.org/10.1261/rna.1020708 Fujino T, Goto Y, Suga H, Murakami H (2013) Reevaluation of the D-amino acid compatibility with the elongation event in translation. J Am Chem Soc 135:1830–1837. https://doi.org/10.1021/ja309570x Katoh T, Tajima K, Suga H (2017) Consecutive elongation of D-amino acids in translation. Cell Chem Biol 24:46–54. https://doi.org/10.1016/j.chembiol.2016.11.012 Fujino T, Goto Y, Suga H, Murakami H (2016) Ribosomal synthesis of peptides with multiple beta-amino acids. J Am Chem Soc 138:1962–1969. https://doi.org/10.1021/jacs.5b12482 Katoh T, Suga H (2018) Ribosomal incorporation of consecutive beta-amino acids. J Am Chem Soc 140:12159–12167. https://doi.org/10.1021/jacs.8b07247 Katoh T, Sengoku T, Hirata K, Ogata K, Suga H (2020) Ribosomal synthesis and de novo discovery of bioactive foldamer peptides containing cyclic beta-amino acids. Nat Chem. https://doi.org/10.1038/s41557-020-0525-1 Ohshiro Y, Nakajima E, Goto Y, Fuse S, Takahashi T, Doi T, Suga H (2011) Ribosomal synthesis of backbone-macrocyclic peptides containing gamma-amino acids. ChemBioChem 12:1183–1187. https://doi.org/10.1002/cbic.201100104 Katoh T, Suga H (2020) Ribosomal elongation of cyclic gamma-amino acids using a reprogrammed genetic code. J Am Chem Soc 142:4965–4969. https://doi.org/10.1021/jacs. 9b12280

6 Mid-Sized Macrocyclic Peptides as a New Drug Modality

107

30. Maini R, Kimura H, Takatsuji R, Katoh T, Goto Y, Suga H (2019) Ribosomal formation of thioamide bonds in polypeptide synthesis. J Am Chem Soc 141:20004–20008. https://doi.org/ 10.1021/jacs.9b11097 31. Ohta A, Murakami H, Higashimura E, Suga H (2007) Synthesis of polyester by means of genetic code reprogramming. Chem Biol 14:1315–1322. https://doi.org/10.1016/j.chembiol. 2007.10.015 32. Goto Y, Suga H (2009) Translation initiation with initiator tRNA charged with exotic peptides. J Am Chem Soc 131:5040–5041. https://doi.org/10.1021/ja900597d 33. Katoh T, Wohlgemuth I, Nagano M, Rodnina MV, Suga H (2016) Essential structural elements in tRNA(Pro) for EF-P-mediated alleviation of translation stalling. Nat Commun 7:11657. https://doi.org/10.1038/ncomms11657 34. Katoh T, Iwane Y, Suga H (2017) Logical engineering of D-arm and T-stem of tRNA that enhances d-amino acid incorporation. Nucleic Acids Res 45:12601–12610. https://doi.org/10. 1093/nar/gkx1129 35. Okuma R, Kuwahara T, Yoshikane T, Watanabe M, Dranchak P, Inglese J, Shuto S, Goto Y, Suga H (2020) A macrocyclic peptide library with a structurally constrained cyclopropane-containing building block leads to thiol-independent inhibitors of phosphoglycerate mutase. Chem Asian J. https://doi.org/10.1002/asia.202000700 36. Yu H, Dranchak P, Li Z, MacArthur R, Munson MS, Mehzabeen N, Baird NJ, Battalie KP, Ross D, Lovell S, Carlow CK, Suga H, Inglese J (2017) Macrocycle peptides delineate locked-open inhibition mechanism for microorganism phosphoglycerate mutases. Nat Commun 8:14932. https://doi.org/10.1038/ncomms14932 37. Desai BJ, Goto Y, Cembran A, Fedorov AA, Almo SC, Gao J, Suga H, Gerlt JA (2014) Investigating the role of a backbone to substrate hydrogen bond in OMP decarboxylase using a site-specific amide to ester substitution. Proc Natl Acad Sci USA 111:15066–15071. https:// doi.org/10.1073/pnas.1411772111 38. Goto Y, Ito Y, Kato Y, Tsunoda S, Suga H (2014) One-pot synthesis of azoline-containing peptides in a cell-free translation system integrated with a posttranslational cyclodehydratase. Chem Biol 21:766–774. https://doi.org/10.1016/j.chembiol.2014.04.008 39. Goto Y, Suga H (2016) A posttranslational cyclodehydratase, PatD, tolerates sequence variation in the C-terminal region of the substrate peptides. Chem Lett 45:1247–1249. https:// doi.org/10.1246/cl.160562 40. Goto Y, Suga H (2020) In vitro biosynthesis of peptides containing exotic azoline analogues. ChemBioChem 21:84–87. https://doi.org/10.1002/cbic.201900521 41. Kato Y, Kuroda T, Huang Y, Ohta R, Goto Y, Suga H (2020) Chemoenzymatic posttranslational modification reactions for the synthesis of Psi[CH2 NH]-containing peptides. Angew Chem Int Ed Engl 59:684–688. https://doi.org/10.1002/anie.201910894 42. Ozaki T, Yamashita K, Goto Y, Shimomura M, Hayashi S, Asamizu S, Sugai Y, Ikeda H, Suga H, Onaka H (2017) Dissection of goadsporin biosynthesis by in vitro reconstitution leading to designer analogues expressed in vivo. Nat Commun 8:14207. https://doi.org/10. 1038/ncomms14207 43. Vinogradov AA, Shimomura M, Goto Y, Ozaki T, Asamizu S, Sugai Y, Suga H, Onaka H (2020) Minimal lactazole scaffold for in vitro thiopeptide bioengineering. Nat Commun 11:2272. https://doi.org/10.1038/s41467-020-16145-4 44. Vinogradov AA, Shimomura M, Kano N, Goto Y, Onaka H, Suga H (2020) Promiscuous enzymes cooperate at the substrate level en route to lactazole A. J Am Chem Soc 142:13886– 13897. https://doi.org/10.1021/jacs.0c05541

Part II

Synthetic Advances in Complex Middle Molecules

Chapter 7

Convergent Total Synthesis of (+)-Cotylenin A Masahiro Uwamori, Ryunosuke Osada, Ryoji Sugiyama, Kotaro Nagatani, and Masahisa Nakada

Abstract Herein, the convergent total synthesis of (+)-cotylenin A is described. A retrosynthetic analysis of cotylenin A generated three fragments—A- and C-ring fragments, and a sugar moiety fragment. The A-ring fragment was prepared via a catalytic asymmetric intramolecular cyclopropanation developed in our laboratory, while the C-ring fragment was prepared via the modified acyl radical cyclization of a known chiral compound. The two fragments were successfully assembled by the Utimoto coupling reaction, while the B-ring, a carbocyclic eight-membered ring, was efficiently constructed by palladium-mediated cyclization, which was discovered during our synthesis of taxol. All hydroxy groups in the 5-8-5 tricyclic scaffold were stereoselectively introduced. Moreover, a new modified reducing reagent, Me4NBH(O2CiPr)3, was developed during the course of this study. The sugar moiety fragment was successfully prepared for the first time via the consecutive carbon–oxygen bond-forming reactions and was terminated by an epoxide opening reaction. Finally, the first enantioselective total synthesis of cotylenin A was successfully accomplished in a highly convergent manner via glycosylation using Wan’s protocol. Moreover, this is the first report to investigate the specific rotation of cotylenin A through the total synthesis.










Keywords Cascade reaction Convergent synthesis Cyclopropane Enantioselective Natural Product Palladium-catalyzed reaction Glycosylation










M. Uwamori
R. Osada
R. Sugiyama
K. Nagatani
M. Nakada (&) Department of Chemistry and Biochemistry, Faculty of Science and Engineering, Waseda University, 3-4-1 Ohkubo, Shinjuku-ku, Tokyo 169-8555, Japan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_7

111

112

7.1

M. Uwamori et al.

Introduction

Cotylenin A and fusicoccin A (Fig. 7.1) are structurally related diterpene glucosides, which were isolated from different sources of Cladosporium sp. 501-7 W [1–3] and Phomopsis amygdali [4], respectively. Although both these compounds exhibit potent plant growth activity, biological studies revealed that only cotylenin A induces the differentiation of murine and human myeloid leukemia cells and the apoptosis of a wide range of human cancer cell lines by the combined treatment with interferon-a [5–9]. Moreover, the combined treatment of cotylenin A with an anti-epidermal growth factor receptor antibody is reported to synergistically suppress tumor growth in vitro and in vivo, which provides a novel pharmacologic strategy for treatment of RAS mutant cancers [10]. The crystal structure of cotylenin A in a complex with 14-3-3 protein and a phosphopeptide of H+-ATPase (QSYpTV-COOH) [11] indicates that cotylenin A binds to inhibitory 14-3-3 interaction sites of C-RAF, pSer233, and pSer259 but not to the activating interaction site of pSer621. Members of the cotylenin family other than cotylenin A have been isolated, too. To the best of our knowledge, cotylenins B-J [12–15] have been reported as naturally occurring cotylenins thus far (Fig. 7.2). Most of these compounds are bioactive, but their biological properties have not yet been completely explored because of their limited availability. Among cotylenins, cotylenin A has attracted considerable research attention in the preceding decades because of its promising anti-cancer activity and unique mechanism of action. However, the loss of ability of Cladosporium sp. 501-7 W, the producer of cotylenin A, to proliferate during preservation on a slant has hindered further biological studies [16]. Hence, a steady supply of cotylenin A is needed for a comprehensive investigation of its biological properties. Total synthesis is a potentially effective strategy for obtaining biologically important derivatives of cotylenins that are difficult to prepare from naturally occurring cotylenins. However, despite the elucidation of the absolute structure of cotylenin A by X-ray crystallographic analysis in 1998, the total synthesis of cotylenin A or its congeners have not yet been reported [17–29], and only one total

Fig. 7.1 Structures of cotylenin A and fusicoccin A

7 Convergent Total Synthesis of (+)-Cotylenin A

113

Fig. 7.2 Structures of cotylenol and cotylenins B-J

synthesis of its aglycone, cotylenol, by Kato and co-workers has been reported till now [30, 31]. The fused 5-8-5 carbocyclic ring system of cotylenin A includes an all-carbon quaternary stereogenic center, an acid-sensitive chiral allylic tertiary alcohol, four contiguous stereogenic centers with a trans-1,2-diol, and a four-substituted alkene bearing an isopropyl group. Moreover, cotylenin A is connected with a structurally unique glucose-fused trioxabicyclo[2.2.1]heptane bearing methyl and epoxyethyl groups. These unique structural features make cotylenin A an important member among other fusicoccan diterpenoids. Cotylenin A has been an attractive synthetic target because of its intriguing biological activity and mechanism of action, limited availability, and unique structural features [17–31]. Hence, herein we developed the enantioselective total synthesis of cotylenin A, and describe the successful results [32].

7.2

Retrosynthetic Analysis of Cotylenin A

Our retrosynthetic analysis of cotylenin A is depicted in Schemes 7.1 and 7.2. The synthesis of cotylenin A was envisioned via the glycosylation of a protected cotylenol 1 and sugar moiety fragment 2. The trans-1,2-diol motif at the C8–C9

114

M. Uwamori et al.

position of 1 could be introduced by the stereoselective reduction of the a-hydroxyketone, which in turn could be prepared by stereoselective C9 hydroxylation of 3. The formidable eight-membered carbocyclic B-ring was envisaged to be formed by the palladium-catalyzed alkenylation of methyl ketone 4 because we have found that the palladium-catalyzed intramolecular alkenylation of a methyl ketone successfully formed the eight-membered carbocyclic ring of taxol in 96% yield (Scheme 7.7) [33]. Compound 4 could be derived from 5, which in turn could be prepared via the assembly of A-ring fragment 6 with C-ring fragment 7 (Scheme 7.2). We envisaged that 6 could be prepared via the stereoselective ring-opening reaction of cyclopropane 8 with sodium cyanide, and 8 in turn could be obtained by the catalytic asymmetric intramolecular cyclopropanation (CAIMCP) [34, 35] of 9, developed in our laboratory. C-ring fragment 7 would be prepared from ketone 10 which could be formed by the acyl radical cyclization [36] of a known chiral aldehyde 11 [37].

7.3

Enantioselective Preparation of 13

Scheme 7.3 shows the actual synthesis of the A-ring fragment. Compound 9, which was prepared from ethyl (E)-hex-4-enoate, was subjected to the CAIMCP to afford cyclopropane 8 in 86% yield and 86% ee. Compound 8 was subjected to sodium cyanide-induced cyclopropane ring-opening reaction to afford the highly crystalline b-keto sulfone 13, whose absolute structure was confirmed by X-ray crystallographic analysis [38].

