The Integrin Interactome: Methods and Protocols [1st ed.] 9781071609613, 9781071609620

Citation preview

Methods in Molecular Biology 2217

Miguel Vicente-Manzanares Editor

The Integrin Interactome Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, UK

For further volumes: http://www.springer.com/series/7651

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-bystep fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.

The Integrin Interactome Methods and Protocols

Edited by

Miguel Vicente-Manzanares Molecular Mechanisms Program, Centro de Investigación del Cáncer and Instituto de Biología Molecular y Celular del Cáncer, Consejo Superior de Investigaciones Cientóficas (CSIC)-University of Salamanca, Salamanca, Spain

Editor Miguel Vicente-Manzanares Molecular Mechanisms Program, Centro de Investigacio´n del Ca´ncer and Instituto de Biologı´a Molecular y Celular del Ca´ncer Consejo Superior de Investigaciones Ciento´ficas (CSIC)-University of Salamanca Salamanca, Spain

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-0961-3 ISBN 978-1-0716-0962-0 (eBook) https://doi.org/10.1007/978-1-0716-0962-0 © Springer Science+Business Media, LLC, part of Springer Nature 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface Since their formal identification in the 1980s, integrins have remained in the spotlight in numerous fields, from basic cell biology and development to immunology, as well as clinical and pharmacological research. The term integrin was originally used to define the beta1 chain of the first identified fibronectin receptor, the integrin alpha5 beta1 [1]. In 1987, the name “integrin” was extended to comprise a whole superfamily of heterodimeric receptors that mediate cell-cell and cell-matrix adhesive adhesion [2]. By naming these receptors “integrins,” the intention was to convey the fact that they connected extracellular signals with intracellular adaptors to provide an integrated cellular response. This was prescient, as further research showed that different integrins controlled a myriad of processes in animal organisms, from cell proliferation to differentiation, cell migration, and boundary definition during organogenesis. Integrins also mediate wound healing and other processes, including hemostasis and immune surveillance. Furthermore, pathologies ranging from cancer to autoimmune disease have an adhesive component that has attracted attention on integrins as therapeutic targets. Some of these therapeutic advances have changed dramatically the standard of care of patients, for example in coronary disease. Conversely, other efforts have failed, e.g., targeting integrins to treat solid tumors, as reviewed elsewhere [3]. Early research showed that integrins were localized at the interface between extracellular matrix fibers and intracellular microfilaments [4]. Research from different fields, including embryology, cell biology, hematology, and immunology, converged to demonstrate that different integrins mediated the interaction of cells with other cells as well as the proteins that form the extracellular matrix. The ability of specific integrins to mediate platelet adhesion to fibrinogen and leukocyte adhesion to endothelial cells during extravasation indicated that some types of integrins were initially inactive to prevent inappropriate platelet, or leukocyte, adhesion. Conversely, these integrins became activated in response to specific cues (wounding, infection) of the cellular microenvironment. This line of thought led to the discovery of the ability of some integrins to undergo conformational changes that promote their ability to interact with their ligands. As these observations coincided in time with the zenith of the characterization of protein phosphorylation as a post-translational modification [5], it was postulated that binding of specific molecules to the cytoplasmic tails of integrins upon phosphorylation or other post-translational modification would promote their conformational extension. The events necessary to promote these associations would be triggered by the activation of other types of receptors, which would inform the cell of the need to activate its integrins. A clear example of this effect is the ability of pro-inflammatory cytokines to activate leukocyte integrins to trigger leukocyte adhesion in the inflammatory region. However, additional experimentation showed that integrin activation also required the application of mechanical force through a mechanosensitive relay made of a relatively small number of stretchable cytoplasmic adaptors, most notably talin [6]. As the complexity of integrin activation started growing, it became clear that integrins were at the epicenter of many signaling routes that would trigger their activation (inside-out signaling), but they also delivered signals that directly promoted proliferation, differentiation, etc. (outside-in signaling). Further complexity is added by the fact that some signaling mediators participate in both types of signaling [7].

v

vi

Preface

The present volume focuses on cutting-edge methods that are currently used to study integrin dynamics and their interaction with molecular partners at the plasma membrane as well as cytoplasmic adaptors. The methods described herein represent the culmination of over 50 years of active research in the field. But how did it start?

Early Days: Purifying and Identifying Adhesive Receptors and Their Ligands The beginning of modern integrin research required two major technological advances: the development of monoclonal antibodies and cDNA cloning. Very early experimentation using sponges and other marine organisms had shown that most organisms were comprised of individual cells that aggregated in a species-specific fashion to form multicellular systems [8]. This indicated that cell-cell adhesion was a key for tissue integrity and postulated that more complicated organs in higher organisms would abide by similar principles. Later, it was shown that leukocytes aggregated when activated with phytohemagglutinins [9], suggesting that activation triggered the expression or function of membrane proteins that mediated this effect. Interest in the role of adhesion in cancer biology stemmed from experiments that showed that, during the in vitro culture of tumor suspensions, the most aggressive cancer cells lost anchorage dependence, that is, they no longer needed substrate attachment to grow [10]. They also lost contact inhibition of growth and locomotion, growing in threedimensional foci and trampling on each other as they migrated, whereas less transformed cells did not display these behaviors [11]. Together, these experiments showed that adhesion was a hallmark of cellular “normalcy,” explained why cancer cells could invade neighboring tissues, and suggested that leukocytes and platelets had tunable receptors to mediate their physiological function only when and where required. The advent of monoclonal antibodies promoted a quantum leap in the field. Researchers could then visualize adhesive plaques in some cell types (but not others) that acted as connections between extracellular fibers and intracellular microfilaments [4]. This suggested that adhesive receptors would be concentrated in these regions. The theoretical ability of those (at the time) yet-to-be-identified receptors to interact with matrix components was employed to develop affinity approaches aimed at identifying the receptors. Progress was slow due to the low affinity of these interactions and the promiscuity of the different receptors. However, advances were made regarding the identification of ligands for the different types of integrins. Among others, these efforts identified the RGD motif as a key interactive sequence for many integrin types, as reviewed elsewhere [12]. Conversely, the use of monoclonal antibodies proved more fruitful. Hybridoma generation against plasma membrane fractions yielded specific antibodies that enabled the identification of proteins that mediated the interaction of platelets with fibrinogen [13], of embryonic muscle precursors with fibronectin [14], and of cytotoxic T cells with their targets [15]. These proteins had variable molecular weights, due to the heterodimeric nature of integrin chains of different classes and glycosylation. Also, immunization with proteins isolated from adhesion plaques and sequencing of co-immunoprecipitated proteins provided an early picture of the molecular interactome of these receptors [16–18]. These efforts peaked in 1986 with the cloning of the beta1 integrin [1] and the identification of talin as a major cytoplasmic interactor of beta1 integrins [19]. One year later, Richard Hynes coordinated a Gordon Research conference focused on adhesion, in which the term “integrin” was applied to various receptors identified in different cell lineages and using diverse strategies. From then on, a frantic race began to define the different families and subtypes of integrins, how they became activated, and what

Preface

vii

type of signals they would elicit. Integrins were found to control proliferation and survival, determine the shape of the cell through their control of cytoskeletal dynamics, and enable tissue migration and leukocyte extravasation, as reviewed in [20]. Together, this progress immediately poised integrins as key therapeutic targets to treat different types of diseases.

The Rise and Fall of Integrins as Therapeutic Targets The central role of integrins in leukocyte extravasation [21], together with the demonstration that integrin alterations underlie some immunodeficiencies [22], suggested that integrins could be involved in the abnormal migration of leukocytes in autoimmune disease. The emergence of antibody humanization techniques enabled the appearance of a therapeutic antibody (natalizumab) that targeted a leukocyte integrin (VLA-4) [23]. This antibody showed early promise to treat multiple sclerosis (MS). The FDA fast-tracked its approval for human use, but several deaths post-approval led to its withdrawal. Indeed, natalizumab completely abrogated leukocyte migration into the CNS. In fact, preventing leukocyte targeting to the CNS gave free rein to neuropathic viruses that caused progressive multifocal leukoencephalopathy (PML) [24]. After this caveat was determined, natalizumab was reapproved to treat MS patients tested for the presence of neuropathic viruses [25]. However, this risk has prevented the widespread use of natalizumab as a primary therapy to treat MS, but the biological effect (the fact that it prevents leukocyte migration to the CNS) was undeniable. Integrins were also considered as potential therapeutic targets to treat cancer, particularly to target angiogenesis and metastasis. However, these efforts failed due to the inability of anti-integrin antibodies to prevent either process. Endothelial cells can use many different integrins to migrate, and it is likely that if one is targeted, alternative integrins take over. This would explain the failure of cilengitide (a cyclic RGD-containing analogue) in phase II treatment of glioblastoma [26]. Furthermore, endothelial cells can adopt migratory modes that rely less on integrins [27]. On the other hand, metastatic cells can even migrate through tissues in the absence of integrins, displaying a similar behavior to that of leukocytes [28]. Finally, different integrin antagonists are used to treat coronary disease by targeting platelet integrins. In general, targeting integrins has proven a feasible strategy to control biological processes in which activatable integrins are required (platelets and leukocytes), but it is not in processes in which multiple integrins can take over the adhesive function, or they play nonessential roles.

The Integrin Adhesome, Optical Super-Resolution, and Quantitative Microscopy Up to the late 2000s, the integrin interactome had been partially defined through immunoprecipitation and localization experiments with antibodies. Several studies defined the integrin adhesome using mass spectrometry [29–31] and retrospective analysis of the literature and bioinformatics analysis [32, 33]. The adhesome is dependent on the integrin and cell type, and this type of work continues today [34]. Two chapters of this book describe these types of approaches. These early unbiased studies redefined the integrin interactome into functional categories, clarifying the connections between different families of

viii

Preface

molecules, including kinases/phosphatases, cytoskeletal adaptors, Rho GTPases and their regulators, and other proteins. Importantly, information regarding the stoichiometry of these interactions still depended on protein-protein interaction assays, which was a relatively unreliable approach due to the potential complexity of the integrin interactome. It is worth noting that this approach had elegantly determined the stoichiometry of actin-binding proteins, as reviewed elsewhere [35]. Stoichiometry began to be clarified when super-resolution microscopy allowed the direct visualization of molecular complexes within adhesive structures. A landmark study from the Waterman group used iPALM (a single molecule localization microscopy technique) to determine the nanoscale structure of integrin-based adhesions [36]. Their study showed that adhesions displayed a hierarchical distribution, in which focal adhesion proteins with different degrees of engagement with actin appeared in different strata within the adhesion. This work provided visual proof of previous data, some from the Waterman group itself [37, 38], in which the correlation of the fluctuations of different adhesive components with actin had revealed a different degree of engagement to the integrins on one side and actin on the other side. Together, these studies demonstrated that adhesion dynamics largely depend on a feedback loop that includes adhesive components and microfilaments.