Scheme 7.1 Retrosynthetic analysis of cotylenin A (I)

7 Convergent Total Synthesis of (+)-Cotylenin A

115

Scheme 7.2 Retrosynthetic analysis of cotylenin A (II)

Scheme 7.3 CAIMCP of 9 and preparation of 13

7.4

Enantioselective Preparation of 7

C-ring fragment 7 was synthesized from a known aldehyde 11 (Scheme 7.4) [37]. The acyl radical cyclization of 11 under the conditions described in ref. 36 gave 14, and the subsequent enol triflate formation afforded 15 in 40% yield. However, the acyl radical cyclization of 11 under the modified conditions discovered by us improved the yield, and enol triflate 15 was obtained in 54% yield (2 steps). The subsequent removal of the TBS group and Dess–Martin oxidation afforded 7.

116

M. Uwamori et al.

Scheme 7.4 Preparation of 7 via acyl radical cyclization

7.5

Preparation of 17

Next, we examined the coupling reaction of 13 with 7. Treatment of 13 with SmI2 afforded the enolate; however, the subsequent aldol reaction with 7 was not successful. Varying the reaction conditions, including the use of additives and other reducing reagents such as lithium naphthalenide and LiDBB, did not change the results. We then focused on the Utimoto coupling reaction [39–41] because it has been reported to proceed even when the reacting aldehyde is bulky, such as 7. Hence, the preparation of a-bromo ketone 17 required for the Utimoto coupling reaction was examined (Scheme 7.5). The enolate generated from 13 using SmI2 or lithium naphthalenide was again found unsuitable for the preparation of 17. However, the reaction of the enolate with acetic anhydride [42] afforded the corresponding enol acetate in 20-30% yield. This result encouraged us to examine other more effective trapping reagents, among which chloro diethylphosphate was found to be most suitable that afforded enol phosphate 16 in quantitative yield. Although the transformation of enol phosphate to a-bromo ketone has never been reported, the treatment of 16 with NBS was found to successfully afford a-bromo ketone 17 (82%).

Scheme 7.5 Preparation of 17

7 Convergent Total Synthesis of (+)-Cotylenin A

7.6

117

Preparation of Methyl Ketone 4

The Utimoto coupling reaction of 17 with 7 successfully afforded 18 as the sole product in 92% yield (Scheme 7.6). The dehydration of 18 with the Burgess reagent stereoselectively afforded enone 5. However, the subsequent Wittig methylenation resulted in the recovery of the starting material, probably because of enolization of 5. Hence, we employed the Takai reaction [43], and the desired exo-methylene compound 19 was successfully obtained. The subsequent dihydroxylation stereoselectively afforded 20 (46% (2 steps)). Further studies revealed that the use of ZrCl4 [44] instead of TiCl4, for methylenation, increased the yield to 71% (2 steps). The selective methylation of the primary hydroxyl group in 20 was successfully achieved by the use of the Meerwein reagent. Then, the acid-labile allylic tertiary

Scheme 7.6 Preparation of methyl ketone 4

118

M. Uwamori et al.

alcohol was protected as a TMS ether to furnish 21 by a one-flask operation. The subsequent conversion of 21 to methyl ketone 4 involved the reaction induced by methyl lithium or methyl magnesium bromide with the enol triflate in the C-ring moiety. Hence, 21 was converted to 22 by DIBAL-H reduction, followed by methylation and oxidation, in a stepwise manner to afford 4.

7.7

Palladium-Catalyzed Intramolecular Alkenylation

We had confirmed the efficient construction of the formidable carbocyclic eight-membered rings by the palladium-catalyzed intramolecular alkenylation of methyl ketones, which were discovered during the synthesis of taxol (Scheme 7.7) [18–29, 45]. Therefore, we examined the reaction of 4 under the same reaction conditions (Scheme 7.8) [18–29]. However, interestingly, no reaction occurred in this case. The reaction was then attempted under several other conditions that were effective for alkenylation, as reported in previous studies [18–29, 45]. However, the desired product was not formed. The structural difference between 4 and previously reported substrates is that 4 is an enol triflate and not an alkenyl halide. Hence, electron-rich tricyclohexylphosphine was employed as a suitable ligand considering that it would enhance the oxidative addition of palladium to enol triflate 4. Consequently, the reaction of 4 with 50 mol% of PdCl2(PCy3)2 at 100 °C led to the formation of 3 as expected. However, epimerization at the C7 position occurred, resulting in a mixture of 3 and its C7 epimer 3a (dr = 1.8:1) in 87% yield. The reaction at lower temperature (50 ° C) prevented the epimerization, but two equivalents of PdCl2(PCy3)2 were required to achieve 95% yield of 3 (Scheme 7.9). It should be noted that the enol triflate group survived through 10 steps (from 14 to 4) although its reaction is hindered by the adjacent all-carbon quaternary center and isopropyl group.

Scheme 7.7 Palladium-catalyzed intramolecular alkenylation of 23

7 Convergent Total Synthesis of (+)-Cotylenin A

119

Scheme 7.8 Attempted palladium-catalyzed intramolecular alkenylation of 4

Scheme 7.9 Successful palladium-catalyzed intramolecular alkenylation of 4

7.8

Enantioselective Total Synthesis of Cotylenol

We next examined the C9 hydroxylation of 3 (Scheme 7.10). It has been reported that C9 hydroxylation of 3 afforded a mixture of 25 and its C9 epimer in the moderate yield (64%) and diastereoselectivity (1.6/1) [30, 31]. However, we found that the reaction with MoOPH/LHMDS in the presence of LiCl improved the yield (71%) as well as diastereoselectivity (2.7/1).

Scheme 7.10 Enantioselective total synthesis of cotylenol

120

M. Uwamori et al.

Subsequently, the stereoselective reduction of a-hydroxy ketone 25 to trans1,2-diol 26 was conducted using NaBH(OAc)3 and Me4NBH(OAc)3 [46]. However, in both the cases, low reproducibility in the yield and stereoselectivity were observed [30, 31]. After several attempts, reduction using Me4NBH(O2CiPr)3, which was developed during the course of this study, was found to afford 26 as the single isomer. Me4NBH(O2CiPr)3 is easily prepared from commercially available Me4NBH4 and iPrCO2H. The high stereoselectivity and reproducibility, which were realized by the use of Me4NBH(O2CiPr)3, would be attributed to the bulky lipophilic ligand (O2CiPr) which could work favorably in the transition state and also improve solubility of the reagent. Finally, the removal of the TMS group of 26 with TBAF afforded cotylenol. All spectroscopic data of the synthesized product were identical to those reported for naturally occurring cotylenol [47, 48].

7.9

Preparation of Sugar Moiety Fragment 32

Next, we addressed the preparation of the sugar moiety fragment of cotylenin A (Scheme 7.11). This sugar moiety features the trioxabicyclo[2.2.1]heptane scaffold, which consists of a-hydroxy aldehyde 28a bearing an epoxyethyl group and glucose-derived a-hydroxy ketone 27 [32]. Hence, the reaction of 27 and 28a was expected to furnish 32. However, reactive 28a rapidly dimerized prior to the reaction with 27 under the given conditions. Thus, the desired product 32 could not be formed. The reaction of dimethyl acetal 28b with 27 did not afford the desired product either. Hence, an alternative fragment was required to construct the trioxabicyclo[2.2.1]heptane scaffold. After several attempts, epoxy aldehyde 28c [32] was designed as an alternative fragment for 28a because we envisioned that the reaction of 27 and 28c under acidic conditions could afford hemiacetal 29, and the generated hydroxy group of 29 could successively react with the keto group in the glucose unit to form new hemiacetals 30a and 30b. Finally, the resulting hydroxy group of the hemiacetal in 30a could undergo the intramolecular epoxide opening reaction to form compound 31. Thus, the reaction of 27 and 28c was expected to produce two diastereomers, 30a and 30b, and 30a could afford 31, while 30b could isomerize to the reacting diastereomer 30a via 29 under equilibrium conditions. The above-envisioned cascade transformations were realized as expected. The reaction of 27 and 28c afforded 29 in the presence of CSA in acetonitrile at 20–25 ° C. The concentration of the starting material (2 M) was important for the effective formation of 30a and 30b, indicating that the aggregation of the products could be a driving force for this reaction. Hemiacetals 30a and 30b were unstable to isolate and easily gave a mixture of 27 and 28c. Indeed, 30a and 30b could not be detected by TLC analysis, but 1H-NMR analysis confirmed their formation as a mixture of diastereomers (1:1). The reaction of 27 and 28c to afford 30a and 30b reached equilibrium after 30 min. Interestingly, after the formation of 30a and 30b in the 2 M solution, dilution of the reaction mixture to 0.1 M with acetonitrile shifted the

7 Convergent Total Synthesis of (+)-Cotylenin A

121

Scheme 7.11 Preparation of sugar moiety fragment 32

equilibrium to the starting materials 27 and 28c, resulting in the disappearance of 30a and 30b. This result also indicated that the aggregation of the products could be important for the stabilization of 30a and 30b in the 2 M solution. The acid-catalyzed intramolecular epoxide opening reaction with the hydroxy group in 30a proceeded slowly, and compound 31 was formed 24 h after the initiation of the reaction. Allowing the reaction to run for 48 h did not change the yield of 31, indicating that hemiacetal 30b, which did not afford the bis-acetal because of the stereochemical restriction, was not converted to 30a. These results could be attributed to the slow interconversion between 30a and 30b. Compound 31 was not fully purified by silica-gel column chromatography and contained a small amount of inseparable impurities. Hence, it was treated as is with sodium hydride to afford epoxide 32 in 23% yield (from 27 and 28c). The 23% yield is reasonable because the transformation from 27 and 28c to 32 includes the formation of four carbon–oxygen bonds; thus, the average yield of each step is ca. 70%. Moreover, this sequence was beneficial because 32 was obtained from 27 and

122

M. Uwamori et al.

28c through only a two-flask operation. Indeed, 32 was successfully prepared on a gram scale using this protocol.

7.10

Completion of the First Enantioselective Total Synthesis of Cotylenin A

After the successful preparation of the sugar moiety fragment 32, we next examined glycosylation with 32. It was found that the C8 alcohol was reactive and the glycosylation of 26 exclusively afforded the C8 glycoside. Hence, the protection of the C8 alcohol was needed, and compound 33, which was prepared in 59% yield (77% brsm) by the reaction of 26 and Ac2O, was subjected to glycosylation with 32 (Scheme 7.12). Glycosylations of 32 and 33 using common reagents such as Tf2O or MeOTf for the activation of thioglycoside induced the decomposition of 32, and Crich’s conditions [49] were used to afford 34, but unfortunately, the yield was low. After several attempts, we finally found that rhodium-catalyzed sulfonium ylide formation and the subsequent Brønsted-acid-catalyzed glycosylation, which were reported by Wan’s group [50], afforded 34 in moderate yield. Because compound 34 was inseparable from a small amount of impurities, which were finally removed in the last step of this synthesis, the subsequent three steps were carried out with slightly impure intermediates.

Scheme 7.12 Enantioselective total synthesis of cotylenin A

7 Convergent Total Synthesis of (+)-Cotylenin A

123

The C8 acetate of 34 resisted hydrolysis but was ultimately removed by the reaction with methyllithium at low temperature. Finally, the removal of the TMS and benzyl groups afforded cotylenin A. The spectroscopic data of the synthesized compound were identical to those of cotylenin A [1–3], indicating that the enantioselective total synthesis of cotylenin A has been successfully accomplished. To the best of our knowledge, the specific rotation of cotylenin A has been reported for the first time in this study.

7.11

Conclusion

We successfully achieved the convergent enantioselective total synthesis of cotylenin A by 25 longest linear steps from geraniol. The A-ring fragment was prepared via the catalytic asymmetric intramolecular cyclopropanation developed in our laboratory, and the C-ring fragment was prepared from a known chiral compound via a modified acyl radical cyclization. These two fragments were then assembled by the Utimoto coupling reaction. The carbocyclic eight-membered ring of cotylenin A was efficiently constructed by palladium-mediated cyclization. All hydroxy groups were stereoselectively introduced in the scaffold. A modified reducing reagent, Me4NBH(O2CiPr)3, was also developed during the course of this synthesis. The sugar moiety fragment was successfully prepared via three consecutive carbon– oxygen bond-forming reactions and it was terminated by an epoxide opening reaction. The glycosylation was accomplished using Wan’s protocol. Further studies to optimize the reaction conditions of the low-yielding steps are underway to facilitate the production of an adequate amount of cotylenin A and its derivatives for biological studies.