Force-Dependent Integrin Activation Among many studies that sought to determine the regulation of integrin activation, a 1996 study showed that focal adhesions depended not only on the actin cytoskeleton but also on the normal function of a molecular motor associated with microfilaments, myosin II [39]. Earlier, focal adhesions had been shown to increase in size and number when a small GTPase, RhoA, was activated [40]. Indeed, later work showed that myosin II activation was dependent on the RhoA/ROCK axis [41, 42]. Larger focal adhesions meant more active integrins; hence it was postulated that myosin II-driven contractility and thus mechanical forces could mediate integrin activation. Given the bidirectional nature of integrin signals, it also posed the question whether integrins could behave as mechanical sensors [43]. As a consequence, the search of the mechanical mechanisms involved in the conformational extension of integrin receptors resulted in the widespread application of biophysical techniques to the adhesion field. Diverse studies have addressed different aspects of this relationship, from studies addressing the effect of ligand immobilization as a generator of mechanical traction [44] to substrate compliance [45, 46]. Others have used optical tweezers [47, 48] or atomic force microscopy [49, 50] to probe the role of mechanics in integrin activation. Also, FRET probes have been used to determine the mechanical stretching of specific molecules involved in adhesion, for example vinculin [51]. At any rate, the introduction of biophysical techniques to measure integrin activation added an extra layer of complexity to the study of the activation of these proteins. However, the highly quantitative nature of most of these techniques has also pushed the field towards a finer understanding of these mechanisms.

Preface

ix

Current Outlook and Future Perspectives Integrin research remains an active and vital field of research, as many questions remain in the field. The content of this book includes chapters on the evolution of some of the techniques described in the introduction, as well as novel methods that provide a more refined perspective on the function and interaction of integrins. The protocols included here include experimental approaches to quantify focal adhesion parameters, integrin activation, and the lateral interaction of integrins with transmembrane binding partners. Other chapters are devoted to an in-depth description of biophysical studies on integrin activation. State-ofthe-art biochemistry, proteomics, and bioinformatics analysis are also discussed in detail. Finally, several chapters are devoted to methods and techniques to study integrin traffic, the presence and role of integrins in exosomes, and sophisticated cellular methods to interrogate platelet integrin function without the caveats of working directly with platelets. This book, authored by a new generation of experts in the field, represents an excellent collection of methods and protocols that brings the know-how of its authors to the forefront. We have written it with the hope that it will provide a solid foundation for yet another generation of scientists of different disciplines as they interrogate the role of integrins in different biological processes.

References 1. Tamkun JW, DeSimone DW, Fonda D, Patel RS, Buck C, Horwitz AF, Hynes RO (1986) Structure of integrin, a glycoprotein involved in the transmembrane linkage between fibronectin and actin. Cell 46 (2):271–282 2. Hynes RO (1987) Integrins: a family of cell surface receptors. Cell 48(4):549–554. https://doi.org/ 10.1016/0092-8674(87)90233-9 3. Vicente-Manzanares M, Sanchez-Madrid F (2018) Targeting the integrin interactome in human disease. Curr Opin Cell Biol 55:17–23. https://doi.org/10.1016/j.ceb.2018.05.010 4. Lazarides E, Burridge K (1975) Alpha-actinin: immunofluorescent localization of a muscle structural protein in nonmuscle cells. Cell 6(3):289–298 5. Hunter T (1987) A thousand and one protein kinases. Cell 50(6):823–829. https://doi.org/10. 1016/0092-8674(87)90509-5 6. del Rio A, Perez-Jimenez R, Liu R, Roca-Cusachs P, Fernandez JM, Sheetz MP (2009) Stretching single talin rod molecules activates vinculin binding. Science 323(5914):638–641. https://doi.org/ 10.1126/science.1162912 7. Shattil SJ, Kim C, Ginsberg MH (2010) The final steps of integrin activation: the end game. Nat Rev Mol Cell Biol 11(4):288–300. https://doi.org/10.1038/nrm2871 8. Wilson HV (1907) On some phenomena of coalescence and regeneration in sponges. J Exp Zool 5:245–258 9. Li JG, Osgood EE (1949) A method for the rapid separation of leukocytes and nucleated erythrocytes from blood of marrow with a phytohemagglutinin from red beans (Phaseolus vulgaris). Blood 4 (5):670–675 10. Stoker M, O’Neill C, Berryman S, Waxman V (1968) Anchorage and growth regulation in normal and virus-transformed cells. Int J Cancer 3(5):683–693. https://doi.org/10.1002/ijc.2910030517 11. Abercrombie M (1979) Contact inhibition and malignancy. Nature 281(5729):259–262. https://doi. org/10.1038/281259a0

x

Preface

12. Ruoslahti E, Pierschbacher MD (1987) New perspectives in cell adhesion: RGD and integrins. Science 238(4826):491–497. https://doi.org/10.1126/science.2821619 13. Jennings LK, Phillips DR (1982) Purification of glycoproteins IIb and III from human platelet plasma membranes and characterization of a calcium-dependent glycoprotein IIb-III complex. J Biol Chem 257(17):10458–10466 14. Neff NT, Lowrey C, Decker C, Tovar A, Damsky C, Buck C, Horwitz AF (1982) A monoclonal antibody detaches embryonic skeletal muscle from extracellular matrices. J Cell Biol 95(2 Pt 1):654–666. https://doi.org/10.1083/jcb.95.2.654 15. Sanchez-Madrid F, Krensky AM, Ware CF, Robbins E, Strominger JL, Burakoff SJ, Springer TA (1982) Three distinct antigens associated with human T-lymphocyte-mediated cytolysis: LFA-1, LFA-2, and LFA-3. Proc Natl Acad Sci USA 79(23):7489–7493. https://doi.org/10.1073/pnas. 79.23.7489 16. Turner CE, Glenney JR Jr, Burridge K (1990) Paxillin: a new vinculin-binding protein present in focal adhesions. J Cell Biol 111(3):1059–1068. https://doi.org/10.1083/jcb.111.3.1059 17. Geiger B, Tokuyasu KT, Dutton AH, Singer SJ (1980) Vinculin, an intracellular protein localized at specialized sites where microfilament bundles terminate at cell membranes. Proc Natl Acad Sci USA 77 (7):4127–4131. https://doi.org/10.1073/pnas.77.7.4127 18. Burridge K, Connell L (1983) Talin: a cytoskeletal component concentrated in adhesion plaques and other sites of actin-membrane interaction. Cell Motil 3(5-6):405–417. https://doi.org/10.1002/cm. 970030509 19. Horwitz A, Duggan K, Buck C, Beckerle MC, Burridge K (1986) Interaction of plasma membrane fibronectin receptor with talin—a transmembrane linkage. Nature 320(6062):531–533. https://doi. org/10.1038/320531a0 20. Hynes RO (2002) Integrins: bidirectional, allosteric signaling machines. Cell 110(6):673–687 21. Springer TA (1994) Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 76(2):301–314. https://doi.org/10.1016/0092-8674(94)90337-9 22. Arnaout MA (1990) Leukocyte adhesion molecules deficiency: its structural basis, pathophysiology and implications for modulating the inflammatory response. Immunol Rev 114:145–180. https://doi. org/10.1111/j.1600-065x.1990.tb00564.x 23. Miller DH, Khan OA, Sheremata WA, Blumhardt LD, Rice GP, Libonati MA, Willmer-Hulme AJ, Dalton CM, Miszkiel KA, O’Connor PW, International Natalizumab Multiple Sclerosis Trial G (2003) A controlled trial of natalizumab for relapsing multiple sclerosis. N Engl J Med 348(1):15–23. https:// doi.org/10.1056/NEJMoa020696 24. Langer-Gould A, Atlas SW, Green AJ, Bollen AW, Pelletier D (2005) Progressive multifocal leukoencephalopathy in a patient treated with natalizumab. N Engl J Med 353(4):375–381. https://doi.org/ 10.1056/NEJMoa051847 25. Stuve O, Gold R, Chan A, Mix E, Zettl U, Kieseier BC (2008) alpha4-Integrin antagonism with natalizumab: effects and adverse effects. J Neurol 255 Suppl 6:58–65. https://doi.org/10.1007/ s00415-008-6011-0 26. Eisele G, Wick A, Eisele AC, Clement PM, Tonn J, Tabatabai G, Ochsenbein A, Schlegel U, Neyns B, Krex D, Simon M, Nikkhah G, Picard M, Stupp R, Wick W, Weller M (2014) Cilengitide treatment of newly diagnosed glioblastoma patients does not alter patterns of progression. J Neurooncol 117 (1):141–145. https://doi.org/10.1007/s11060-014-1365-x 27. Chilla A, Margheri F, Biagioni A, Del Rosso M, Fibbi G, Laurenzana A (2018) Mature and progenitor endothelial cells perform angiogenesis also under protease inhibition: the amoeboid angiogenesis. J Exp Clin Cancer Res 37(1):74. https://doi.org/10.1186/s13046-018-0742-2 28. Lammermann T, Bader BL, Monkley SJ, Worbs T, Wedlich-Soldner R, Hirsch K, Keller M, Forster R, Critchley DR, Fassler R, Sixt M (2008) Rapid leukocyte migration by integrin-independent flowing and squeezing. Nature 453(7191):51–55. https://doi.org/10.1038/nature06887

Preface

xi

29. Byron A, Humphries JD, Craig SE, Knight D, Humphries MJ (2012) Proteomic analysis of alpha4beta1 integrin adhesion complexes reveals alpha-subunit-dependent protein recruitment. Proteomics 12(13):2107–2114. https://doi.org/10.1002/pmic.201100487 30. Humphries JD, Byron A, Bass MD, Craig SE, Pinney JW, Knight D, Humphries MJ (2009) Proteomic analysis of integrin-associated complexes identifies RCC2 as a dual regulator of Rac1 and Arf6. Sci Signal 2(87):ra51. https://doi.org/10.1126/scisignal.2000396 31. Schiller HB, Friedel CC, Boulegue C, Fassler R (2011) Quantitative proteomics of the integrin adhesome show a myosin II-dependent recruitment of LIM domain proteins. EMBO Rep 12 (3):259–266. https://doi.org/10.1038/embor.2011.5 32. Zaidel-Bar R, Itzkovitz S, Ma’ayan A, Iyengar R, Geiger B (2007) Functional atlas of the integrin adhesome. Nat Cell Biol 9(8):858–867. https://doi.org/10.1038/ncb0807-858 33. Horton ER, Byron A, Askari JA, Ng DHJ, Millon-Fremillon A, Robertson J, Koper EJ, Paul NR, Warwood S, Knight D, Humphries JD, Humphries MJ (2015) Definition of a consensus integrin adhesome and its dynamics during adhesion complex assembly and disassembly. Nat Cell Biol 17 (12):1577–1587. https://doi.org/10.1038/ncb3257 34. Myllymaki SM, Kamarainen UR, Liu X, Cruz SP, Miettinen S, Vuorela M, Varjosalo M, Manninen A (2019) Assembly of the beta4-integrin interactome based on proximal biotinylation in the presence and absence of heterodimerization. Mol Cell Proteomics 18(2):277–293. https://doi.org/10.1074/ mcp.RA118.001095 35. Pollard TD (2016) Actin and actin-binding proteins. Cold Spring Harb Perspect Biol 8(8):a018226. https://doi.org/10.1101/cshperspect.a018226 36. Kanchanawong P, Shtengel G, Pasapera AM, Ramko EB, Davidson MW, Hess HF, Waterman CM (2010) Nanoscale architecture of integrin-based cell adhesions. Nature 468(7323):580–584. https:// doi.org/10.1038/nature09621 37. Hu K, Ji L, Applegate KT, Danuser G, Waterman-Storer CM (2007) Differential transmission of actin motion within focal adhesions. Science 315(5808):111–115 38. Brown CM, Hebert B, Kolin DL, Zareno J, Whitmore L, Horwitz AR, Wiseman PW (2006) Probing the integrin-actin linkage using high-resolution protein velocity mapping. J Cell Sci 119 (Pt 24):5204–5214. https://doi.org/10.1242/jcs.03321 39. Chrzanowska-Wodnicka M, Burridge K (1996) Rho-stimulated contractility drives the formation of stress fibers and focal adhesions. J Cell Biol 133(6):1403–1415 40. Ridley AJ, Hall A (1992) The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 70(3):389–399 41. Kimura K, Ito M, Amano M, Chihara K, Fukata Y, Nakafuku M, Yamamori B, Feng J, Nakano T, Okawa K, Iwamatsu A, Kaibuchi K (1996) Regulation of myosin phosphatase by Rho and Rho-associated kinase (Rho-kinase). Science 273(5272):245–248 42. Ishizaki T, Naito M, Fujisawa K, Maekawa M, Watanabe N, Saito Y, Narumiya S (1997) p160ROCK, a Rho-associated coiled-coil forming protein kinase, works downstream of Rho and induces focal adhesions. FEBS Lett 404(2-3):118–124. https://doi.org/10.1016/s0014-5793(97)00107-5 43. Bershadsky AD, Balaban NQ, Geiger B (2003) Adhesion-dependent cell mechanosensitivity. Annu Rev Cell Dev Biol 19:677–695 44. Schurpf T, Springer TA (2011) Regulation of integrin affinity on cell surfaces. EMBO J 30 (23):4712–4727. https://doi.org/10.1038/emboj.2011.333 45. Peng X, Huang J, Xiong C, Fang J (2012) Cell adhesion nucleation regulated by substrate stiffness: a Monte Carlo study. J Biomech 45(1):116–122. https://doi.org/10.1016/j.jbiomech.2011.09.013 46. Paszek MJ, Boettiger D, Weaver VM, Hammer DA (2009) Integrin clustering is driven by mechanical resistance from the glycocalyx and the substrate. PLoS Comput Biol 5(12):e1000604. https://doi. org/10.1371/journal.pcbi.1000604

xii

Preface

47. Schwingel M, Bastmeyer M (2013) Force mapping during the formation and maturation of cell adhesion sites with multiple optical tweezers. PLoS One 8(1):e54850. https://doi.org/10.1371/ journal.pone.0054850 48. Choquet D, Felsenfeld DP, Sheetz MP (1997) Extracellular matrix rigidity causes strengthening of integrin-cytoskeleton linkages. Cell 88(1):39–48. https://doi.org/10.1016/s0092-8674(00)81856-5 49. Li F, Redick SD, Erickson HP, Moy VT (2003) Force measurements of the alpha5beta1 integrinfibronectin interaction. Biophys J 84(2 Pt 1):1252–1262. https://doi.org/10.1016/S0006-3495( 03)74940-6 50. Franz CM, Muller DJ (2005) Analyzing focal adhesion structure by atomic force microscopy. J Cell Sci 118(Pt 22):5315–5323. https://doi.org/10.1242/jcs.02653 51. Grashoff C, Hoffman BD, Brenner MD, Zhou R, Parsons M, Yang MT, McLean MA, Sligar SG, Chen CS, Ha T, Schwartz MA (2010) Measuring mechanical tension across vinculin reveals regulation of focal adhesion dynamics. Nature 466(7303):263–266. https://doi.org/10.1038/nature09198

Acknowledgement I would like to thank Francisco Sa´nchez-Madrid for his trust and friendship over the years, his advice and critical reading of this preface. I would also like to thank Rick Horwitz for his mentorship, friendship and advice, and his terrific input to the field for the largest part of his distinguished career.

xiii

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PART I

INTEGRIN ACTIVATION AND FOCAL ADHESION MEASUREMENTS

1 Measurement of Integrin Activation and Conformational Changes on the Cell Surface by Soluble Ligand and Antibody Binding Assays . . . . . . . . . . Zhengli Wang and Jieqing Zhu 2 Quantification of Integrin Activation and Ligation in Adherent Cells . . . . . . . . . . Zaki Al-Yafeai and A. Wayne Orr 3 Multiparametric Analysis of Focal Adhesions in Bidimensional Substrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vanessa C. Talayero and Miguel Vicente-Manzanares 4 Focal Adhesion Isolation Assay Using ECM-Coated Magnetic Beads . . . . . . . . . . Wesley Sturgess and Vinay Swaminathan

PART II

3 17

27 39

PROXIMITY AND MICROSCOPY-BASED METHODS TO DETERMINE INTEGRIN INTERACTIONS

5 Functional Integrin Regulation Through Interactions with Tetraspanin CD9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ´ lvaro Torres-G
omez, Beatriz Carden ˜ es, Ester Dı´ez-Sainz, A ˜ as Esther M. Lafuente, and Carlos Caban 6 Proximity-Dependent Biotinylation (BioID) of Integrin Interaction Partners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Satu-Marja Myllym€ a ki, Xiaonan Liu, Markku Varjosalo, and Aki Manninen 7 Analyzing the Integrin Adhesome by In Situ Proximity Ligation Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Brian A. Perrino, Yeming Xie, and Cristina Alexandru

PART III

v xiii xvii

47

57

71

BIOCHEMICAL, PROTEOMICS AND COMPUTATIONAL METHODS TO DETERMINE INTEGRIN INTERACTIONS

8 Single-Protein Tracking to Study Protein Interactions During Integrin-Based Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 A. V. Radhakrishnan, Tianchi Chen, Jose Filipe Nunes Vicente, Thomas Orre´, Amine Mehidi, Olivier Rossier, and Gre´gory Giannone 9 Biochemical Characterization of the Integrin Interactome . . . . . . . . . . . . . . . . . . . 115 Rejina B. Khan, Lorena Varela, Alana R. Cowell, and Benjamin T. Goult

xv

xvi

10

Contents

Network Analysis of Integrin Adhesion Complexes . . . . . . . . . . . . . . . . . . . . . . . . . 149 Frederic Li Mow Chee and Adam Byron

PART IV 11

12

Surface Patterning for the Control of Receptor Clustering and Molecular Forces of Integrin-Mediated Adhesions . . . . . . . . . . . . . . . . . . . . . . 183 Federica Pennarola and Elisabetta Ada Cavalcanti-Adam Dynamics and Physics of Integrin Activation in Tumor Cells by Nano-Sized Extracellular Ligands and Electromagnetic Fields . . . . . . . . . . . . . 197 Alkiviadis-Constantinos Cefalas, Vassilios Gavriil, Angelo Ferraro, Zoe Kollia, and Evangelia Sarantopoulou

PART V 13

14 15

17

INTEGRIN ACTIVATION AND INTERACTIONS IN SPECIFIC SYSTEMS AND PROCESSES

Genetic Instruction of Megakaryocytes and Platelets Derived from Human Induced Pluripotent Stem Cells for Studies of Integrin Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237 Ana Kasirer-Friede and Sanford J. Shattil Quantitative Analysis of Integrin Trafficking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 Enoir Farage and Patrick T. Caswell Methods to Study Integrin Functions on Exosomes. . . . . . . . . . . . . . . . . . . . . . . . . 265 Eiji Kawamoto, Eun Jeong Park, and Motomu Shimaoka

PART VI 16

BIOPHYSICAL METHODS TO DETERMINE INTEGRIN ACTIVATION AND ITS CELLULAR AND MOLECULAR EFFECTS

INTEGRIN LIGANDS AND THE EXTRACELLULAR MATRIX

Functional Bioinformatics Analyses of the Matrisome and Integrin Adhesome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Edward Roy Horton Quantifying Polarized Extracellular Matrix Secretion in Cultured Endothelial Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 Fabiana Clapero, Dora Tortarolo, Donatella Valdembri, and Guido Serini

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

313

Contributors CRISTINA ALEXANDRU • Department of Physiology and Cell Biology, University of Nevada, Reno School of Medicine, Reno, NV, USA ZAKI AL-YAFEAI • Department of Molecular and Cellular Physiology, LSU Health Sciences Center Shreveport, Shreveport, LA, USA ADAM BYRON • Cancer Research UK Edinburgh Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK CARLOS CABAN˜AS • Department of Immunology, Ophthalmology and Otorhinolaryngology, School of Medicine, Universidad Complutense de Madrid, Madrid, Spain; Instituto de Investigaci
on Sanitaria i+12, Hospital 12 de Octubre, Madrid, Spain; Centro de Biologı´a Molecular Severo Ochoa (CSIC-UAM), Madrid, Spain BEATRIZ CARDEN˜ES • Centro de Biologı´a Molecular Severo Ochoa (CSIC-UAM), Madrid, Spain PATRICK T. CASWELL • Wellcome Trust Centre for Cell-Matrix Research, Division of Cell Matrix Biology and Regenerative Medicine, School of Biological Sciences, University of Manchester, Manchester, UK ELISABETTA ADA CAVALCANTI-ADAM • Department of Cellular Biophysics, Max Planck Institute for Medical Research, Heidelberg, Germany ALKIVIADIS-CONSTANTINOS CEFALAS • National Hellenic Research Foundation, Theoretical and Physical Chemistry Institute, Athens, Greece FREDERIC LI MOW CHEE • Cancer Research UK Edinburgh Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK TIANCHI CHEN • Universite´ de Bordeaux, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France FABIANA CLAPERO • Candiolo Cancer Institute, FPO—IRCCS, Candiolo, Italy; Department of Oncology, University of Torino School of Medicine, Torino, Italy ALANA R. COWELL • School of Biosciences, University of Kent, Kent, UK ESTER DI´EZ-SAINZ • Centro de Biologı´a Molecular Severo Ochoa (CSIC-UAM), Madrid, Spain ENOIR FARAGE • Wellcome Trust Centre for Cell-Matrix Research, Division of Cell Matrix Biology and Regenerative Medicine, School of Biological Sciences, University of Manchester, Manchester, UK ANGELO FERRARO • National Hellenic Research Foundation, Theoretical and Physical Chemistry Institute, Athens, Greece; The National Technical University of Athens, School of Electrical and Computer Engineering, Zografou, Greece VASSILIOS GAVRIIL • National Hellenic Research Foundation, Theoretical and Physical Chemistry Institute, Athens, Greece GRE´GORY GIANNONE • Universite´ de Bordeaux, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France BENJAMIN T. GOULT • School of Biosciences, University of Kent, Kent, UK EDWARD ROY HORTON • Biotech Research and Innovation Centre (BRIC), University of Copenhagen, Copenhagen, Denmark