References 1. Sassa T, Tojyo T, Munakata K (1970) Isolation of a new plant growth substance with cytokinin-like activity. Nature 227:379. https://doi.org/10.1038/227379a0 2. Sassa T (1971) Cotylenins, leaf growth substances produced by a fungus, part I. Isolation and characterization of cotylenins A and B. Agric Biol Chem 35:1415–1418. https://doi.org/10. 1080/00021369.1971.10860078 3. Sassa T, Ooi T, Nukina M, Kato N (1998) Structural confirmation of cotylenin A, a novel fusicoccane-diterpene glycoside with potent plant growth-regulating activity from cladosporium fungus sp. 501-7W. Biosci Biotechnol Biochem 62:1815–1819. https://doi.org/10.1271/ bbb.62.1815 4. Ballio A, Brufani M, Casinovi CG, Cerrini S, Fedeli W, Pellicciari R, Santurbano B, Vaciago A (1968) The structure of fusicoccin A. Experientia 24:631–635. https://doi.org/10. 1007/BF02153818 5. Asahi K, Honma Y, Hazeki K, Sassa T, Kubohara Y, Sakurai A, Takahashi N (1997) Cotylenin A, a plant-growth regulator, induces the differentiation in murine and human

124

6.

7.

8.

9.

10.

11.

12. 13. 14.

15. 16.

17.

18.

19.

20.

21.

M. Uwamori et al. myeloid leukemia cells. Biochem Biophys Res Commun 238:758–763. https://doi.org/10. 1006/bbrc.1997.7385 Honma Y (2002) Cotylenin A—a plant growth regulator as a differentiation-inducing agent against myeloid leukemia. Leuk Lymphoma 43:1169–1178. https://doi.org/10.1080/ 10428190290026222 Honma Y, Ishii Y, Yamamoto-Yamaguchi Y, Sassa T, Asahi K (2003) Cotylenin A, a differentiation-inducing agent, and IFN-a cooperatively induce apoptosis and have an antitumor effect on human non-small cell lung carcinoma cells in nude mice. Cancer Res 63:3659–3666 Honma Y, Kasukabe T, Yamori T, Kato N, Sassa T (2005) Antitumor effect of cotylenin A plus interferon-a: possible therapeutic agents against ovary carcinoma. Gynecol Oncol 99:680–688. https://doi.org/10.1016/j.ygyno.2005.07.015 Matsunawa W, Ishii Y, Kasukabe T, Tomoyasu S, Ota H, Honma Y (2006) Cotylenin A-induced differentiation is independent of the transforming growth factor-b signaling system in human myeloid leukemia HL-60 cells. Leuk Lymphoma 47:733–740. https://doi.org/10. 1080/10428190500375839 Molzan M, Kasper S, Roeglin L, Skwarczynska M, Sassa T, Inoue T, Breitenbuecher F, Ohkanda J, Kato N, Schuler M, Ottmann C (2013) Stabilization of physical RAF/14-3-3 interaction by cotylenin A as treatment strategy for RAS mutant cancers. ACS Chem Biol 8:1869–1875. https://doi.org/10.1021/cb4003464 Ottmann C, Weyand M, Sassa T, Inoue T, Kato N, Wittinghofer A, Oecking C (2009) A structural rationale for selective stabilization of anti-tumor interactions of 14-3-3 proteins by cotylenin A. J Mol Biol 386:913–919. https://doi.org/10.1016/j.jmb.2009.01.005 Sassa T, Togashi M, Kitaguchi T (1975) The structures of cotylenins A, B, C, D and E. Agr Biol Chem 39:1735–1744. https://doi.org/10.1080/00021369.1975.10861845 Sassa T, Takahama A (1975) Isolation and identification of cotylenins F and G. Agr Biol Chem 39:2213–2215. https://doi.org/10.1080/00021369.1975.10861916 Takahama A, Sassa T, Ikeda M, Nukina M (1979) Isolation and structures of minor metabolites, cotylenins H and I. Agr Biol Chem 43:647–650. https://doi.org/10.1080/ 00021369.1979.10863477 Sassa T, Sakata Y, Nukina M, Ikeda M (1981) Germination-stimulating activity and chemical structure of cotylenin. Nippon Kagakukaishi: 895–898 Ono Y, Minami A, Noike M, Higuchi Y, Toyomasu T, Sassa T, Kato N, Dairi T (2011) Dioxygenases, key enzymes to determine the aglycon structures of fusicoccin and brassicicene, diterpene compounds produced by fungi. J Am Chem Soc 133:2548–2555. https://doi.org/10.1021/ja107785u Nagatani K, Hoshino Y, Tezuka H, Nakada M (2017) Enantioselective preparation of C-ring fragment of cotylenin A via catalytic asymmetric intramolecular cyclopropanation of a-diazo b-keto ester. Tetrahedron Lett 58:959–962. https://doi.org/10.1016/j.tetlet.2017.01.076 Kato N, Tanaka S, Takeshita H (1986) Total synthesis of cycloaraneosene, a fundamental hydrocarbon of epi-fusicoccane diterpenoids, and the structure revision of its congener, hydroxycycloaraneosene. Chem Lett 15(11):1989–1992. https://doi.org/10.1246/cl.1986.1989 Kato N, Tanaka S, Takeshita H (1988) Synthetic photochemistry. XLII. total synthesis of cycloaraneosene, a fundamental hydrocarbon of 5–8-5 membered tricyclic diterpenoid from Sordaria araneosa. Bull Chem Soc Jpn 61:3231–3237. https://doi.org/10.1246/bcsj.61.3231 Kato N, Nakanishi K, Wu X, Nishikawa H, Takeshita H (1994) Total synthesis of fusicogigantones A and B and fusicogigantepoxide via the singlet oxygen-oxidation of fusicoceadienes. “fusicogigantepoxide B”, a missing congener metabolite. Tetrahedron Lett 35:8205–8208. https://doi.org/10.1016/0040-4039(94)88283-5 Paquette LA, Sun L-Q, Friedrich D, Savage PB (1997) Highly enantioselective total synthesis of natural epoxydictymene. An alkoxy-directed cyclization route to highly strained trans-oxabicyclo[3.3.0]octanes. Tetrahedron Lett 38:195–198. https://doi.org/10.1016/ S0040-4039(96)02287-3

7 Convergent Total Synthesis of (+)-Cotylenin A

125

22. Paquette LA, Sun L-Q, Friedrich D, Savage PB (1997) Total synthesis of (+)epoxydictymene. application of alkoxy-directed cyclization to diterpenoid construction. J Am Chem Soc 119:8438–8450. https://doi.org/10.1021/ja971526v 23. Michalak K, Michalak M, Wicha J (2005) Studies towards the total synthesis of di- and sesterterpenes with dicyclopenta[a, d]cyclooctane skeletons: three-component approach to the A/B rings building block. Molecules 10:1084–1100. https://doi.org/10.3390/10091084 24. Williams DR, Robinson LA, Nevill CR, Reddy JP (2007) Strategies for the synthesis of fusicoccanes by Nazarov reactions of dolabelladienones: total synthesis of (+)-fusicoauritone. Angew Chem Int Ed 46:915–918. https://doi.org/10.1002/anie.200603853 25. Dake GR, Fenster EE, Patrick BO (2008) A synthetic approach to the fusicoccane A-B ring fragment based on a Pauson-Khand cycloaddition/Norrish type 1 fragmentation. J Org Chem 73:6711–6715. https://doi.org/10.1021/jo800933f 26. Srikrishna A, Nagaraju G (2011) Enantiospecific approach to AB-ring system of the diterpenes fusicoccanes. Indian J Chem, Sect B: Org Chem Incl Med Chem 50B:73–76 27. Fujitani B, Hanaya K, Higashibayashi S, Shoji M, Sugai T (2017) Construction of containing 4-keto-D-glucose skeleton. 2,6,9,11-tetraoxatricyclo[6.2.1.03,8]undecane Tetrahedron 73:7217–7222. https://doi.org/10.1016/j.tet.2017.11.008 28. Kuwata K, Hanaya K, Higashibayashi S, Sugai T, Shoji M (2017) Synthesis of the 1,2-seco fusicoccane diterpene skeleton by Stille coupling reaction between the highly functionalized A and C ring segments of cotylenin A. Tetrahedron 73:6039–6045. https://doi.org/10.1016/j. tet.2017.08.056 29. Kuwata K, Hanaya K, Sugai T, Shoji M (2017) Chemo-enzymatic synthesis of (R)5-hydroxymethyl-2-isopropyl-5-methylcyclopent-1-en-1-yl trifluoromethylsulfonate, a potential chiral building block for multicyclic terpenoids. Tetrahedron: Asymmetry 28:964–968. https://doi.org/10.1016/j.tetasy.2017.05.007 30. Okamoto H, Arita H, Kato N, Takeshita H (1994) Total synthesis of (-)-cotylenol, a fungal metabolite having a leaf growth activity. Chem Lett 23(12):2335–2338. https://doi.org/10. 1246/cl.1994.2335 31. Kato N, Okamoto H, Takeshita H (1996) Total synthesis of optically active cotylenol, a fungal metabolite having a leaf growth activity. Intramolecular ene reaction for an eight-membered ring formation. Tetrahedron 52:3921–3932. https://doi.org/10.1016/S00404020(96)00059-2 32. Uwamori M, Osada R, Sugiyama R, Nagatani K, NakadaM (2020) Enantioselective total synthesis of cotylenin A. J Am Chem Soc 142:5556–5561. https://dx.doi.org/10.1021/jacs. 0c01774 33. Hirai S, Utsugi M, Iwamoto M, Nakada M (2015) Formal total synthesis of (–)-taxol through Pd-catalyzed eight-membered carbocyclic ring formation. Chem-Eur J 21:355–359. https:// doi.org/10.1002/chem.201404295 34. Honma M, Sawada T, Fujisawa Y, Utsugi M, Watanabe H, Umino A, Matsumura T, Hagihara T, Takano M, Nakada M (2003) Asymmetric catalysis on the intramolecular cyclopropanation of a-diazo-b-keto sulfones. J Am Chem Soc 125:2860–2861. https://doi. org/10.1021/ja029534l 35. Honma M, Takeda H, Takano M, Nakada M (2009) Development of catalytic asymmetric intramolecular cyclopropanation of a-diazo-b-keto sulfones and applications to natural product synthesis. Synlett 11:1695–1712. https://doi.org/10.1055/s-0029-1217363 36. Yoshikai K, Hayama T, Nishimura K, Yamada K-I, Tomioka K (2005) Thiol-catalyzed acyl radical cyclization of alkenals. J Org Chem 70:681–683. https://doi.org/10.1021/jo048275a 37. Maruoka K, Ooi T, Nagahara S, Yamamoto H (1991) Organoaluminum-catalyzed rearrangement of epoxides A facile route to the synthesis of optically active b-siloxy aldehydes. Tetrahedron 47:6983–6998. https://doi.org/10.1016/S0040-4020(01)96153-8 38. CCDC 1983382 (13) contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.1983382

126

M. Uwamori et al.

39. Nozaki K, Oshima K, Utimoto K (1988) Facile routes to boron enolates. Et3B-mediated Reformatsky type reaction and three components coupling reaction of alkyl iodides, methyl vinyl ketone, and carbonyl compounds. Tetrahedron Lett 29:1041–1044. https://doi.org/10. 1016/0040-4039(88)85330-9 40. Tsuna K, Noguchi N, Nakada M (2011) Convergent total synthesis of (+)-ophiobolin A. Angew Chem Int Ed 50:9452–9455. https://doi.org/10.1002/anie.201104447 41. Tsuna K, Noguchi N, Nakada M (2013) Enantioselective total synthesis of (+)-ophiobolin A. Chem-Eur J 19:5476–5486. https://doi.org/10.1002/chem.201204119 42. Lebsack AD, Overman LE, Valentekovich RJ (2001) Enantioselective total synthesis of shahamin K. J Am Chem Soc 123:4851–4852. https://doi.org/10.1021/ja015802o 43. Takai K, Kakiuchi Y, Kataoka K, Utimoto K (1994) A novel catalytic effect of lead on the reduction of a zinc carbenoid with zinc metal leading to a geminal dizinc compound. Acceleration of the Wittig-type olefination with the RCHX2-TiCl4-Zn systems by addition of lead. J Org Chem 59:2668–2670. https://doi.org/10.1021/jo00089a002 44. Colby EA, O’Brien KC, Jamison TF (2004) Synthesis of amphidinolide T1 via catalytic, stereoselective macrocyclization. J Am Chem Soc 126:998–999. https://doi.org/10.1021/ ja039716v 45. Utsugi M, Kamada Y, Miyamoto H, Nakada M (2008) Synthetic studies on the taxane skeleton: effective construction of eight-membered carbocyclic ring by palladium-catalyzed intramolecular a-alkenylation of a methyl ketone. Tetrahedron Lett 49:4754–4757. https://doi. org/10.1016/j.tetlet.2008.05.105 46. Evans DA, Chapman KT, Carreira EM (1988) Directed reduction of ß-hydroxy ketones employing tetramethylammonium triacetoxyborohydride. J Am Chem Soc 110:3560–3578. https://doi.org/10.1021/ja00219a035 47. Sassa T, Negoro T, Ueki H (1972) The stereostructure of cotylenol, the aglycone of cotylenins leaf growth substances. Agr Biol Chem 36:2281–2285. https://doi.org/10.1080/00021369. 1972.10860584 48. Sassa T, Takahama A, Shindo T (1975) The stereostructure of cotylenol, the aglycone of cotylenins leaf growth substances. Agr Biol Chem 39:1729–1734. https://doi.org/10.1080/ 00021369.1975.10861844 49. Crish D, Smith M (2001) 1-Benzenesulfinyl piperidine/trifluoromethanesulfonic anhydride: a potent combination of shelf-stable reagents for the low-temperature conversion of thioglycosides to glycosyl triflates and for the formation of diverse glycosidic linkages. J Am Chem Soc 123:9015–9020. https://doi.org/10.1021/ja0111481 50. Meng L, Wu P, Fang J, Xiao Y, Xiao X, Tu G, Ma X, Teng S, Zeng J, Wan Q (2019) Glycosylation enabled by successive rhodium(II) and Brønsted acid catalysis. J Am Chem Soc 141:11775–11780. https://doi.org/10.1021/jacs.9b04619