xvii

xviii

Contributors

ANA KASIRER-FRIEDE • Hematology-Oncology Division, Department of Medicine, University of California, San Diego, La Jolla, CA, USA EIJI KAWAMOTO • Department of Molecular Pathobiology and Cell Adhesion Biology, Mie University Graduate School of Medicine, Tsu-city, Mie, Japan; Department of Emergency and Disaster Medicine, Mie University Graduate School of Medicine, Tsu-city, Mie, Japan REJINA B. KHAN • School of Biosciences, University of Kent, Kent, UK ZOE KOLLIA • National Hellenic Research Foundation, Theoretical and Physical Chemistry Institute, Athens, Greece ESTHER M. LAFUENTE • Department of Immunology, Ophthalmology and Otorhinolaryngology, School of Medicine, Universidad Complutense de Madrid, Madrid, Spain; Instituto de Investigaci
on Sanitaria i+12, Hospital 12 de Octubre, Madrid, Spain XIAONAN LIU • HiLIFE—Institute of Biotechnology, University of Helsinki, Helsinki, Finland AKI MANNINEN • Oulu Center for Cell-Matrix Research, Biocenter Oulu, Faculty of Biochemistry and Molecular Medicine, University of Oulu, Oulu, Finland AMINE MEHIDI • Universite´ de Bordeaux, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; Department of Biochemistry, University of Geneva, Geneva, Switzerland € SATU-MARJA MYLLYMAKI • Oulu Center for Cell-Matrix Research, Biocenter Oulu, Faculty of Biochemistry and Molecular Medicine, University of Oulu, Oulu, Finland; Cell and Tissue Dynamics Program, HiLIFE—Institute of Biotechnology, University of Helsinki, Helsinki, Finland JOSE FILIPE NUNES VICENTE • Universite´ de Bordeaux, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France A. WAYNE ORR • Department of Molecular and Cellular Physiology, LSU Health Sciences Center Shreveport, Shreveport, LA, USA; Department of Pathology, LSU Health Sciences Center Shreveport, Shreveport, LA, USA THOMAS ORRE´ • Universite´ de Bordeaux, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France EUN JEONG PARK • Department of Molecular Pathobiology and Cell Adhesion Biology, Mie University Graduate School of Medicine, Tsu-city, Mie, Japan FEDERICA PENNAROLA • Department of Cellular Biophysics, Max Planck Institute for Medical Research, Heidelberg, Germany; Department of Physics and Astronomy, Heidelberg University, Heidelberg, Germany BRIAN A. PERRINO • Department of Physiology and Cell Biology, University of Nevada, Reno School of Medicine, Reno, NV, USA A. V. RADHAKRISHNAN • Universite´ de Bordeaux, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France OLIVIER ROSSIER • Universite´ de Bordeaux, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France; CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, Bordeaux, France EVANGELIA SARANTOPOULOU • National Hellenic Research Foundation, Theoretical and Physical Chemistry Institute, Athens, Greece GUIDO SERINI • Candiolo Cancer Institute, FPO—IRCCS, Candiolo, Italy; Department of Oncology, University of Torino School of Medicine, Torino, Italy

Contributors

xix

SANFORD J. SHATTIL • Hematology-Oncology Division, Department of Medicine, University of California, San Diego, La Jolla, CA, USA MOTOMU SHIMAOKA • Department of Molecular Pathobiology and Cell Adhesion Biology, Mie University Graduate School of Medicine, Tsu-city, Mie, Japan WESLEY STURGESS • Wallenberg Centre for Molecular Medicine, Division of Oncology, Department of Clinical Sciences, Lund University, Lund, Sweden VINAY SWAMINATHAN • Wallenberg Centre for Molecular Medicine, Division of Oncology, Department of Clinical Sciences, Lund University, Lund, Sweden VANESSA C. TALAYERO • Molecular Mechanisms Program, Centro de Investigaci
on del Ca´ncer and Instituto de Biologı´a Molecular y Celular del Ca´ncer, Consejo Superior de Investigaciones Cientı´ficas (CSIC)—University of Salamanca, Salamanca, Spain ´ LVARO TORRES-GO´MEZ • Department of Immunology, Ophthalmology and A Otorhinolaryngology, School of Medicine, Universidad Complutense de Madrid, Madrid, Spain; Instituto de Investigaci
on Sanitaria i+12, Hospital 12 de Octubre, Madrid, Spain DORA TORTAROLO • Candiolo Cancer Institute, FPO—IRCCS, Candiolo, Italy; Department of Oncology, University of Torino School of Medicine, Torino, Italy DONATELLA VALDEMBRI • Candiolo Cancer Institute, FPO—IRCCS, Candiolo, Italy; Department of Oncology, University of Torino School of Medicine, Torino, Italy LORENA VARELA • School of Biosciences, University of Kent, Kent, UK MARKKU VARJOSALO • HiLIFE—Institute of Biotechnology, University of Helsinki, Helsinki, Finland MIGUEL VICENTE-MANZANARES • Molecular Mechanisms Program, Centro de Investigaci
on del Ca´ncer and Instituto de Biologı´a Molecular y Celular del Ca´ncer, Consejo Superior de Investigaciones Cientı´ficas (CSIC)—University of Salamanca, Salamanca, Spain ZHENGLI WANG • Blood Research Institute, Versiti, Milwaukee, WI, USA YEMING XIE • Department of Physiology and Cell Biology, University of Nevada, Reno School of Medicine, Reno, NV, USA JIEQING ZHU • Blood Research Institute, Versiti, Milwaukee, WI, USA; Department of Biochemistry, Medical College of Wisconsin, Milwaukee, WI, USA

Part I Integrin Activation and Focal Adhesion Measurements

Chapter 1 Measurement of Integrin Activation and Conformational Changes on the Cell Surface by Soluble Ligand and Antibody Binding Assays Zhengli Wang and Jieqing Zhu Abstract Soluble ligand and conformation-dependent antibody binding assay of integrins on the cell surface is an effective approach to evaluate the activation status of integrins in live cells. The ligands or antibodies are usually labeled with biotin or a fluorescent dye and incubated with integrin-expressing cells in suspension. The cell-bound ligands and antibodies are then detected by flow cytometry. Here we describe the detailed protocols of soluble ligand or antibody binding assay for αIIbβ3, αVβ3, α5β1, and αLβ2 integrins that are transiently or stably expressed in the model cell lines such as HEK293 or CHO-k1 cells. Key words Integrin, Activation, Ligand binding, Conformation-dependent antibodies, Flow cytometry

1

Introduction Integrin activation is associated with affinity regulation that is modulated by conformational changes [1]. The changes in affinity can be measured by the increased or decreased binding of soluble ligands, while the changes in conformation can be probed by the binding of conformation-dependent monoclonal antibodies (mAbs) or ligand-induced-binding-site (LIBS) mAbs [2, 3]. Both soluble ligand and mAb binding can be performed with integrins presented on the surface of live cells and then measured by flow cytometry. This assay is an effective tool to assess integrin activation that is influenced by mutations or by the cellular activators such as talin and kindlin in the cytosol. When the ligand and LIBS mAb binding are analyzed at the same time, a correlation between integrin activation and conformational changes can be established, which provides useful information about the activation regulation of integrins on the cell surface. The integrin soluble ligand-binding assay has been widely used by many groups on different types of

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021

3

4

Zhengli Wang and Jieqing Zhu

integrins. Here we describe the ligands, mAbs, and methods used for analyzing the activation and conformation of human αIIbβ3, αVβ3, α5β1, and αLβ2 integrins on the surface of live cells, which we have been using in our structural and functional analysis of these integrins [4–8]. The method may be applied to study the activation and conformation of other members of the integrin family.

2 2.1

Materials DNA Constructs

1. Integrin constructs: Human αIIb cDNA in pEF1/V5-HisA: Addgene Plasmid #27288; human β3 cDNA in pcDNA3.1/ Myc-His(+)A: Addgene Plasmid #27289; human αV cDNA in pEF1/V5-HisA: Addgene Plasmid #27290; human α5 cDNA in pcDNA3.1/Myc-His(+)A [8]; human α5 cDNA with C-terminal EGFP tag in pEGFP-N3: Addgene Plasmid #15238; human β1 cDNA in pEF1/V5-HisA [8]; human αL cDNA in pcDNA3.1/Hydro(
) [8]; human β2 cDNA in pcDNA3.1(+) [8]. 2. Ligands constructs: human fibronectin type III repeats 9-10 (Fn9-10) in pGEX-6p-1; human intercellular adhesion molecule 1 (ICAM-1) with C-terminal human IgG1 Fc tag in pIRES2-EGFP [4]. 3. Talin and kindlin constructs: mouse talin1 head domain (1-433) with N-terminal EGFP tag (EGFP-TH) in pEGFPC1 [9]; Kindlin-2 with N-terminal mCherry tag in pmCherryC1 (mCherry-K2) [7].

2.2

Integrin Ligands

See Note 1 for considerations of the number of binding sites of each ligand. 1. Two multivalent ligands can be used for αIIbβ3 integrin: (a) A mouse anti-human αIIbβ3 mAb named PAC-1 (BD Biosciences). PAC-1 is an IgM molecule that contains an Arg-Tyr-Asp (RYD) motif in its heavy chain CDR3 loop [10], which mimics the integrin-binding Arg-Gly-Asp (RGD) motif found in many integrin ligands (see Note 2). (b) Human fibrinogen (Plasminogen, von Willebrand factor, and fibronectin depleted. Enzyme Research Laboratories). 2. Two multivalent ligands can be used for αVβ3 integrin: (a) Human fibronectin (Sigma-Aldrich). (b) Human vitronectin (Sigma-Aldrich). 3. Two ligands can be used for α5β1 integrin: (a) The multivalent human fibronectin. (b) The monovalent fibronectin fragment FN9-10.

Integrin-Ligand Interaction on the Cell Surface

5

4. One multivalent ligand can be used for αLβ2 integrin: human ICAM-1 with C-terminal human IgG1 Fc tag (ICAM-1-Fc), which can be produced as described below, or purchased from Sino Biological. 2.3 Preparation of FN9-10 Fragment

1. E. coli strain BL21. 2. Glutathione Sepharose 4B (GE Healthcare). 3. HRV 3C protease (Sino Biological).

2.4 Preparation of ICAM-1-Fc Fragment

1. Polyethylenimine (PEI) (Polysciences). 2. Anti-ICAM-1 mAb (Sino Biological). 3. Peroxidase-conjugated mouse anti-human IgG1 Fc (Jackson ImmunoResearch). 4. Purified ICAM-1-Fc (Sino Biological).

2.5 Integrin Antibodies

For most of the antibodies listed below, unlabeled and labeled versions can be obtained from several vendors. Unlabeled antibodies can be conjugated with biotin or fluorescent dyes of choice as described in Subheading 3. 1. Anti-αIIbβ3 and αVβ3 antibodies: (a) Mouse anti-human β3 mAb AP3 is a nonfunctional (nonstimulation, noninhibitory) antibody that can be used for detecting the cell surface expression of both αIIbβ3 and αVβ3 integrins [11] (Kerafast). (b) Mouse anti-human αIIb mAb 10E5 is an αIIbβ3 complex– specific blocking antibody [12, 13]. (c) Mouse anti-αIIbβ3 mAb AP-2 is αIIbβ3 complex–specific blocking antibody [14] (Kerafast). (d) Mouse anti-human αVβ3 mAb LM609 is αVβ3 complex– specific blocking antibody [15] (Sigma-Aldrich). 2. There are several conformation-dependent (or LIBS) mAbs we have been using for probing the conformational changes of human αIIb and β3 subunits: (a) Mouse anti-αIIb mAb 370.3 [8] or 370.2 [16] (Kerafast) reports conformational activation of αIIb subunit. (b) Mouse anti-β3 mAbs AP-5 [17] (Kerafast), 319.4 [8] (Kerafast), and LIBS-1 [18] report the conformational extension of β3 subunit [4, 8]. 3. Anti-α5β1 antibodies: (a) MAR4 (BD Bioscience) is a nonfunctional mouse antihuman β1 mAb used for detecting cell surface expression of β1 integrin.

6

Zhengli Wang and Jieqing Zhu

(b) VC5 (BD Bioscience) is a nonfunctional mouse antihuman α5 mAb used for detecting cell surface expression of α5 integrin. 4. Two conformation-dependent mAbs can be used for probing β1 integrin conformational activation: (a) 12G10 (Bio-Rad) is a mouse mAb that reports β1 conformational change at the βI domain [19]. (b) 9EG7 (BD Bioscience) is a rat mAb that reports β1 extension by binding to the I-EGF2 domain [20, 21]. 5. Anti-αLβ2 antibodies: (a) TS2/4 (BioLegend) is a nonfunctional mouse anti-human αL mAb used for detecting cell surface expression of αLβ2 integrin. 6. Two conformation-dependent mAbs can be used for probing β2 integrin conformational activation: (a) The mouse mAb m24 (BioLegend) binds to the β2 βI domain and reports β2 headpiece opening [22]. (b) KIM127 (ATCC CRL-2838) is a mouse mAb that reports β2 integrin headpiece extension by binding to the I-EGF2 domain [23]. 2.6 Secondary Antibodies and Other Probes

Depending on the selection of ligands and primary antibodies, the following secondary antibodies or probes can be used as detection reagents: 1. Alexa Fluor 647-labeled goat anti-mouse IgM (Jackson ImmunoResearch). 2. Alexa Fluor Scientific).

647-labeled

streptavidin

(ThermoFisher

3. Biotin-labeled mouse anti-human IgG1 (ThermoFisher Scientific). 4. Alexa Fluor 488-labeled goat anti-mouse IgG (ThermoFisher Scientific). 5. Alexa Fluor 647-labeled goat anti-rat IgG (Abcam). 6. The following probes can be used for protein labeling: (a) Alexa Fluor 488 NHS ester (ThermoFisher Scientific). (b) Alexa Fluor 647 NHS ester (ThermoFisher Scientific). (c) EZ-Link Sulfo-NHS-Biotin (ThermoFisher Scientific). 7. Labeling requires 1 M sodium bicarbonate buffer, pH 8.3, which can be stored at
20  C for at least 2 months. 2.7 Integrin Inhibitors

The integrin-specific small molecule or peptide inhibitors can be used to block integrin ligand binding for negative controls, which give nonspecific background binding of ligands to integrin

Integrin-Ligand Interaction on the Cell Surface

7

expressing cells. Alternatively, 5 or 10 mM EDTA can be used as universal inhibitors for blocking integrin–ligand interaction. Some of the specific reagents are as follows: 1. Eptifibatide (Santa Cruz Biotechnology) is a high-affinity peptide inhibitor that is specific for αIIbβ3 integrin. 2. Cilengitide (Sigma-Aldrich) is a high-affinity peptide inhibitor that is specific for αVβ3 integrin. 3. A286982 (Santa Cruz Biotechnology) is a high-affinity smallmolecule inhibitor that is specific for αLβ2 integrin. 4. The disulfide cyclized peptide ACRGDGWCG can be used as a high-affinity inhibitor for α5β1 integrin [24–26]. The inhibitors can be prepared as 1–10 mM stock solutions. 2.8 Cell Culture Medium and Transfection

1. Culture medium: DMEM supplemented with MEM nonessential amino acids, L-glutamine, penicillin–streptomycin, and 10% heat-inactivated fetal bovine serum. Cell are routinely passed using trypsin–EDTA. 2. Endofectin Max (Genecopeia) or a similar transfection reagent.

3

Methods

3.1 Protein Ligand Preparation

Most multivalent protein ligands including PAC-1, fibrinogen, fibronectin, vitronectin, and ICAM-1-Fc are resuspended in conventional PBS.

3.2 FN9-10 Fragment Preparation

The monovalent ligand Fn9-10 can be expressed as a GST fusion protein using vector pGEX-6p-1 in E. coli BL21 by conventional protein purification methods [27].

3.3 ICAM-1-Fc Fragment Preparation

ICAM-1-Fc can be expressed as a secreted protein in HEK293FT cells via transient transfection using polyethylenimine (PEI). The cell culture supernatant can be directly used for ligand binding assay without purification. The ICAM-1-Fc concentration in the supernatant can be determined by conventional ELISA using antiICAM-1 mAb as the capture antibody and peroxidase-conjugated mouse anti-human IgG1 Fc as the detection antibody. Purified ICAM-1-Fc is used as a standard for determining protein concentration.

3.4 Protein Ligand and Antibody Labeling

Depending on the combination of fluorescent dyes, the ligands and antibodies can be labeled with biotin or Alexa Fluor dyes. We routinely use the amine-reactive biotin and Alexa Fluor dye for labeling protein ligands and antibodies. As an example, we describe the labeling protocol of 1 mg human fibrinogen with Alexa Fluor 647. For antibody labeling, the protocol can be scaled down to

8

Zhengli Wang and Jieqing Zhu

0.2 mg in 0.2 ml (see Note 3). For antibody biotinylation, we use the EZ-Link sulfo-NHS-biotinylation kit according to the manufacturer’s instructions. If there is no sufficient amount of antibody that can be labeled, an alternative protocol can be used for unlabeled antibodies as described below. 1. Dissolve or dilute 1 mg of fibrinogen into 0.5 ml 0.1 M sodium bicarbonate buffer (pH 8.3). If the protein solution contains chemicals such as Tris or glycine that are reactive to the dye, it can be dialyzed against PBS using Slide-A-Lyzer dialysis cassettes or similar dialysis devices. 2. Dissolve the Alexa Fluor 647 dye in DMSO at 10 mg/ml. It needs to be freshly prepared each time of use. 3. Add 10 μl of Alexa Fluor 647 dye into the 0.5 ml fibrinogen solution. 4. Incubate the reaction for 1 h at room temperature on an Eppendorf tube roller. 5. Immediately transfer the solution into a Slide-A-Lyzer dialysis cassette with 10 kDa molecular weight cut off and dialyze against 2000 ml PBS at 4  C. 6. Change PBS twice a day and dialyze for 3 days. 7. Store the labeled fibrinogen in aliquots at
80  C. 3.5

Cell Transfection

We routinely use the fast-growing HEK293FT cells as carrier cells for integrin expression (see Note 4). There is no detectable expression of endogenous αIIbβ3, αVβ3, and αLβ2 integrins. However, the endogenous expression of α5 and β1 integrins can be easily detected by flow cytometry. We have generated the α5 and β1 double knockout HEK293FT cells for the ligand-binding assay of α5β1 integrin [4, 6]. The HEK293FT cells can be transfected at the time of cell seeding without the need for cell preparation 1 day ahead of transfection. Here we describe a protocol of HEK293FT transfection using EndoFectin Max in 12-well cell culture plate. 1. HEK293FT cells are detached by trypsin–EDTA, washed with DMEM–FBS, and resuspended at 0.5  106/ml in DMEMFBS without antibiotics. 2. For each well of 12-well plate, dilute 0.5 μg each of integrin α and β DNA constructs in 50 μl DMEM. For cotransfection with EGFP and/or mCherry constructs, 0.5 μg of each DNA construct can be added. 3. Adequate controls need to be included for the transfection: the cells transfected with empty parent DNA vectors as mock transfection; for the transfections including EGFP-TH and/or mCherry-K2, the cells transfected with pEGFP-C1 and/or pmCherry-C1 along with integrins.

Integrin-Ligand Interaction on the Cell Surface

9

4. For 1 μg of total plasmid DNA, dilute 3 μl of EndoFectin Max in 50 μl DMEM. The weight–volume ratio of DNA/EndFectin is 1:3. If the EGFP and mCherry constructs are cotransfected with integrin, the total amount of EndoFectin needs to be adjusted accordingly. 5. Mix the diluted DNA and EndoFectin and incubate at room temperature for 15 min. 6. Add the HEK293FT cells from step 1 to a 12-well plate, 1 ml/ well. 7. Add the DNA–EndoFectin mixture to the cells drop by drop, shake the plate to mix thoroughly. 8. Incubate the cells at 37  C in a CO2 incubator for a total of 24–48 h before ligand binding assay. 3.6 Soluble Ligand and Antibody Binding Assay

Depending on the usage of ligands and antibodies, the different combinations of probing reagents can be selected as shown in Table 1. The following protocol provides a general procedure for the washing and incubation steps for cells transfected in 12-well plate. A V-bottom 96-well plate and multichannel pipette are recommended for the binding assay when handling multiple samples in various conditions (see Note 5). 1. Aspirate the cell culture medium, add 0.5 ml PBS to wash the cell in the well, aspirate the PBS and add 0.1 ml trypsin–EDTA, and incubate the plate at 37  C for 30–60 s (see Note 6). 2. Add 0.5 ml DMEM-FBS to suspend the cells thoroughly and transfer the cells to a 1.5-ml tube. 3. Spin down at 500  g for 5 min. 4. Aspirate the medium and add 0.5 ml HBSGB (25 mM HEPES pH 7.4, 150 mM NaCl, 2.75 mM glucose, 0.5% BSA) to resuspend the cells. 5. Spin down at 500  g for 5 min. 6. Aspirate the HBSGB and add 100 μl HBSGB (no EDTA or metal ions) to resuspend the cells thoroughly. 7. Add the suspension cells to 3 wells of a V-bottom 96-well plate, 22.5 μl/well. 8. Add 22.5 μl HBSGB-EDTA (HBSGB + 10 mM EDTA) or HBSGB-2Ca/Mg (HBSGB + 2 mM CaCl2 + 2 mM MgCl2) plus 20 μM integrin inhibitors to one well of cells for an inhibitory condition as a negative control for ligand binding; Add 22.5 μl HBSGB-2Ca/Mg to one well of cells as a physiological metal ion condition; Add 22.5 μl HBSGB2Ca/Mn (HBSGB + 0.4 mM CaCl2 + 4 mM MnCl2) to one well of cells for an activating condition as a positive control. HBSGB-EDTA condition cannot be used for the LIBS mAb