Chapter 8

Electrochemical Synthesis of Oligosaccharides as Middle-Sized Molecules Md. Azadur Rahman, Kumpei Yano, Sujit Manmode, Yuta Isoda, Norihiko Sasaki, Toshiyuki Itoh, and Toshiki Nokami Abstract Automated electrochemical assembly, which is an electrochemical method for automated solution-phase synthesis of oligosaccharides, has been developed. The method is an enabling tool for preparation of chitooligosaccharides and their derivatives as biologically active middle-sized molecules. Electrochemical activation is useful to synthesize cyclic oligosaccharides and is combined with ionic liquid tag strategy to facilitate analysis and purification of synthesized oligosaccharides. Keywords Oligosaccharide tion Thioglycoside




8.1 8.1.1


Automated synthesis
Electrochemical glycosyla-

Introduction Biologically Active Oligosaccharides

Carbohydrates are one of the most abundant organic molecules and they form monosaccharides [1], oligosaccharides [2], and polysaccharides in nature [3]. All these forms of carbohydrates are biologically active; however, structure-activity relationship of many oligosaccharides and polysaccharides is still uncertain because of their structural complexity and diversity. Therefore, chemical synthesis of oligosaccharides is particularly important to elucidate their functions and the minimal structure of polysaccharides for biological activity. Chitin, which has 1,4-b-linked N-acetyl glucosamine as a repeating unit, is the second most abundant polysaccharides and its oligomers show biological activities (Fig. 8.1) [4]. For example, TMG-chitotriomycin, which is a pseudo-tetrasaccharide of chitin is known as an inhibitor of N-acetyl glucosaminidase [5, 6]. Mycorrhizal Md. A. Rahman
K. Yano
S. Manmode
Y. Isoda
N. Sasaki
T. Itoh
T. Nokami (&) Department of Chemistry and Biotechnology, Tottori University, 4-101 Koyamacho-Minami, Tottori City, 680-8552 Tottori, Japan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_8

127

128

Md. A. Rahman et al.

Fig. 8.1 Chitin and chitooligosaccharides

lipopolysaccharides (Myc-LCOs) and nodulation factors (Nod factors) have been reported as target compounds of total synthesis by several groups [7].

8.1.2

Automated Synthesis of Oligosaccharides

To access biologically active oligosaccharides efficiently, chemical synthesis of oligosaccharide especially automated synthesis is crucial. There have been several methods for automated synthesis of oligosaccharides based on solid-phase synthesis (Fig. 8.2) [8, 9]. Glycosyl acceptor is immobilized on the solid phase and chain elongation by the glycosylation and deprotection sequence is carried out on the solid materials such as synthetic polymers and porous metals. Methods for automated fluorous-phase synthesis have also been reported [10]. These methods facilitate chemical synthesis of complex oligosaccharides; however, productivity of these methods is lower than that of conventional solution-phase synthesis.

Fig. 8.2 Automated solid-phase synthesis

8 Electrochemical Synthesis of Oligosaccharides …

129

Fig. 8.3 Electrochemical glycosylation of aryl glycoside and thioglycoside

8.1.3

Electrochemical Glycosylation

Electrochemical methods for activation of glycosyl donors have been reported using aryl glycosides as glycosyl donors in the pioneering work of Noyori (Fig. 8.3) [11]. Thioglycosides, which have lower oxidation potentials than aryl glycosides, have also been utilized [12, 13]. Electrochemical activation of thioglycosides is a useful process with precise control of reaction; however, application of electrochemical glycosylation for oligosaccharide synthesis has been limited. We have been interested in electrochemical glycosylation because of its potential application for automated solution-phase synthesis.

8.2 8.2.1

Automated Electrochemical Assembly Electrochemical Generation of Glycosyl Triflates

Glycosyl triflates, which are highly reactive glycosylation intermediate, can be generated by electrochemical method [14–17]. Thus-generated glycosyl triflates can be accumulated, when anodic oxidation of thioglycosides was carried out using an divided electrolysis cell in the presence of tetrabutylammonium triflate (Bu4NOTf) at low temperature (Fig. 8.4). The cathodic reaction was reduction of proton to hydrogen gas. Therefore, thioglycosides were electrochemically converted to glycosyl triflates and hydrogen gas and diaryl disulfide were byproducts which did not disturb NMR analysis of glycosyl triflates. NMR measurements of electrochemically generated glycosyl triflates indicated that stereochemistry of accumulated glycosyl triflates was a-isomer, because of strong electron-withdrawing property of the anomeric substituent of triflate.

130

Md. A. Rahman et al.

Fig. 8.4 Electrochemical conversion of thioglycoside to glycosyl triflate

Fig. 8.5 Oligosaccharide synthesis via electrochemically generated glycosyl triflate

8.2.2

Automated Electrochemical Solution-Phase Synthesis of Oligosaccharides

Quantitative generation and accumulation of glycosyl triflates by electrochemical method encouraged us to use the protocol for iterative glycosylation using thioglycosides [18]. Dichloromethane solution of thioglycoside with a free-hydroxyl group was added to electrochemically generated glycosyl triflate and the corresponding disaccharide was obtained in 92% yield (Fig. 8.5). Thus-obtained disaccharide can be activated under the same electrochemical conditions and utilized for synthesis of longer oligosaccharides. We proposed automated solution-phase synthesis of oligosaccharide based on the electrochemical protocol. To perform the procedure by automated manner, we developed an automated electrochemical synthesizer by assembling commercially available instruments such as chiller, syringe pump, power supply and sequencer (programmable logic controller) (Fig. 8.6). Automated synthesis of oligoglucosamine up to hexasaccharide has been demonstrated as a proof of concept study.

8.3 8.3.1

Synthesis of Oligosaccharides TMG-Chitotriomycin

TMG-chitotriomycin, which is a selective inhibitor of N-acetyl glucosaminidase of insects and fungi, is a pseudo-chitotetrasaccharide with N,N,N-trimethyl-glucosaminium (TMG) part at the non-reducing end of the tetrasaccharide [5]. A potential precursor of

8 Electrochemical Synthesis of Oligosaccharides …

Fig. 8.6 Automated synthesizer

solution-phase

synthesis

of

131

oligosaccharide

using

electrochemical

TMG-chitotriomycin can be prepared by automated electrochemical assembly; however, the precursor was obtained as a mixture of stereoisomers [19]. The revised synthesis was started with b-isomer of disaccharide to prepare the precursor with a single stereoisomer. Further manipulation of the precursor afforded TMG-chitotriomycin in 21% yield (10 steps from disaccharide) (Fig. 8.7) [20].

Fig. 8.7 Total synthesis of TMG-chitotriomycin via automated electrochemical assembly

132

Md. A. Rahman et al.

Table 8.1 Effect of electrolyte on stereoselectivity of electrochemical glycosylation

entry 1 2 3 4 5 6 7

8.3.2

Bu4NOTf 0.10 M 0.025 M 0.05 M 0.025 M 0.025 M 0.025 M -

X of Bu4NX NTf2 (0.05 M) NTf2 (0.075 M) BF4 (0.075 M) ClO4 (0.075 M) NTf2 (0.10 M)

ratio

yield (

ratio)

100:0 100:0 50:50 25:75 25:75 25:75 0:100

82% (16:84) 74% ( 7:93) 69% ( 6:94) 62% ( 5:95) 59% (12:88) 51% ( 3:97) 35% (28:72)

Myc-IV (16C:0, S)

Total synthesis of TMG-chitotriomycin prompted us to prepare other chitooligosaccharides derivatives; however, stereoselectivity of disaccharide synthesis had to be improved. To generate and accumulate glycosyl triflates excess amount of Bu4NOTf has been used as electrolyte and reservoir of triflate anion. This may cause a poor b-selectivity in the glycosylation of a thioglycoside building block with 2-deoxy-2-azido group (Table 8.1). Although careful optimization of electrolyte for electrochemical glycosylation indicated that lower concentration of Bu4NOTf improved b-selectivity, resistance became higher at lower concentration of electrolyte. Thus, we added second electrolyte such as Bu4NNTf2 and found that mixed electrolyte systems were appropriate for b-selective glycosylation [21]. Thus-obtained disaccharide was used as a disaccharide building block for preparation of the precursor of Myc-IV (16C:0, S) (Fig. 8.8). Automated electrochemical assembly of two building blocks afforded the precursor, which can be converted to Myc-IV (16C:0, S) after further transformation including introduction of lipid part and sulfate group [22].

8.3.3

Cyclic Oligosaccharides

Cyclic oligosaccharides of glucosamine have been reported as novel cyclic oligosaccharides [23]. They were prepared by a conventional chemical method; however, control of stereochemistry was not perfect even in the presence of

8 Electrochemical Synthesis of Oligosaccharides …

133

Fig. 8.8 Total synthesis of Myc-IV (16C:0, S) via automated electrochemical assembly

neighboring group. On the other hand, electrochemical glycosylation of linear oligoglucosamines afforded corresponding cyclic oligoglucosamines in stereoselective manner (Fig. 8.9) [24]. Other electrolytes such as Bu4NNTf2 also gave a cyclic oligoglucosamine as a mixture of stereoisomers. Therefore, choice of electrolyte is important to obtain the desired product in stereoselective manner.

8.3.4

Ionic Liquid Tag for Oligosaccharide Synthesis

Ionic liquid Tag (IL-Tag) has been used as a separation tag for organic synthesis including oligosaccharide synthesis; however, its application for electrochemical glycosylation had never been reported [25]. We envisioned that an IL-tag plays multiple roles such as separation tag, electrolyte, and stereo-controlling group [26]. Actually, a/b-selectivity of disaccharide changed drastically in presence and absence of the IL-tag and the substituted position of IL-tag also influenced selectivity of electrochemical glycosylation (Fig. 8.10). We also demonstrated synthesis of the potential precursor of TMG-chitotriomycin using IL-tag as an intramolecular electrolyte (Fig. 8.11) [27]. Synthesis of the precursor was achieved by one-pot electrochemical glycosylation and Fmoc deprotection sequence in the absence of external electrolyte [28].

134

Md. A. Rahman et al.

Fig. 8.9 Electrochemical synthesis of cyclic oligosaccharides

Fig. 8.10 Effect of ionic liquid tag on stereoselectivity of electrochemical glycosylation

8 Electrochemical Synthesis of Oligosaccharides …

135

Fig. 8.11 Electrochemical synthesis of the TMG-chitotriomycin precursor in the absence of external electrolyte

8.4

Conclusion

In this chapter, development of automated electrochemical assembly and its application to synthesis of oligoglucosamines were described. The method enables us to access biologically active oligosaccharide such as TMG-chitotriomycin and Myc-IV (16C:0 S) rapidly. Electrochemical glycosylation is useful to convert linear oligosaccharide to cyclic oligosaccharide and its combination with IL-tag strategy, which is a potential empowering tool of automated electrochemical assembly, has also been demonstrated. Scope of thioglycoside building block is not limited to glucosamines and electrochemical glycosylation is useful for polymerization of building blocks. Development of novel methods to functionalize the anomeric thioaryl group is also in progress in our laboratory.

References 1. Witczak ZJ (2008) Monosaccharides: occurrence, significance, and properties. In: Fraser-Reid BO, Tatsuta K, Thiem J (eds) Glycoscience, 2nd edn. Springer, Berlin, Heidelberg, New York, pp 815–840 2. Witczak ZJ (2008) Oligosaccharides: occurrence, significance, and properties. In: Fraser-Reid BO, Tatsuta K, Thiem J (eds) Glycoscience, 2nd edn. Springer, Berlin, Heidelberg, New York, pp 1143–1183 3. BeMiller JN (2008) Polysaccharides: occurrence, significance, and properties. In: Fraser-Reid BO, Tatsuta K, Thiem J (eds) Glycoscience, 2nd edn. Springer, Berlin, Heidelberg, New York, pp 1413–1435

136

Md. A. Rahman et al.

4. Liaqat F, Eltem R (2018) Chitooligosaccharides and their biological activities: a comprehensive review. Carbohydr Poly 184:243–259. https://doi.org/10.1016/j.carbpol.2017.12.067 5. Usuki H, Nitoda T, Ichikawa M, Yamaji N, Iwashita T, Komura H, Kanzaki H (2008) TMG-chitotriomycin, an enzyme inhibitor specific for insect and fungal b-N-acetylglucosaminidases, Produced by Actinomycete Streptomyces anulatus NBRC 13369. J Am Chem Soc 130:4146–4152. https://doi.org/10.1021/ja077641f 6. Yang Y, Li Y, Yu B (2009) Total synthesis and structural revision of TMG-chitotriomycin, a specific inhibitor of insect and fungal b-N-acetylglucosaminidases. J Am Chem Soc 131:12076–12077. https://doi.org/10.1021/ja9055245 7. Yang Y, Yu B (2014) Recent advances in the synthesis of chitooligosaccharides and congeners. Tetrahedron 70:1023–1046. https://doi.org/10.1016/j.tet.2013.11.064 8. Plante OJ, Palmacci ER, Seeberger PH (2001) Automated solid-phase synthesis of oligosaccharides. Science 291:1523–1527. https://doi.org/10.1126/science.1057324 9. Ganesh NV, Fujikawa K, Tan YH, Stine KJ, Demchenko AV (2012) HPLC-assisted automated oligosaccharide synthesis. Org Lett 14:3036–3039. https://doi.org/10.1021/ol301105y 10. Tang SL, Pohl NLB (2015) Automated solution-phase synthesis of b-1,4-Mannuronate and b-1,4-Mannan. Org Lett 17:2642–2645. https://doi.org/10.1021/acs.orglett.5b01013 11. Noyori R, Kurimoto I (1986) Electrochemical glycosylation method. J Org Chem 51:4320– 4322. https://doi.org/10.1021/jo00372a050 12. Amatore C, Jutand A, Mallet JM, Meyer G, Sinaÿ P (1990) Electrochemical glycosylation using phenyl S-glycosides. J Chem Soc Chem Commun 718–719. https://doi.org/10.1039/ C39900000718 13. Balavoine G, Gref A, Fischer JC, Lubineau A (1990) Anodic glycosylation from aryl thioglycosides. Tetrahedron Lett 31:5761–5764. https://doi.org/10.1016/s0040-4039(00) 97951-6 14. Nokami T, Shibuya A, Tsuyama H, Suga S, Bowers AA, Crich D, Yoshida J (2007) Electrochemical generation of glycosyl triflate pools. J Am Chem Soc 129:10922–10928. https://doi.org/10.1021/ja072440x 15. Nokami T, Shibuya A, Manabe S, Ito Y, Yoshida J (2009) a and b-glycosyl sulfonium ions; generation and reactivity. Chem Eur J 15:2252–2255. https://doi.org/10.1002/chem. 200802293 16. Nokami T, Nozaki Y, Saigusa Y, Shibuya A, Manabe S, Ito Y, Yoshida J (2011) Glycosyl sulfonium ions as storable intermediates for glycosylations. Org Lett 13:1544–1547. https:// doi.org/10.1021/ol200242u 17. Nokami T, Shibuya A, Saigusa Y, Manabe S, Ito Y, Yoshida J (2012) Electrochemical generation of 2,3-oxazolidinone glycosyl triflates as an intermediate for stereoselective glycosylation. Beilstein J Org Chem 8:456–460. https://doi.org/10.3762/bjoc.8.52 18. Nokami T, Hayashi R, Saigusa Y, Shimizu A, Liu CY, Mong KK, Yoshida J (2013) Automated solution-phase synthesis of oligosaccharides via iterative electrochemical assembly of thioglycosides. Org Lett 15:4520–4523. https://doi.org/10.1021/ol402034g 19. Nokami T, Isoda Y, Sasaki N, Takaiso A, Hayase S, Itoh T, Hayashi R, Shimizu A, Yoshida J (2015) Automated Electrochemical assembly of the protected potential TMG-chitotriomycin precursor based on rational optimization of the carbohydrate building block. Org Lett 17:1525–1528. https://doi.org/10.1021/acs.orglett.5b00406 20. Isoda Y, Sasaki N, Kitamura K, Takahashi S, Manmode S, Takeda-Okuda N, Tamura J, Nokami T, Itoh T (2017) Total synthesis of TMG-chitotriomycin based on automated electrochemical assembly of a disaccharide building block. Beilstein J Org Chem 13:919– 924. https://doi.org/10.3762/bjoc.13.93 21. Isoda Y, Kitamura K, Takahashi S, Nokami T, Itoh T (2019) Mixed electrolytes driven stereoselective electrochemical glycosylation. ChemElectroChem 6:4149–4152. https://doi. org/10.1002/celc.201900215 22. Yano K, Itoh T, Nokami T (2020) Total synthesis of Myc-IV(C16:0, S) via automated electrochemical assembly. Carbohydr Res 492:108018. https://doi.org/10.1016/j.carres.2020. 108018

8 Electrochemical Synthesis of Oligosaccharides …

137

23. Gening ML, Tsvetkov YE, Denis V. Titov DV, Gerbst AG, Yudina ON, Grachev AA, Shashkov AS, Vidal S, Imberty A, Saha T, Kand D, Talukdar P, Pier GB, Nifantiev NE (2013) Linear and cyclic oligo-b-(1 ! 6)-D-glucosamines: Synthesis, conformations, and applications for design of a vaccine and oligodentate Glycoconjugates. Pure Appl Chem 85:1879–1891. https://doi.org/10.1351/PAC-CON-12-09-06 24. Manmode S, Tanabe S, Yamamoto T, Sasaki N, Nokami T, Itoh T (2019) Electrochemical glycosylation as an enabling tool for the stereoselective synthesis of cyclic oligosaccharides. ChemistryOpen 8:869–872. https://doi.org/10.1002/open.201900185 25. He X, Chan TH (2006) Ionic-tag-assisted oligosaccharide synthesis. Synthesis 1645–1651. https://doi.org/10.1055/s-2006-926447 26. Nokami T, Sasaki N, Isoda Y, Itoh T (2016) Ionic liquid tag with multiple functions in electrochemical glycosylation. ChemElectroChem 3:2012–2016. https://doi.org/10.1002/celc. 201600311 27. Sasaki N, Nokami T, Itoh T (2017) Synthesis of TMG-chitotriomycin precursor based on electrolyte-free electrochemical glycosylation using an ionic liquid tag. Chem Lett 46:683– 685. https://doi.org/10.1246/cl.170126 28. Nokami T, Tsuyama H, Shibuya A, Nakatsutsumi T, Yoshida J (2008) Oligosaccharide synthesis based on a one-pot electrochemical glycosylation-fmoc deprotection sequence. Chem Lett 37:942–943. https://doi.org/10.1246/cl.2008.942

Chapter 9

Efficient Synthesis of Biologically Active Peptides Based on Micro-flow Amide Bond Formation Shinichiro Fuse

Abstract Peptide drugs have garnered much attention in recent years because they possess the merits of both protein drugs and small-molecule-based drugs. In particular, specialty peptides such as N-methylated peptides and cyclic peptides have become increasingly important as drug candidates. Developing an inexpensive process for peptide chain elongation, that would also be high-yielding and scalable, however, is a highly challenging task even under the most promising reported conditions. We have performed amidations using highly active, high-atom economy, and inexpensive coupling agents. These highly active agents accelerated both the desired and the side reactions. The undesired reactions were suppressed, however, by taking advantage of micro-flow technology that allows precise control of both the reaction time and the temperature. Here we introduce our originally developed amidations for use with biologically active peptides including highly racemizable peptides, cyclic peptides, and bulky N-methylated peptides.










Keywords Peptide N-methylated peptide Cyclic peptide Micro-flow Continuous-flow Racemization Epimerization Amidation




9.1 9.1.1










Introduction The Importance of Peptide Drugs and the Challenges Presented by Their Production

Peptides are traditionally not regarded as promising drug candidates due to poor oral availability, metabolic stability, and cell membrane permeability. However, the success of protein drugs has changed this situation [1, 2]. These protein drugs are not suitable for oral administration and they cannot penetrate a cellular membrane. S. Fuse (&) Department of Basic Medicinal Sciences, Graduate School of Pharmaceutical Sciences, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 K. Fukase and T. Doi (eds.), Middle Molecular Strategy, https://doi.org/10.1007/978-981-16-2458-2_9

139

140

S. Fuse

In addition, their production cost is very high, but the failure rate in drug development is lower than that of traditional small-molecular drugs. People believe peptide drugs possess the merits of both protein drugs (high success rate) and small-molecule-based drugs (low production cost, orally available, membrane-permeable). In recent years, specialty peptides such as N-methylated peptides and cyclic peptides have become increasingly important as drug candidates because they improve characteristics such as metabolic stability, target selectivity, and membrane permeability [3–7]. However, high-yielding, scalable, and low-cost production of these peptides remains a highly challenging task even under the most promising reported conditions.

9.1.2

Problems in Traditional Peptide Synthesis

German chemist Theodor Curtius reported the first chemical coupling of amino acids via mixed anhydrides in 1881 [8]. Numerous coupling approaches have been reported in the long history of peptide synthesis [9–18]. The most frequently used approaches for peptide synthesis over the past few decades have included active ester formation. This approach usually requires an excess amount of expensive and low atom economy coupling agents and additives [15, 19, 20]. Representative condensation agents and their molecular weights are listed in Fig. 9.1. The molecular weights of N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide (EDCI) hydrochloride, O-(7-azabenzotriazole-1-yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate (HATU), and (1-cyano-2-ethoxy-2-oxoethylideneaminooxy)dimethylamino(morpholino) carbenium hexafluorophosphate (COMU) are much higher than the average molecular weight of amino acids (ca. 110). The large amount of waste generated from low atom economy reagents complicates the purification of desired peptides. In addition, frequently used condensation agents containing 1-hydroxybenzotriazole such as HATU are essentially explosive [21, 22]. Therefore, their use on a large

Fig. 9.1 Structure of representative condensation agents and their molecular weights

9 Efficient Synthesis of Biologically Active Peptides Based …

141

scale requires special care. It is well known that DCC and DIC have detrimental health effects. In addition, anaphylaxis induced by HATU, HBTU and HCTU has been reported quite recently [23]. Although detrimental health effects from the use of COMU have not been reported, the cost is too high for use in peptide production [24].

9.1.3

Micro-Flow Synthesis

Since the late 1990s, micro-flow synthesis has garnered much attention due to advantages over conventional batch synthesis, as shown below [25–33]. Here, “micro-flow” is defined as having reaction channels equal to, or less than, an inner diameter of 1 mm. (1) Precise control of short reaction time (99%

ZnCl2 in THF/2-Me THF Batch System

F

Br Me

N CF3

(S)-TFPO Pd cat.