10

Zhengli Wang and Jieqing Zhu

Table 1 Selection of integrin ligands, antibodies, and detection probes Cell transfection

First incubation

Second incubation

αIIb + β3

PAC-1

Alexa 488-AP3 + Alexa 647-goat-antimouse IgM

αIIb + β3

Alexa 647-fibrinogen

Alexa 488-AP3

αIIb + β3 + EGFPTH  mCherry-K2

PAC-1

Alexa 647-goat-anti-mouse IgM + PE-AP3 (or biotin-AP3 + PE-streptavidin)

αIIb + β3 + EGFPTH  mCherry-K2

Alexa 647-fibrinogen

PE-AP3 (or biotin-AP3 + PE-streptavidin)

αIIb + β3

Alexa 488-AP3 + Alexa 647-streptavidin Biotin-labeled LIBS mAb (319.4, AP5, 370.3, or LIBS1)

αIIb + β3 + EGFPTH  mCherry-K2

Biotin-labeled LIBS mAb PE-AP3 + Alexa 647-streptavidin (319.4, AP5, 370.3, or LIBS1)

αV + β3

Alexa 647-fibronectin or Alexa 647-vitronectin

Alexa 488-AP3

αV + β3 + EGFPTH  mCherry-K2

Alexa 647-fibronectin or Alexa 647-vitronectin

PE-AP3 (or biotin-AP3 + PE-streptavidin)

α5 + β1  EGFPTH  mCherry-K2

Alexa 647-fibronectin or Alexa 647-Fn9-10

PE-MAR4

α5 + β1  EGFPTH  mCherry-K2

Biotin-12G10

PE-MAR4 + Alexa 647-streptavidin

α5 + β1  EGFPTH  mCherry-K2

9EG7

PE-MAR4 + Alexa 647-goat-anti-rat IgG (preabsorbed)

αL + β2  EGFPTH  mCherry-K2

ICAM-1-Fc + biotin-goat-antihuman IgG

PE-TS2/4 + Alexa 647-streptavidin

αL + β2  EGFPTH  mCherry-K2

Biotin-labeled LIBS mAb (KIM127 or m24)

PE-TS2/4 + Alexa 647-streptavidin

binding assay. To obtain a background fluorescence for LIBS mAb binding assay, it is optional to add a condition of mAb isotype control or a condition of no LIBS mAb. As a positive control for LIBS mAb binding, a ligand-mimetic integrin inhibitor such as eptifibatide for αIIbβ3, which is known to induce integrin conformational changes, can be used in the buffer of HBSGB-2Ca/Mg or HBSGB-2Ca/Mn. 9. Mix the cells by shaking the plate and incubate the plate at 25  C for 15 min. 10. According to Table 1, prepare 10 stock solution of ligands or LIBS mAbs in HBSGB (no EDTA or metal ions): 50 μg/ml

Integrin-Ligand Interaction on the Cell Surface

11

PAC-1; 500 μg/ml Alexa 647-fibrinogen or Alexa 647-Fn910; 150 μg/ml Alexa 647-fibronectin or Alexa 647-vitronectin; 420 μg/ml ICAM-1-Fc plus 450 μg/ml biotin-goat-anti-human IgG (Fc specific); 100 μg/ml biotinlabeled LIBS mAb. The total volumes needed can be determined based on the number of wells  5 μl. 11. Add the 10 stock solution of ligand or LIBS mAb to the cells, 5 μl/well. The final reaction volume is 50 μl in one well. 12. Mix the cells by shaking the plate and incubate the cells at 25  C for 30 min. 13. Prepare the 1 detection solution according to Table 1 in HBSGB-1Ca/Mg buffer (HBSGB + 1 mM CaCl2 + 1 mM MgCl2) at 5–10 μg/ml. For example, for PAC-1 binding, prepare 5 μg/ml Alexa 488-AP3 plus 5 μg/ml Alexa 647-goat-anti-mouse IgM in HBSGB-1Ca/Mg. The total volumes needed can be determined based on the number of wells  50 μl. 14. Spin down the cells at 500  g for 5 min. 15. Aspirate the supernatant and add 150 μl HBSGB-1Ca/Mg to each well to resuspend the cells. 16. Spin down the cells at 500  g for 5 min. It is optional to fix the cells with 4% formaldehyde in PBS. 17. Aspirate the supernatant and add the 1 detection solution at 50 μl/well to resuspend the cells by pipetting. 18. Incubate the cells on ice for 30 min, protected from light. 19. Spin down the cells at 500  g for 5 min. 20. Aspirate the supernatant and add 150 μl HBSGB-1Ca/Mg to each well to resuspend the cells by pipetting. The cells are ready for flow cytometry. It is optional to wash the cells once with HBSGB-1Ca/Mg and then fix the cells with 4% formaldehyde in PBS before flow cytometry analysis. 3.7

Flow Cytometry

We routinely use the BD Accuri C6 flow cytometer for analyzing the cells. As shown in Fig. 1, the single-cell events are gated (gate P1) on the FSC-SSC plot for analyzing the integrin expression (Alexa 488-AP3 for β3 integrin) and ligand binding (Alexa 647-goat-anti-mouse IgM for PAC-1 binding). Based on the mock-transfected cells, the integrin-expressing cells are gated (gate P2 in P1) and acquired for data analysis. We recommend acquiring at least 5000 events of integrin-expressing cells for each condition (see Note 7). If integrin, EGFP and/or mCherry constructs are cotransfected, the integrin–EGFP double-positive or integrin–EGFP–mCherry triple-positive cells will be gated and acquired for data analysis. Due to the overlap of EGFP with PE and/or mCherry, proper compensation needs to be set up when acquiring the fluorescent events.

12

Zhengli Wang and Jieqing Zhu

Fig. 1 Data acquisition and analysis of soluble PAC-1 binding to αIIbβ3-transfected HEK293FT cells. The cells were analyzed with a BD Accuri C6 flow cytometer. For the gate setup, only the EDTA or Ca/Mg condition is shown 3.8

Data Analysis

Figure 1 shows how integrin-expressing cells for each condition are gated (gate P2 in P1) for calculating the mean fluorescence intensity (MFI) of integrin expression (MFI of Alexa 488-AP3 for β3 integrin) and the ligand or LIBS mAb binding (MFI of Alexa 647-goat-anti-mouse IgM for PAC-1 binding). The MFI of integrin expression can be averaged over the three conditions. That is, MFIintegrin-averaged ¼ (MFIintegrin-EDTA (or integrin inhibitor) + MFIintegrin-Ca/Mg + MFIintegrin-Ca/Mn)/3. The MFI of ligand binding is then normalized to the MFI of integrin expression after subtracting the MFI of background binding under the negative control condition (EDTA or integrin inhibitor). That is MFIligand-normalized ¼ (MFIligand-Ca/Mg or Ca/Mn
MFIligand-EDTA (or integrin inhibitor))/MFIintegrin-averaged  100. The MFI of LIBS mAb binding can be normalized to integrin expression with or without subtracting the background MFI. When the EGFP-TH and/or mCherryK2 constructs are cotransfected with integrin, the MFI of EGFP and/or mCherry needs to be compared among the transfections

Integrin-Ligand Interaction on the Cell Surface

13

such as integrins with different mutations. This is to check that the differences observed in integrin activation are not due to the differences in the expression of EGFP-TH and/or mCherry-K2 among the transfections (see Note 8). 3.9 An Alternative Method

4

For the LIBS mAb binding assay using transiently transfected cells, if the fluorescence-labeled antibodies are not available, an alternative method is to use EGFP as a cotransfection monitor. The transfected cells need to be incubated separately with an antibody that reports total surface expression of integrin and a LIBS mAb that reports integrin conformational changes. During flow cytometry analysis, the EGFP-positive cells are gated and acquired for calculating the MFI of integrin expression and LIBS mAb binding. The MFI of LIBS mAb binding is then normalized to the MFI of integrin expression as described above (see Note 9).

Notes 1. All the protein ligands except FN9-10 used in this protocol are multivalent ligands that have more than one integrin binding site. Therefore, the avidity effect is involved in the soluble ligand binding. 2. In general, there is a good correlation between PAC-1 binding and fibrinogen binding for αIIbβ3 integrin activation assay. However, PAC-1 is more sensitive to the weak activating αIIbβ3 mutations than fibrinogen. 3. Over labeling needs to be avoided when labeling the protein ligands with the fluorescent dyes. 4. To have a stable and reproducible results, high and stable efficiency of transient transfection is critical. It is critical to maintaining a healthy status of HEK293FT cells for high transfection efficiency. If an obvious decrease in transfection efficiency is observed, a new batch of cells needs to be thawed. 5. Saturated concentrations of ligands and antibodies are recommended. 6. It is important to detach the adhesion cells into single cells during trypsin–EDTA digestion. Alternatively, the cells can be filtered with a 40 μm cell strainer before ligand binding and flow cytometry analysis. 7. For the activating integrin mutants that greatly decrease integrin surface expression, the total number of transfected cells needs to be increased in the binding assay in order to acquire a sufficient number of integrin-expressing cells for MFI calculation.

14

Zhengli Wang and Jieqing Zhu

8. If stable cell lines need to be generated, we recommend HEK293 or CHO-K1 cells. However, it should be noted that there is an endogenous expression of α5β1 and αV integrins in these cell lines. For making αIIbβ3 stable cell lines in CHO-K1 cells, the αIIbβ3 complex–specific antibody such as 10E5 or AP-2 needs to be used for detecting the cell surface expression. And the expression of αVβ3 (formed by hamster αV and human β3) can be detected by the αVβ3 complex–specific antibody such as LM609. 9. This protocol can be easily adapted to the usage of other wellcharacterized nonfunctional and conformation-dependent antibodies [3].