N

S O O

Flow-Batch System

F

F 3C Me

N CF3

N

S O O >99: 1 dr

(Het)Ar-X (X= Br or Cl) Pd catalyst

F

O

NaBH 4 EtOH

F 3C Me

N CF3

OH CH3

N

S O O

HSD-016 65% yield, >99: 1 dr

Scheme 19.14 Generation of a-trifluoromethyl oxiranyl zincate in flow system

342

H. Amii

I

4.0 mL/min

F

Li

(0.1 M in THF)

F

0.013 s M1

R1

MeLi (Br free) (0.20 M in Et 2O) (2.0 eq)

4.0 mL/min M2

82%

R2 OMe N Me

F

Ph

O Ph

O

7.85 s

4.0 mL/min

(0.30 M in THF) (3.0 eq)

I

Li

4.0 mL/min

F

I

(0.10 M in THF)

-40 C M1

LDA (0.42 M in THF) 1.0 mL/min (1.05 eq) R 3SnCl (0.30 M in THF) (3.0 eq)

O R2

SnR3

0.082 s R1

I

M2

4.0 mL/min

LDA (0.42 M in THF) (4.0 eq)

Ph

F

N

or Ph

Cbz

F

-40 C

Li

10 s R2

I M3

SnR3 F

-40 C F

0.0081 s R3 3.8 mL/min

SnPh3

O

M4

-40 C

R 2 88%

Ph

or

10 s R4 Cbz

6.0 mL/min

Ph

F N

Ph

SnBu3 Ph 38%

Scheme 19.15 Flash generation and reactions of a-fluoromethyllithiums

with microreactor performed in good yields within short reaction times by virtue of thermal stability of CF3Cu and control of mixing. Fluoroform (HCF3) is one of the greenhouse gases that is generated in huge quantities from polytetrafluoroethylene production. HCF3 is a menace substance to environment. Transformation of fluoroform to valuable fluorinated compounds is quite attractive [36]. Grushin and Urakawa demonstrated a continuous flow process for the preparation of a powerful trifluoromethylating reagent “ligandless CuCF3” from fluoroform, which is one of the best CF3 sources from the practical viewpoints such as availability, cost, and atom economy (Scheme 19.17) [37]. Fine-tuning of the residence time and flow rates for HCF3, the cuprating reagent (from CuCl and t-

19

Recent Advances in the Integrated Microflow … Back pressure regulator (BPR) Ar CF3 200 psi

EtOAc Ar

I

343

in NMP

CF3CO 2K in NMP CuI, pyridine

160 min

CF3

N

CF3

N O

91%

N CF3

BnN 91%

83%

Scheme 19.16 Trifluoromethyation with trifluoroacetate in a flow system

BuOK) in DMF, and the stabilizer (Et3N
3HF) allowed for the continuous production of ligandless CuCF3, which acted as a highly efficient agent to afford a series of trifluoromethylated compounds. The Ruppert–Prakash reagent (Me3SiCF3) is the most powerful and welldocumented reagent for trifluoromethylation. In 2016, Shibata et al. demonstrated a flow method for the trifluoromethylation of carbonyl compounds by the Ruppert– Prakash reagent (Scheme 19.18) [38]. The KOH/Celite-packed column was easy to set up and trifluoromethylation was performed under air. The reactions proceeded smoothly and were applicable to the synthesis of trifluoromethylated drugs such as efavirenz. They also devised a cinchona alkaloid ammonium phenoxides/ Celite-packed column and disclosed preliminary results of enantioselective trifluoromethylation in flow system. Mass flow controller (MFC) HCF 3 4.5 mL/min

[K(DMF)][Cu(Ot-Bu) 2]

Et 3 [K(DMF)n][(CF 3)Cu(Ot-Bu)]

1.5 mL/h

CuCF3

18 mL/h

ligandless CuCF3

O Ph

Br

Br

O

in air

CO 2H CF3

Ph

B(OH) 2

>90%

CF3

CF3

87% CO 2H 89%

Scheme 19.17 Continuous process for production of CuCF3 from fluoroform

99%

344

H. Amii O Ph

Ph

Me 3SiCF3 (2.0 eq)

KOH/Celite

rt

F 3C

OH

Ph

Ph

flow system

DMF

86%

O Cl NO 2

rt

Me 3SiCF3

F 3C

Cl

Catalyst/Celite

OR

flow system

41% yield, 36% ee OMe

OMe OH

OH N N

(R = H, Me 3Si)

NO 2

toluene/CH2Cl 2 (2/1)

N

O

N

H

H PhO

F 3C

Cl

PhO

O

N H

Catalyst

Efavirenz

Scheme 19.18 Flow trifluoromethylation of carbonyl compounds by Ruppert–Prakash reagent

By the use of flow reactors, 2H-azirines were formed from O-mesylation and base-promoted cyclization of oximes (Scheme 19.19). The continuous flow techniques involving in situ generation of 2H-azirines and the subsequent nucleophilic trifluoromethylation by Ruppert–Prakash reagent gave trifluoromethylated aziridines with excellent diastereoselectivity [39].

HO

N

N

N Py SiO2

40 C Me

Et 3N in MeCN

10 mL, 10 min

MsCl in MeCN

SiO2

N 50 C

BPR

HN

H

Me

100 psi

10 mL, 5 min

CF3

fluoride monolith column

10 mL, 50 C

83%

Scheme 19.19 Continuous trifluoromethylation

flow

synthesis

4−Py

Tol

Me 3SiCF3 in THF

of

2H-azirines

and

diastereoselective

19

Recent Advances in the Integrated Microflow …

345

Ley and co-workers invented a useful method for the utilization of fluoroform. The generation of trifluoromethyl anion from deprotonation of HCF3 and its complete consumption through trapping with electrophiles such as ketones, aldehydes, and chlorosilanes were accomplished by the use of a flow microreactor system (Scheme 19.20). The reactions were continuously monitored by portable FT-IR and by bench-top 19F NMR devices connected to the flow reactor setup. Employing in-line analytical tools allowed perfect controlling the trifluoromethylation by gaseous fluoroform [40]. A flow nucleophilic trifluoromethylation of carbonyl compounds and their derivatives using fluoroform was developed by Shibata in 2019 [41]. This method is also applicable to the transformation of N-sulfinylimines to trifluoromethyl amines with high degree of diastereoselectivity. Furthermore, they demonstrated the synthetic utility of the present protocol, for instance, the formal synthesis of the anti-HIV drug Efavirenz by flow microreactors (Scheme 19.21).

19.3.2 Radical Fluoroalkylations The application of continuous flow devices has become common and general in organic synthesis. There have been a lot of examples using flow microreactors with improvement of chemical conversions and selectivities compared to conventional batch systems. There are several beneficial effects using of flow microreactor systems; for instance, extremely fast mixing ability attributed to shortened diffusion path length, and highly efficient heat transfer ability based on large surface-area-to-volume ratio. In some cases, product selectively can be enhanced dramatically by the use of continuous flow microreactor devices. Daikin Industries,

3 bar THF KOt-Bu in THF

NMR 43 MHz

0.2 mL/min

BPR 75 psi

FT-IR

0.2 mL/min

OH Ph

CF3 88%

16 mL −20 C

O Ph H 0.3 M in DMF

HCF 3

0.2 mL/min HCF 3 (in THF) Et 3SiCl (in THF) KHMDS (in THF)

−40 C

Et 3SiCF3 99%

Scheme 19.20 Real-time spectroscopic analysis for quantitative consumption of fluoroform during nucleophilic trifluoromethylation

346

H. Amii 0.1 MPa, 25 mL/min flow reactor

HCF 3

0.2 mL/min CF3 O

KHMDS in toluene

N H

1.1 mL/min

N Me

S

Me

0.23 mL −10 C

O

S

t-Bu

52%, dr = 26:1

t-Bu 0.33 mL/min

0.3 M in toluene

Scheme 19.21 Stereoselective nucleophilic trifluoromethylation using fluoroform

Ltd. developed the flow microreactor synthesis of fluorinated epoxides [42]. In the first step (radical addition of perfluoroalkyl iodides to allyl alcohol), there is a technical issue concerning the violent exothermic reaction induced by decomposition of AIBN. Furthermore, in the second step (intramolecular nucleophilic substitution of b-iodoalcohols to afford epoxides), efficient mixing technique is required for biphasic aqueous-organic systems. To resolve all the troublesome problems, flow microreactor provided improved reaction controls over traditional batch reactors; high-yield synthesis of fluorinated epoxides was achieved (Scheme 19.22). Despite the remarkable progress in functionalization of alkenes, styrene and its derivatives remain a challenging class of substrates for radical trifluoromethylation with respect to serious side reactions such as polymerization. In 2016 Noël disclosed the selective photocatalytic trifluoromethylation of styrenes under mild conditions (Scheme 19.23) [43]. The use of fac-Ir(ppy)3, visible light and CF3I was applicable to radical trifluoromethylation and hydrotrifluoromethylation of styrenes. The adoption of photochemical continuous flow microreactors allowed to reduce the reaction time drastically and to increase the selectivity of the reactions.

I C 4F 9

OH

Step 1 1.0 mm φ

CF 3 (CF 2)3 I

Step 2

OH + AIBN

C 4F 9

1.0 mm φ

O

KOH aq.

Step 2

Step 1 In Flow Microreactor System

20 min, 97% yield

In Macrobatch System

8 h, 97% yield

17 min, 97% yield 2 h, 84% yield

Scheme 19.22 Flow microreactor synthesis of fluorinated epoxides

19

Recent Advances in the Integrated Microflow …

347

Mass flow controller Blue LED (3.12 W × 2)

MFC

CF3I

CF3 t-Bu

t-Bu

rt tR = 30 min

fac-Ir(ppy)3 CsOAc in MeCN/DMF

Batch: 92% yield (E/Z = 73/27) Flow: 97% yield (E/Z = 94/6)

Scheme 19.23 Photocatalytic trifluoromethylation of styrenes

A continuous flow procedure for the two-step synthesis of a-trifluoromethyl ketones was developed by Kappe (Scheme 19.24). In the first step, ketones were converted into the corresponding silyl enol ethers by reaction with TMSOTf and iPr2NEt. In the second step, in situ formed silyl enol ethers underwent a visible light-induced trifluoromethylation. For the net transformation from ketones to a-trifluoromethylated products, a flow reactor with transparent FEP tubing and a household compact fluorescent lamp was effective in the presence of Eosin Y as an inexpensive metal-free photoredox catalyst [44]. In 2017, Ryu et al. presented photoredox-catalyzed hydrodifluoroalkylation of alkenes (Scheme 19.25). The reactions proceeded smoothly upon treatment with Hantzsch ester as a hydrogen source under visible light irradiation [45]. Due to the high efficiency with regard to irradiation, continuous photo flow systems shortened reaction time significantly.

O R1

R2

Eosin Y TMSOTf in THF i-Pr 2NEt in THF

O

0.5 mL/min R

OTMS R2 1

Br

Br

O

Br

Br CO 2Na Eosin Y

2 mL, 2 min

0.5 mL/min

white CFL (30 W) O R2

R1

CF3

28 mL, 15 min

CF3SO2Cl in THF Products:

ONa

0.5 mL/min O CF3

O

O

O CF3

CF3 86%

82%

72%

Scheme 19.24 Continuous flow a-trifluoromethylation of ketones

S

CF3 63%

348

H. Amii white LED (5 W × 2) F F

F F

Br

+

CO 2Me

(0.09 M) (5 eq) Hantzsch ester Ru(bpy) 3Cl 2 6H2O (1 mol%) Et 3N, DMF EtO 2C Hantzsch ester: Me

CO 2Me 82% Flow Photoreactor depth: 1 mm, width: 2 mm, length: 3 m, total inner volume: 6 mL

CO 2Et

residence time 20 min

Me

N H

Scheme 19.25 Photoredox-catalyzed hydrodifluoroalkylation of alkenes

A visible light-induced S-perfluoroalkylation for cysteine derivatives using Ru (bpy)2+ 3 as photocatalyst and perfluoroalkyl iodides was developed by Noël and co-workers (Scheme 19.26) [46, 47]. With exposure to blue LED light, a variety of fluoroalkyl groups were introduced to protected cysteines using CnF2n+1-I and I-CF2COOEt as coupling partners in short reaction time (5 min in flow). Kappe devoted a continuous flow radical C–C bond formation under Fenton-type reaction conditions to aromatic trifluoromethylation (Scheme 19.27). The reactions were performed in the presence of hydrogen peroxide and DMSO as reagents with an Fe(II) catalyst [48, 49]. Employing an operationally simple and rapid flow protocol, electron-rich aromatic and heteroaromatic substrates reacted with electrophilic radicals such as CF3, C4F9, etc. to afford perfluoroalkyl (hetero) arenes [50].

O

Blue LED (3.2 W)

MeO

SH

HN R Ru(bpy) 3Cl 2 in CH3CN

O S

MeO rt

HN

Rf-I TMEDA in CH3CN Products:

S

MeO S

HN

R

O

O MeO

Rf

Boc

(CF 2) 3CF3

92%

CF 2CO 2Et

HN

Boc 81%

Boc

H N

SCF3

O

Bn

N H 85%

Scheme 19.26 Photocatalytic fluoroalkylation of thiols in continuous flow

CO 2Me

19

Recent Advances in the Integrated Microflow … Ar-H CF3I (1.6 eq) or C 4F 9-I FeSO 4 7H 2O (0.4 eq) H 2SO 4 (0.8 eq) in DMSO/MeCN (2:1)

349

Na 2S 2O3 (1 M) 4.75 mL/min

rt

5.0 mL/min Ar-CF3

1.8 mL, 22 s 30% H 2O 2 (1.6 eq) Products: CF3 N Ph 90%

0.25 μL/min OH CF3

or Ar-C4F 9

CF3

Cl

Me

N H

Cl NH 2

OH 52%

36%

C 4F 9 71%

Scheme 19.27 Continuous flow homolytic aromatic perfluoroalkylation

Cu(I)-catalyzed radical trifluoromethylation of coumarins in a continuous flow reactor was developed (Scheme 19.28). Showing wide substrate tolerance, the reactions proceeded smoothly for a wide range of coumarin substrates upon treatment with CF3SO2Na and TBHP [51]. For (hetero)aromatic fluoroalkylations, photoredox methodology is straightforward and powerful because of direct C–H transformation. Photocatalytic trifluoromethylation and perfluoroalkylation of aromatic compounds were developed in continuous microflow (Scheme 19.29). A diverse set of arenes, heteroarenes and benzofused heterocycles participated in continuous flow perfluoroalkylation to give the corresponding perfluoroalkyl (hetero)arenes within several minutes [52]. As a matter of course, trifluoromethylated and perfluoroalkylated heterocycles are important building blocks for the synthesis of numerous pharmaceutical products, agrochemicals and are widely applied in material sciences. To date, several splendid examples of photocatalytic trifluoromethylation and perfluoroalkylation have been published (Scheme 19.30) [53–58]. Wirth and co-workers invented the electrochemical microreactors possessing electrodes connected to the flow path for rapid and efficient electrochemical reactions [59]. Kolbe electrolysis of trifluoroacetic acid in the presence of electron-poor

Et 2N

O

O

CF 3SO2Na CuCl in DMSO

60 C

40 min t-BuOOH (TBHP) in DMSO

CF3

66 μL/min Et 2N

O 71%

Scheme 19.28 Continuous flow trifluoromethylation of coumarins with CF3 radical

O

350

H. Amii Blue LED

X

X

Ru(bpy) 3Cl 2 (0.5-1 mol%) TMEDA (2-3 eq) in CH 3CN

Rf

Rf-I in CH 3CN

Products:

Me N

H N

CF 3 95% Me

O CF 3

95% H N

S CF 3 65%

Me

CF 3

Me

CF3 F CF 3

99%

73%

Me

H N CF 2CO 2Et 95%

Me

Scheme 19.29 Trifluoromethylation and perfluoroalkylation of heterocycles by photoredox catalysis in continuous flow

X

R1 X

CF3

R2 O

No l (2014)

O

Stephenson (2015)

CF3I Eosin Y TMEDA MeCN, rt white LED

Ru(bpy) 3Cl 2 CF3CO-O-COCF3 (TFAA) pyridine N-oxide MeCN, 30-45 C blue LED

Fagnoni (2016)

CF3

X

R1 X

Y

Y CF3

R2 O

O

No l (2017) CF3SO2Na [Ir{dF(CF3)ppy} 2(dtbpy)]PF 6 (NH 4) 2S2O8 DMSO, 40 C UV (450 nm)

O CH 2Cl 2, rt hv (310 nm)

F 3CO 2S

N

SO2CF3

Scheme 19.30 Examples of trifluoromethylation of aromatics by photoredox catalysis in flow

alkenes was carried out under constant current at continuous flow (Scheme 19.31). The electrochemical trifluoromethylation of alkenes in acetonitrile/water as a solvent mixture resulted in dimerization of the radical trifluoromethylated intermediates [60].

19

Recent Advances in the Integrated Microflow …

351

Electrochemical Microreactor (ECMR) Galvanostat Pt anode + CO 2Me CF3CO 2H Et 3N in MeCN/H 2O

CF3CO 2H -e - -CO 2 -H+

-

Pt cathode 2.8 mA/cm 2 69 s

CF3 CO 2Me

R1 R2

F 3C

CF3 CO 2Me 52% (DL:meso = 5:3)

R1 F 3C

R2

Scheme 19.31 Trifluoromethylation of electron-deficient alkenes in an electrochemical microreactor

Undevided flow cell G (+) | 316L (-) O O S F

SH

PhOH

O O Ph S O

Cs 2CO3 81% KF (5 eq), pyridine (6 eq) CH3CN/ 1M HCl (1:1 v/v) 3.3 V, tr = 5 min

SuFEx reaction in flow

79%

Scheme 19.32 Synthesis of sulfonyl fluorides through electrochemical oxidative coupling of thiols and KF

Sulfonyl fluorides are valuable synthetic motifs which are currently of high interest due to the popularity of the sulfur (VI) fluoride exchange (SuFEx) as click chemistry concept. In 2020, Noël showed a flow electrochemical approach to synthesize sulfonyl fluorides through oxidative coupling of thiols with potassium fluoride (Scheme 19.32). Furthermore, a continuous flow process for the electrochemical synthesis of sulfonyl fluorides was directly supplied to the subsequent SuFEx reactions with alcohols [61].

19.3.3 Reactions of Difluorocarbene In 2016, Charette demonstrated the generation of difluorocarbene (:CF2) and its addition to alkenes and alkynes in flow microreactors [62]. The adoption of continuous flow technology led to the controlled generation of difluorocarbene from the reaction of Me3SiCF3 with a catalytic amount of NaI (Scheme 19.33). In situ generated difluorocarbene reacted effectively with a wide range of alkenes and alkynes to afford the corresponding difluorocyclopropanes and difluorocyclopropenes, respectively.

352

H. Amii R1

R3

R2

R4 or

R1

F F R1 R2

BPR R2

8 bar

1.0 mL/min 120 C, 10 min

Me 3SiCF3 NaI in THF

R3 R4 F F or R1

R2

Products: F F

F F

F F

F F

CO 2i-Bu Ph 98%

76%

F F

SiMe3

Me

Ph

Me

CO 2Me

Pr

54%

58%

99%

Scheme 19.33 Difluorocarbene addition to alkenes and alkynes in continuous flow

A scalable continuous flow procedure for the production of eflornithine, an important active pharmaceutical ingredient (API), was developed by Kappe and co-workers (Scheme 19.34). Eflornithine is a widely used drug for the treatment of sleeping sickness and hirsutism, and it is on the World Health Organization’s list of essential medicines. With precious control of flow operation, the protected amino acid was subjected to difluoromethylation involving the generation of difluorocarbene from fluoroform and the follow-up hydrolysis to afford the target eflornithine in 86% isolated yield (over two steps) [63–65]. In 2020, Fu and Jamison presented the utilization of chlorodifluoromethane (ClCF2H, as a difluorocarbene source) in flow deuteriodifluoromethylation and gem-difluoroalkenylation of aldehydes (Scheme 19.35). Mechanistic studies revealed that the formation of difluorinated oxaphosphetane intermediate. Alkaline hydrolysis of the oxaphosphetane upon treatment with D2O gave a-deuteriodifluoromethylated benzyl alcohols. On the contrary, Wittig reaction (retro [2+2] cycloaddition of difluoro oxaphosphetane under thermal conditions) afforded the gem-difluoroalkenes [66].

Ar

Cl N Ar Cl

HF 2C (Ar = 4-ClC6H 4)

N N

N

CO 2Me 0.8 mL/min

in MeTHF

−30 C BPR

4 mL 1.2 mL/min

HCF 3

NH 2

0 C 1 mL

LiHMDS in THF

CO 2Me

MFC

25 C 14 mL

160 C 40 mL 1.0 mL/min

12 bar

H 2N HF 2C

CO 2H

99% conv. DL -(±)-difluoromethylornithine (DFMO, eflornithine)

conc. HCl

Scheme 19.34 Scalable continuous flow process for the synthesis of eflornithine using fluoroform as difluoromethyl source

19

Recent Advances in the Integrated Microflow …

353 Alkaline hydrolysis D 2O

ClCF 2H (g)

MFC O PPh 3

LiOt-Bu in THF O Ar

H PPh 3 in THF

0 C

Ar O

Ar = O

F

F

rt

Ar

OH CF 2D 72%

heat F

Ar 100 C Wittig pathway

F 70%

Scheme 19.35 Deuteriodifluoromethylation and gem-difluoroalkenylation of aldehydes using ClCF2H in continuous flow

19.4

Conclusion

All the examples using flow microreactor devices presented in this chapter give an impression on the possibilities of incorporating fluorine and fluoroalkyl groups in various forms. Even now, organofluorine compound have been gaining a significant importance in wide field of science and technology, and there have been the associated an increasing need for new and more efficient and versatile synthetic methods. As highlighted here, the outstanding achievements have been made vastly in the synthesis of fluorine-containing molecules utilizing flow microreactors. Flow chemistry will continue to make breakthroughs of ingenious synthetic strategies of fluorinated materials. Acknowledgements The financial support of the Ministry of Education, Culture, Sports, Science and Technology of Japan (Grant-in-Aid for Scientific Research on Innovative Areas “Middle Molecular Strategy”) is gratefully acknowledged.

References 1. Gouverneur V, Müller K (eds) (2012) Fluorine in pharmaceutical and medicinal chemistry: from biophysical aspects to clinical applications. World Scientific Publishing Company, London 2. Kirsch P (ed) (2013) Modern fluoroorganic chemistry: synthesis, reactivity, applications, 2nd edn. Wiley-VCH, Weinheim 3. Haufe G, Leroux F (eds) (2018) Fluorine in life sciences: pharmaceuticals, medicinal diagnostics, and agrochemicals, 1st edn. Academic Press, London 4. Inoue M, Sumii Y, Shibata N (2020) Contribution of organofluorine compounds to pharmaceuticals. ACS Omega 5:10633–10640. https://doi.org/10.1021/acsomega.0c00830 5. Yoshida J (2008) Flash chemistry: fast organic synthesis in microsystems. Wiley, Chichester 6. Wirth T (2013) Microreactors in organic chemistry and catalysis. Wiley-VCH, Weinheim 7. Vaccaro L (ed) (2016) Sustainable flow chemistry: methods and applications. Wiley, Chichester

354

H. Amii

8. Amii H, Nagaki A, Yoshida J (2013) Flow microreactor synthesis in organo-fluorine chemistry. Beilstein J Org Chem 9:2793–2802. https://doi.org/10.3762/bjoc.9.314 9. Fülöp Z, Szemesi P, Bana P, Éles J, Greiner I (2020) Evolution of flow-oriented design strategies in the continuous preparation of pharmaceuticals. React Chem Eng 5:1527–1555. https://doi.org/10.1039/d0re00273a 10. Rehm TH (2016) Photochemical fluorination reactions—a promising research field for continuous-flow synthesis. Chem Eng Technol 39:66–80. https://doi.org/10.1002/ceat. 201500195 11. Chambers RD, Spink RCH (1999) Microreactors for elemental fluorine. Chem Commun 1999:883–884. https://doi.org/10.1039/A901473J 12. Jähnisch K, Baerns M, Hessel V, Ehrfeld W, Haverkamp V, Löwe H, Guber CWA (2000) Direct fluorination of toluene using elemental fluorine in gas/liquid microreactors. J Fluorine Chem 105:117–128. https://doi.org/10.1016/S0022-1139(00)00300-6 13. Löb P, Löwe H, Hessel V (2004) Direct fluorination of toluene using elemental fluorine in gas/liquid microreactors. J Fluorine Chem 125:1677–1694. https://doi.org/10.1016/j.jfluchem. 2004.09.006 14. Chambers RD, Fox MA, Sandford G, Trmcic J, Goeta A (2007) Elemental fluorine part 20. Direct fluorination of deactivated aromaticsystems using microreactor techniques. J Fluorine Chem 128:29–33. https://doi.org/10.1016/j.jfluchem.2006.09.010 15. Breen JR, Sandford G, Yufit DS, Howard JAK, Fray J, Patel B (2011) Continuous gas/liquid– liquid/liquid flow synthesis of 4-fluoropyrazole derivatives by selective direct fluorination. Beilstein J Org Chem 7:1048–1054. https://doi.org/10.3762/bjoc.7.120 16. Harsanyi A, Conte A, Pichon L, Rabion A, Grenier S, Sandford G (2017) One-step continuous flow synthesis of antifungal WHO essential medicine flucytosine using fluorine. Org Process Res Dev 21:273–276. https://doi.org/10.1021/acs.oprd.6b00420 17. Johansen MB, Lindhardt AT (2018) Nucleophilic fluorination facilitated by a CsF–CaF2 packed bed reactor in continuous flow. Chem Commun 54:825. https://doi.org/10.1039/ C7CC09035H 18. Gustafsson T, Gilmour R, Seeberger PH (2008) Fluorination reactions in microreactors. Chem Commun 2008:3022–3024. https://doi.org/10.1039/B803695K 19. Baumann M, Baxendale IR, Ley SV (2008) The use of diethylaminosulfur trifluoride (DAST) for fluorination in a continuous-flow microreactor. Synlett, 2111–2114. https://doi.org/10. 1055/s-2008-1078026 20. Baumann M, Baxendale IR, Martin LJ, Ley SV (2009) Development of fluorination methods using continuous-flow microreactors. Tetrahedron 65:6611–6625. https://doi.org/10.1016/j. tet.2009.05.083 21. Carter CF, Lange H, Ley SV, Baxendale IR, Wittkamp B, Goode JG, Gaunt NL (2010) ReactIR flow cell: a new analytical tool for continuous flow chemical processing. Org Process Res Dev 14:393–404. https://doi.org/10.1021/op900305v 22. Watson DA, Su M, Teverovskiy G, Zhang Y, Garcia-Fortanet J, Kinzel T, Buchwald SL (2009) Formation of ArF from LPdAr(F): catalytic conversion of aryl triflates to aryl fluorides. Science 325:1661–1664. https://doi.org/10.1126/science.1178239 23. Noël T, Maimone TJ, Buchwald SL (2011) Accelerating palladium-catalyzed C-F bond formation: use of a microflow packed—bed reactor. Angew Chem Int Ed 50:8900–8903. https://doi.org/10.1002/anie.201104652 24. Ikawa T, Masuda S, Akai S (2018) Microflow fluorinations of Benzynes: efficient synthesis of fluoroaromatic compounds. Chem Pharm Bull 66:1153–1164. https://doi.org/10.1248/cpb. c18-00578 25. Nagaki A, Uesugi Y, Kim H, Yoshida J (2013) Synthesis of functionalized aryl fluorides using organolithium reagents in flow microreactors. Chem Asian J 8:705–708. https://doi.org/ 10.1002/asia.201201191 26. Wang B, Wang Y, Jiang Y, Chu Y, Qi S, Ju W, Xu D (2018) Asymmetric fluorination of indanone-2-carboxylates using a polystyrene-supported diphenylamine-linked bis(oxazoline) complex. Org Biomol Chem 16:7702–7710. https://doi.org/10.1039/C8OB01943F