Acknowledgments This work was supported by NIH grant HL131836 to J. Zhu. References 1. Springer TA, Dustin ML (2012) Integrin inside-out signaling and the immunological synapse. Curr Opin Cell Biol 24(1):107–115. https://doi.org/10.1016/j.ceb.2011.10.004 2. Askari JA, Buckley PA, Mould AP, Humphries MJ (2009) Linking integrin conformation to function. J Cell Sci 122(Pt 2):165–170. https://doi.org/10.1242/jcs.018556. 122/2/165 [pii] 3. Byron A, Humphries J, Askari JA, Craig SE, Mould AP, Humphries M (2009) Anti-integrin monoclonal antibodies. J Cell Sci 122 (22):4009–4011 4. Thinn AMM, Wang Z, Zhu J (2018) The membrane-distal regions of integrin alpha cytoplasmic domains contribute differently to integrin inside-out activation. Sci Rep 8 (1):5067. https://doi.org/10.1038/s41598018-23444-w 5. Wang Z, Thinn AMM, Zhu J (2017) A pivotal role for a conserved bulky residue at the alpha1-helix of the alphaI integrin domain in ligand binding. J Biol Chem 292 (50):20756–20768. https://doi.org/10. 1074/jbc.M117.790519 6. Cai X, Thinn AMM, Wang Z, Shan H, Zhu J (2017) The importance of N-glycosylation on beta3 integrin ligand binding and conformational regulation. Sci Rep 7(1):4656. https:// doi.org/10.1038/s41598-017-04844-w 7. Liu J, Wang Z, Thinn AM, Ma YQ, Zhu J (2015) The dual structural roles of the membrane distal region of the alpha-integrin

cytoplasmic tail during integrin inside-out activation. J Cell Sci 128(9):1718–1731. https:// doi.org/10.1242/jcs.160663 8. Zhang C, Liu J, Jiang X, Haydar N, Zhang C, Shan H, Zhu J (2013) Modulation of integrin activation and signaling by alpha1/alpha10 -helix unbending at the junction. J Cell Sci 126(Pt 24):5735–5747. https://doi.org/10. 1242/jcs.137828 9. Bouaouina M, Lad Y, Calderwood DA (2008) The N-terminal domains of talin cooperate with the phosphotyrosine binding-like domain to activate beta1 and beta3 integrins. J Biol Chem 283(10):6118–6125. https://doi.org/ 10.1074/jbc.M709527200 10. Taub R, Gould RJ, Garsky VM, Ciccarone TM, Hoxie J, Friedman PA, Shattil SJ (1989) A monoclonal antibody against the platelet fibrinogen receptor contains a sequence that mimics a receptor recognition domain in fibrinogen. J Biol Chem 264(1):259–265 11. Kouns WC, Newman PJ, Puckett KJ, Miller AA, Wall CD, Fox CF, Seyer JM, Jennings LK (1991) Further characterization of the loop structure of platelet glycoprotein IIIa: partial mapping of functionally significant glycoprotein IIIa epitopes. Blood 78(12):3215–3223 12. Peerschke EI, Coller BS (1984) A murine monoclonal antibody that blocks fibrinogen binding to normal platelets also inhibits fibrinogen interactions with chymotrypsin-treated platelets. Blood 64(1):59–63

Integrin-Ligand Interaction on the Cell Surface 13. Xiao T, Takagi J, Wang J-h, Coller BS, Springer TA (2004) Structural basis for allostery in integrins and binding of fibrinogen-mimetic therapeutics. Nature 432(7013):59–67 14. Pidard D, Montgomery RR, Bennett JS, Kunicki TJ (1983) Interaction of AP-2, a monoclonal antibody specific for the human platelet glycoprotein IIb–IIIa complex, with intact platelets. J Biol Chem 258:12582–12586 15. Cheresh DA (1987) Human endothelial cells synthesize and express an Arg-Gly-Aspdirected adhesion receptor involved in attachment to fibrinogen and von Willebrand factor. Proc Natl Acad Sci U S A 84:6471–6475 16. Chen Y, Ju LA, Zhou F, Liao J, Xue L, Su QP, Jin D, Yuan Y, Lu H, Jackson SP, Zhu C (2019) An integrin alphaIIbbeta3 intermediate affinity state mediates biomechanical platelet aggregation. Nat Mater 18(7):760–769. https://doi.org/10.1038/s41563-019-03236 17. Honda S, Tomiyama Y, Pelletier AJ, Annis D, Honda Y, Orchekowski R, Ruggeri Z, Kunicki TJ (1995) Topography of ligand-induced binding sites, including a novel cation-sensitive epitope (AP5) at the amino terminus, of the human integrin b3 subunit. J Biol Chem 270:11947–11954 18. Frelinger AL, Cohen I, Plow EF, Smith MA, Roberts J, Lam SCT, Ginsberg MH (1990) Selective inhibition of integrin function by antibodies specific for ligand-occupied receptor conformers. J Biol Chem 265:6346–6352 19. Mould AP, Garratt AN, Askari JA, Akiyama SK, Humphries MJ (1995) Identification of a novel anti-integrin monoclonal antibody that recognises a ligand-induced binding site epitope on the b1 subunit. FEBS Lett 363:118–122 20. Askari JA, Tynan CJ, Webb SE, MartinFernandez ML, Ballestrem C, Humphries MJ (2010) Focal adhesions are sites of integrin

15

extension. J Cell Biol 188(6):891–903. https://doi.org/10.1083/jcb.200907174. jcb.200907174 [pii] 21. Su Y, Xia W, Li J, Walz T, Humphries MJ, Vestweber D, Cabanas C, Lu C, Springer TA (2016) Relating conformation to function in integrin alpha5beta1. Proc Natl Acad Sci U S A 113(27):E3872–E3881. https://doi.org/10. 1073/pnas.1605074113 22. Chen X, Xie C, Nishida N, Li Z, Walz T, Springer TA (2010) Requirement of open headpiece conformation for activation of leukocyte integrin αXβ2. Proc Natl Acad Sci U S A 107(33):14727–14732 23. Nishida N, Xie C, Shimaoka M, Cheng Y, Walz T, Springer TA (2006) Activation of leukocyte β2 integrins by conversion from bent to extended conformations. Immunity 25 (4):583–594 24. Koivunen E, Wang B, Ruoslahti E (1995) Phage libraries displaying cyclic peptides with different ring sizes: ligand specificities of the RGD-directed integrins. Biotechnology (N Y) 13(3):265–270. https://doi.org/10.1038/ nbt0395-265 25. Humphries JD, Askari JA, Zhang XP, Takada Y, Humphries MJ, Mould AP (2000) Molecular basis of ligand recognition by integrin a5b1. II. Specificity of Arg-Gly-Asp binding is determined by Trp157 OF THE alpha subunit. J Biol Chem 275:20337–20345 26. Xia W, Springer TA (2014) Metal ion and ligand binding of integrin alpha5beta1. Proc Natl Acad Sci U S A 111(50):17863–17868. https://doi.org/10.1073/pnas.1420645111 27. Li M, Feng Z, Zhang G, Li D (2006) Highlevel expression of a recombinant fragment of human fibronectin containing the Cell I-Hep II-IIICS71 domain in Escherichia coli as a soluble protein. Biotechnol Lett 28 (14):1141–1146. https://doi.org/10.1007/ s10529-006-9066-y

Chapter 2 Quantification of Integrin Activation and Ligation in Adherent Cells Zaki Al-Yafeai and A. Wayne Orr Abstract Integrin activation is a crucial event for multiple biological functions. Therefore, methods to detect integrin activation are vital. Since the main cellular function of integrins is adhesion, we and others utilize this feature to measure integrin activation. Here, we describe how to detect the activity of the fibronectinbinding integrin α5β1 using a fusion of glutathione S-transferase (GST) to the 9th, 10th, and 11th type III repeats on fibronectin (GST-FNIII9-11). Moreover, we detail how to measure αvβ3 integrin activity using the ligand-mimetic WOW-1 antibody that selectively binds unoccupied αvβ3 integrins. Finally, we describe methods of testing ligation of fibronectin-binding integrins utilizing monoclonal antibodies against ligandinduced binding sites (LIBS). Key words α5β1, αvβ3, Fibronectin, WOW-1, GST-FNIII9-11, LIBS

1

Introduction Methods to determine integrin activation have been intensely investigated in nonadherent cells. Indeed, the best-characterized pathways leading to integrin activation have been described in platelets [1]. However, measuring integrin activity in adherent cells has been a challenge, since these cells are already bound to the underlying matrix. Integrins primarily exist in either an inactive, bent state or an extended, active state. Upon activation, integrin affinity is enhanced to bind to its ligand. The activity assay for α5β1, the classic fibronectin receptor, was first described in 2001 by Tzima et al. [2]. This assay is based on the fact that α5β1 binds to ninth (the synergy site with PHSRN sequence) and tenth (Arg-GlyAsp or RGD sequence) fibronectin type III repeats [3] (Fig. 1). αvβ3 integrin also binds to fibronectin, among others, and plays a crucial role in cell adhesion and spreading. In 1999, the Shattil group [4] developed a method to detect αvβ3 activity by replacing the heavy chain hypervariable region 3 of Fab region of PAC1, the classic αIIbβ3 antibody, with a single integrin-binding domain from

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_2, © Springer Science+Business Media, LLC, part of Springer Nature 2021

17

18

Zaki Al-Yafeai and A. Wayne Orr

Fig. 1 Model for methods to measure integrin activation. GST-FNIII9-11 peptide binds specifically to extended, active α5β1 integrins, whereas WOW-1 antibody recognizes αvβ3 integrins. Monoclonal antibodies target and detect active α5β1 and αvβ3 integrins. SNAKA51 binds active α5 subunit, 12G10 binds active β1, whereas LIBS6 selectively detects active β3 integrins

multivalent adenovirus penton base that contains multiple RGD subunits [5] (Fig. 1). Once a ligand binds active integrins, conformational changes take place that expose neoepitopes, known as ligand-induced binding sites (LIBS) [6–8]. Multiple monoclonal antibodies have been described to bind these LIBS and can be exploited to study integrin activity. Among the anti-LIBS, we describe SNAKA51 [9], 12G10 [10], and LIBS6 [11] that specifically recognize LIBS on active α5, β1, and β3 integrins, respectively (Fig. 1).

2

Materials Prepare all solutions using ultrapure water and analytical grade reagents. Follow proper laboratory safety procedures and waste disposal regulations as recommended by product guides and institutional regulations.

2.1 α5β1 Activation Assay

1. E. coli-expressing pGEX-GST-FNIII9-11. 2. Lysogeny broth (LB).

2.1.1 Growing GST-FNIII9-11 Peptide

3. 100 μg/ml ampicillin.

2.1.2 Purification of GST-FNIII9-11 Peptide

1. Phosphate buffer saline buffer (PBS).

4. 0.5 M isopropyl β-D-1-thiogalactopyranoside (IPTG).

2. Protein isolation column with Glutathione-Sepharose beads (GE Healthcare). 3. Protease inhibitor cocktail V (RPI). 4. 0.3 mM glutathione in 50 mM Tris–HCl. Adjust pH to 11 with NaOH (elution buffer).

Integrin Affinity in Adherent Cells

19

5. 1 M Tris–HCl, pH 8. 6. 6 M Guanidine hydrochloride. 7. GelCode Blue Stain Reagent (Thermo Fisher). 2.1.3 Activation Assay

1. Cells. 2. 1 mM MgCl2 in PBS. 3. GST-FNIII9-11 peptide. 4. 2
Laemmli Buffer: 10% (w/v) sodium dodecyl sulfate, 40% glycerol, 2 M β-mercaptoethanol in 0.33 M Tris–HCl, pH 6.8 + 0.1 g bromophenol blue.

2.1.4 Immunoblotting

1. SDS-PAGE gel. 2. Polyvinylidene fluoride or polyvinylidene difluoride (PVDF) membrane. 3. Blocking buffer: 5% dry milk in TBST. 4. Mouse anti-GST antibody (Santa Cruz Biotechnology, 1:1000 dilution) in 1% BSA TBS solution. 5. HRP-conjugated rabbit anti-mouse secondary antibody (1:1000). 6. Luminol detection kit.

2.2 αvβ3 Activation Assay

1. Stimulated cells.

2.2.1 Activation Assay

3. WOW-1 antibody.

2. 1 mM MgCl2 in PBS. 4. 2
Laemmli Buffer.