19

Recent Advances in the Integrated Microflow …

355

27. Cantilo D, Frutos O, Rincón JA, Mateos C, Kappe CO (2014) Light-induced benzylic fluorinations. J Org Chem 79:8486−8490. https://doi.org/10.1021/op500213j 28. Wakami H, Yoshida J (2005) Grignard exchange reaction using a microflow system: from bench to pilot plant. Org Process Res Dev 9:787–791. https://doi.org/10.1021/op0501500 29. Perez K, Picard B, Vuluga D, Burel F, Hreiz R, Falk L, Commenge JM, Nagaki A, Yoshida J, Chataigner I, Maddaluno J, Legros J (2020) Bromine−lithium exchange on a gemdibromoalkene, part 2: comparative performance of flow micromixers. Org Process Res Dev 24:787–791. https://doi.org/10.1021/acs.oprd.0c00203 30. Uneyama K, Katagiri T, Amii H (2008) a-Trifluoromethylated carbanion synthons. Acc Chem Res 41:817–829. https://doi.org/10.1021/ar7002573 31. Nagaki A, Tokuoka S, Yamada S, Tomida Y, Oshiro K, Amii H, Yoshida J (2011) Perfluoroalkylation in flow microreactors: generation of perfluoroalkyllithiums in the presence and absence of electrophiles. Org Biomol Chem 9:7559–7563. https://doi.org/10.1039/ c1ob06350b 32. Fujita T, Konno N, Watabe Y, Ichitsuka T, Nagaki A, Yoshida J, Ichikawa J (2018) Flash generation and borylation of 1-(trifluoromethyl)vinyllithium toward synthesis of a(trifluoromethyl)styrenes. J Fluorine Chem 207:72–76. https://doi.org/10.1016/j.jfluchem. 2018.01.004 33. Zhang H, Buchwald SL (2017) Palladium-catalyzed Negishi coupling of a-CF3 oxiranyl zincate: access to chiral CF3-substituted benzylic tertiary alcohols. J Am Chem Soc 139:11590–11594. https://doi.org/10.1021/jacs.7b06630 34. Colella M, Tota A, Takahashi Y, Higuma R, Ishikawa S, Degennaro L, Luisi R, Nagaki A (2020) Fluoro-substituted methyllithium chemistry: external quenching method us-ing flow microreactors. Angew Chem Int Ed 59:10924–10928. https://doi.org/10.1002/anie. 202003831 35. Chen M, Buchwald SL (2013) Rapid and efficient trifluoromethylation of aromatic and heteroaromatic compounds using potassium trifluoroacetate enabled by a flow system. Angew Chem Int Ed 52:11628–11631. https://doi.org/10.1002/anie.201306094 36. Zanardi A, Novikov MA, Martin E, Benet-Buchholz J, Grushin VV (2011) Direct cupration of fluoroform. J Am Chem Soc 133:20901–20913. https://doi.org/10.1021/ja2081026 37. Mazloomi Z, Bansode A, Benavente P, Lishchynskyi A, Urakawa A, Grushin VV (2014) Continuous process for production of CuCF3 via direct cupration of fluoroform. Org Process Res Dev 18:1020–1026. https://doi.org/10.1021/op500109v 38. Okusu S, Hirano K, Yasuda Y, Tokunaga E, Shibata N (2016) Flow trifluoromethylation of carbonyl compounds by Ruppert–Prakash reagent and its application for pharmaceuticals, efavirenz and HSD-016. RSC Adv 6:82716–82720. https://doi.org/10.1039/C6RA19790F 39. Baumann M, Baxendale IR (2016) Continuous-flow synthesis of 2H-azirines and their diastereoselective trans-formation to aziridines. Synlett 27:159–163. https://doi.org/10.1055/ s-0035-1560391 40. Musio B, Gala E, Ley SV (2018) Real-time spectroscopic analysis enabling quantitative and safe consumption of fluoroform during nucleophilic trifluoromethylation in flow. ACS Sustain Chem Eng 6:1489–1495. https://doi.org/10.1021/acssuschemeng.7b04012 41. Hirano K, Gondo S, Punna N, Tokunaga E, Shibata N (2019) Gas/liquid-phase micro-flow trifluoromethylation using fluoroform: trifluoromethylation of aldehydes, ketones, chalcones, and N-sulfinylimines. ChemistryOpen 8:406–410. https://doi.org/10.1002/open.201800286 42. Taguchi M, Nakaya H, Ichihara K, Senba Y, Hiraga Y (2009) Manufacture of fluorine-containing epoxides from fluoroalkyl iodides and unsaturated alcohols. Jpn Kokai Tokkyo Koho JP 2009-67687, April 2 43. Straathof NJW, Cramer SE, Hessel V, Noël T (2016) Practical photocatalytic trifluoromethylation and hydrotrifluoromethylation of styrenes in batch and flow. Angew Chem Int Ed 55:15549–15553. https://doi.org/10.1002/anie.201608297 44. Cantillo D, de Frutos O, Rincon JA, Mateos C, Kappe CO (2014) Continuous flow a-trifluoromethylation of ketones by metal-free visible light photoredox catalysis. Org Lett 16:896–899. https://doi.org/10.1021/ol403650y

356

H. Amii

45. Sumino S, Uno M, Fukuyama T, Ryu I, Matsuura M, Yamamoto A, Kishikawa Y (2017) Photoredox-catalyzed hydrodifluoroalkylation of alkenes using difluorohaloalkyl compounds and a Hantzsch Ester. J Org Chem 82:5469–5474. https://doi.org/10.1021/acs.joc.7b00609 46. Straathof NJW, Tegelbeckers BJP, Hessel V, Wang X, Noël T (2014) A mild and fast photocatalytic trifluoromethylation of thiols in batch and continuous-flow. Chem Sci 5:4768– 4773. https://doi.org/10.1039/C4SC01982B 47. Bottecchia C, Wei XJ, Kuijpers KPL, Hessel V, Noël T (2016) Visible light-induced trifluoromethylation and perfluoroalkylation of cysteine residues in batch and continuous flow. J Org Chem 81:7301–7307. https://doi.org/10.1021/acs.joc.6b01031 48. Bravo A, Bjørsvik HR, Fontana F, Liguori L, Mele A, Minisci F (1997) New methods of free-radical perfluoroalkylation of aromatics and alkenes. Absolute rate constants and partial rate factors for the homolytic aromatic substitution by n-perfluorobutyl radical. J Org Chem 62:7128–7136. https://doi.org/10.1021/jo970302s 49. Kino T, Nagase Y, Ohtsuka Y, Yamamoto K, Uraguchi D, Tokuhisa K, Yamakawa T (2010) Trifluoromethylation of various aromatic compounds by CF3I in the presence of Fe(II) compound, H2O2 and dimethylsulfoxide. J Fluorine Chem 131:98–105. https://doi.org/10. 1016/j.jfluchem.2009.09.007 50. Monteiro JL, Carneiro PF, Elsner P, Roberge DM, Wuts PGM, Kurjan KC, Gutmann B, Kappe CO (2017) Continuous flow homolytic aromatic substitution with electrophilic radicals: a fast and scalable protocol for trifluoromethylation. Chem Eur J 23:176–186. https:// doi.org/10.1002/chem.201604579 51. Zhang X, Huang P, Li Y, Duan C (2015) A mild and fast continuous-flow trifluoromethylation of coumarins with the CF3 radical derived from CF3SO2Na and TBHP. Org Biomol Chem 13:10917–10922. https://doi.org/10.1039/C5OB01516B 52. Straathof NJW, Gemoets HPL, Wang X, Schouten JC, Hessel V, Noël T (2014) Rapid trifluoromethylation and perfluoroalkylation of five-membered heterocycles by photoredox catalysis in continuous flow. ChemSusChem 7:1612–1617. https://doi.org/10.1002/cssc. 201301282 53. Straathof NJW, van Osch DJGP, Schouten A, Wang X, Schouten JC, Hessel V, Noël T (2014) Visible light photocatalytic metal-free perfluoroalkylation of heteroarenes in continuous flow. J Flow Chem 4:12–17. https://doi.org/10.1556/JFC-D-13-00032 54. Beatty JW, Douglas JJ, Miller R, McAtee RC, Cole KP, Stephenson CRJ (2016) Photochemical perfluoroalkylation with pyridine N-oxides: mechanistic insights and performance on a kilogram scale. Chemistry 1:456–472. https://doi.org/10.1016/j.chempr.2016.08. 002 55. Beatty JW, Douglas JJ, Cole KP, Stephenson CRJ (2015) A scalable and operationally simple radical trifluoromethylation. Nat Commun 6:7919. https://doi.org/10.1038/ncomms8919 56. Abdiaja I, Bottecchia C, Alcazar J, Noël T (2017) Visible-light-induced trifluoromethylation of highly functionalized arenes and heteroarenes in continuous flow. Synthesis 49:4978–4985. https://doi.org/10.1055/s-0036-1588527 57. Zhang X, Li Y, Hao X, Jin K, Zhang R, Duan C (2018) Recyclable alkylated Ru(bpy)2+ 3 complex as a visible-light photoredox catalyst for perfluoroalkylation. Tetrahedron 74:1742– 1748. https://doi.org/10.1016/j.tet.2018.01.040 58. Torti E, Protti S, Fagnoni M (2018) N-Aryltrifluoromethanesulfonimides as new trifluoromethylating agents for the (photo)catalyst-free functionalization of (hetero)aromatics. Chem Commun 54:4144–4147. https://doi.org/10.1039/C8CC00938D 59. Arai K, Watts K, Wirth T (2014) Difluoro- and trifluoromethylation of electron-deficient alkenes in an electrochemical microreactor. ChemistryOpen 3:23–28. https://doi.org/10.1002/ open.201300039 60. Uneyama K (1991) Electrochemical trifluoromethylation of olefins; product-selectivity and mechanistic aspects. Tetrahedron 47:555–562. https://doi.org/10.1016/S0040-4020(01) 87045-9 61. Cao Y, Adriaenssens B, Bartolomeu A, Laudadio G, de Oliveira KT, Noël T (2020) Accelerating sulfonyl fluoride synthesis through electrochemical oxidative coupling of thiols

19

62. 63.

64.

65.

66.

Recent Advances in the Integrated Microflow …

357

and potassium fluoride in flow. J Flow Chem 10:191–197. https://doi.org/10.1007/s41981019-00070-9 Rulliére P, Cyr P, Charette AB (2016) Difluorocarbene addition to alkenes and alkynes in continuous flow. Org Lett 18:1988–1991. https://doi.org/10.1021/acs.orglett.6b00573 Gutmann B, Hanselmann P, Bersier M, Roberge D, Kappe CO (2017) Continuous-flow difluoromethylation with chlorodifluoromethane under biphasic conditions. J Flow Chem 7:46–51. https://doi.org/10.1556/1846.2017.00005 Kockinger M, Ciaglia T, Bersier M, Hanselmann P, Gutmann B, Kappe CO (2018) Utilization of fluoroform for difluoromethylation in continuous flow: a concise synthesis of a-difluoromethyl-amino acids. Green Chem 20:108–112. https://doi.org/10.1039/ c7gc02913f Köckinger M, Hone CA, Gutmann B, Hanselmann P, Bersier M, Torvisco A, Kappe CO (2018) Scalable continuous flow process for the synthesis of eflornithine using fluoroform as difluoromethyl source. Org Process Res Dev 22:1553–1563. https://doi.org/10.1021/acs.oprd. 8b00318 Fu WC, Jamison TF (2020) Deuteriodifluoromethylation and gem-Difluoroalkenylation of aldehydes using ClCF2H in continuous flow. Angew Chem Int Ed 59:13885–13890. https:// doi.org/10.1002/anie.202004260