2.2.2 Immunoblotting

1. SDS-PAGE gel. 2. Polyvinylidene fluoride or polyvinylidene difluoride (PVDF) membrane. 3. Blocking buffer: 5% dry milk in TBST. 4. Rabbit anti-His antibody to detect WOW-1 (Cell Signaling Technology-2365; 1:1000) in 1% BSA solution. 5. HRP-conjugated rabbit anti-mouse secondary antibody (1:1000). 6. Luminol detection kit.

2.3 Integrin Ligation Measurement 2.3.1 Western Blotting

1. LIBS antibodies, azide free. (a) 12G10 (5 μg/ml): Mouse anti-human β1 integrins. (b) SNAKA51: (10 μg/ml) Mouse anti-human α5 integrins. (c) LIBS6: Used at 15 μg/ml to activate αvβ3, and at 5 μg/ml to detect ligated integrin. 2. SDS-PAGE gel.

20

Zaki Al-Yafeai and A. Wayne Orr

3. Polyvinylidene fluoride or polyvinylidene difluoride (PVDF) membrane. 4. Blocking buffer: 5% dry milk in TBST. 5. HRP-conjugated secondary antibodies (1:1000). 6. Luminol detection kit. 2.4 Immunocytochemistry

1. Stimulated cells on glass coverslips. 2. 4% formaldehyde in PBS. 3. 0.1% Triton X-100 in TBST. 4. TBST. 5. Blocking buffer: 10% serum, 1% BSA in PBS. 6. Primary antibodies in TBST. (a) 12G10 (5 μg/ml): Mouse anti-human β1 integrins. (b) SNAKA51: (10 μg/ml) Mouse anti-human α5 integrins. (c) LIBS6: Used at 15 μg/ml to activate αvβ3, and at 5 μg/ml to detect ligated integrin. 7. Secondary antibodies in blocking buffer: Alexa Fluor–conjugated anti-mouse (Molecular Probes, 1 μg/ml). Centrifuge the secondary antibody solution at maximum speed for 30 min to spin down dye crystals. Pipet solution to leave a small amount at bottom to not disturb dye crystals. 8. 40 ,6-diamidino-2-phenylindole (DAPI) in PBS (5 mg/ml stock, working dilution 1:50,000). 9. Fluoromount G.

3

Methods

3.1 α5β1 Activation Assay 3.1.1 Growing GST-FNIII9-11 Peptide

1. Inoculate two 5 ml aliquots of LB media containing ampicillin with E. coli expressing pGEX-FNIII9-11 and incubate at 37  C overnight with agitation in a bacterial incubator (shake at approximately 180 rpm). 2. The next day, add each aliquot to 500 ml of LB with ampicillin and incubate at 37  C until proper levels of bacteria as determined by O.D. measurement (600 nm) of 0.6. 3. Add 500 μM IPTG to the bacteria for 4 h incubation at 37  C to induce expression of GST-FNIII9-11 protein. 4. Pellet the bacteria/protein by centrifugation at 3000
g for 20 min at 4  C. Drain the supernatant. Store pellets at 80  C until ready for protein purification.

Integrin Affinity in Adherent Cells 3.1.2 Purification of GST-FNIII9-11 Peptide

21

1. Wash glutathione column with cold 1
PBS for 1 h. 2. Resuspend bacterial pellet in 7 ml PBS containing protease inhibitor [1
]. Aliquot sample into four tubes and sonicate three times for 15 s each at a low setting (Output ~2; constant sonication). Centrifuge sonicated samples at 17,000
g for 15 min at 4  C. Collect supernatant and store on ice. 3. Remove 3 μl of supernatant and place in a labeled microcentrifuge tube with 7 μl of 2
Laemmli Buffer. Keep for later. 4. Load bacterial supernatant onto the column and allow to pass through by gravity flow. Collect flow through and keep on ice. Remove 3 μl of flow through and place in a labeled microcentrifuge tube with 7 μl of 2
Laemmli Buffer. Keep for later. 5. Rinse column with cold 1
PBS for 2 h. 6. Prepare ten 1.5 ml microcentrifuge tubes with 150 μl of Tris– HCl, pH 8 [1 M]. Label 1–10. Keep on ice. 7. Allow PBS to pass through column until approximately 1 cm above beads then add 15 ml elution buffer. Collect eluent into previously prepared 1.5 ml microcentrifuge tubes. Keep on ice. 8. Remove 3 μl from each of the 10 fractions and place in 10 clean, labeled microcentrifuge tubes with 7 μl of 2
Laemmli Buffer. 9. Run an SDS-PAGE gel with supernatant sample, flow through sample, and fraction samples to determine the fractions to keep for dialysis. 10. After gel running, stain SDS-PAGE gel with GelCode Blue stain reagent as per manufacturer’s directions. 11. Pool fractions with highest purified protein concentration into dialysis cassette. 12. Dialyze overnight in 1
PBS at 4  C. 13. Next day, change to fresh 1
PBS for 4 h. 14. Collect dialyzed protein sample and determine protein concentration on Nanodrop spectrophotometer using the Protein A-480 reading. 15. Aliquot into appropriate volumes and store at 80 until use.



C

16. Clean column by running 5 ml of Guanidine HCL [6 M] over the column. Do a final rinse with 1
PBS for 1 h. 3.1.3 Activation Assay

1. After integrin activation is stimulated, media is gently removed from cells and replaced by PBS containing 1 mM MgCl2 (see Notes 1–3). 2. Immediately add GST-FNIII9-10 (20 μg/ml) (see Notes 4 and 5).

22

Zaki Al-Yafeai and A. Wayne Orr

3. After 30 min of incubation with GST-FNIII9-11, wash the cells gently with PBS (with 1 mM MgCl2) and lyse using 2
Laemmli Buffer (50 μl is considered ideal in our hands). 4. Boil each sample for 5 min at 95  C, then centrifuge the samples for 10 s at 14,000
g. 3.1.4 Immunoblotting

1. Load samples onto an SDS-PAGE gel (see Note 6). 2. Run gel and transfer onto PVDF membrane. 3. Block membrane for 30 min in blocking buffer. 4. Rinse membrane with TBST and incubate with anti-GST primary antibody overnight at 4  C. 5. The following day, rinse thoroughly with TBST and apply an HRP-conjugated secondary antibody for 2 h. 6. Rinse with TBST and use a luminol detection system to develop onto X-ray film. 7. Anti-GST antibody will be visible around 70 kDa.

3.2 αvβ3 Activation Assay

1. After integrin activation is stimulated, remove media from cells and replace by PBS containing 1 mM MgCl2 (see Note 1–3).

3.2.1 Activation Assay

2. Immediately add WOW-1 (30 μg/ml), regardless of the number of cells (see Note 5). 3. After 30 min of incubation with WOW1, wash the cells gently with PBS (with 1 mM MgCl2) and lyse using 2
Laemmli Buffer (50 μl is considered ideal in our hands). 4. Boil each sample for 5 min at 95  C, then centrifuge the samples for 10 s at 14,000
g.

3.2.2 Immunoblotting

1. Load samples onto an SDS-PAGE gel (see Note 6). 2. Run gel and transfer onto PVDF membrane. 3. Block membrane for 30 min in blocking buffer. 4. Rinse membrane with TBST and incubate with anti-His primary antibody overnight at 4  C. 5. The following day, rinse thoroughly with TBST and apply an HRP-conjugated secondary antibody for 2 h. 6. Rinse with TBST and use a luminol detection system to develop onto X-ray film. 7. Anti-His antibody will be visible around 28 kDa.

Integrin Affinity in Adherent Cells

3.3 Integrin Ligation Measurement 3.3.1 By Western Blot

23

1. After integrin activation is stimulated, add the appropriate antiLIBS antibody to the cells (see Note 7). (a) 12G10: (5 μg/ml): Mouse anti-human β1 integrins. (b) SNAKA51: (10 μg/ml) Mouse anti-human α5 integrins. (c) LIBS6: (5 μg/ml) to detect ligated β3 integrin. 2. Lyse cells using 2
Laemmli Buffer (50 μl is considered ideal in our hands). 3. Boil each sample for 5 min at 95  C, then centrifuge the samples for 10 s at 14,000
g. 4. Load samples onto an SDS-PAGE gel. 5. Run gel and transfer onto PVDF membrane. 6. Block membrane for 30 min in blocking buffer (see Note 8). 7. Rinse thoroughly with TBST and apply goat anti-mouse IgG HRP-conjugated secondary antibody for 2 h. 8. Rinse with TBST and use a luminol detection system to develop onto X-ray film. 9. Detect 12G10 around 60 kDa, SNAKA51 around 116 kDa, and LIBS6 around 28 kDa.

3.3.2 By Immunocytochemistry

1. After stimulation, fix cells in 4% formaldehyde in PBS (see Note 7). 2. Permeabilize cells using 0.1% Triton X-100 in TBST for 10 min. 3. Rinse coverslips with TBST three times. 4. Aspirate TBST and add blocking buffer at room temperature for 1 h minimum. 5. Prepare the primary antibody solution by mixing the antibody in TBST. (a) 12G10: (5 μg/ml): Mouse anti-human β1 integrins. (b) SNAKA51: (10 μg/ml) Mouse anti-human α5 integrins. (c) LIBS6: (5 μg/ml) to detect ligated integrin. 6. Aspirate blocking buffer and do a quick TBST rinse. 7. Add primary antibody and incubate at 4  C overnight. 8. Next day, prepare the secondary antibody solution. 9. Rinse primary antibody with TBST three times. 10. Apply secondary antibody at room temperature for 2 h (1:1000). 11. Rinse secondary antibody with TBST three times. 12. Rinse triton from TBST solution by rinsing PBS.

24

Zaki Al-Yafeai and A. Wayne Orr

13. To counterstain, aspirate PBS and apply DAPI solution for 5–10 min. Do not rinse DAPI from coverslips, just lightly dab without drying. 14. Mount coverslips to microscope slides with one drop of Fluoromount G. Allow 1–2 h for coverslip to cure. 15. Store slides at 20  C until ready to image on fluorescent microscope.

4

Notes 1. Using media containing MgCl2 can replace PBS containing 1 mM MgCl2 for α5β1 and αvβ3 activation assays. 2. MnCl2 (0.5 mM) activates all integrins, therefore it can be used as a positive control for α5β1 and αvβ3 activation assays. 3. EDTA (2 mM) can be used as a negative control for integrin activity assays. 4. For soluble stimuli (e.g., oxLDL), coincubation of the stimulus with GST-FNIII9-10 would give the same results. 5. The amount of GST-FNIII9-11 (20 μg/ml) and WOW-1 (30 μg/ml) used for α5β1 and αvβ3 assays is fixed regardless of the number of cells. 6. Recombinant GST-FNIII9-11 and His-WOW-1 also can used as a positive control for α5β1 and αvβ3 integrin activation assays respectively. 7. Include ligation-insensitive antibody as a control for integrin ligation assays. Non–function-blocking antibodies such as VC5 and mAbs 11 for α5, K20 for β1, or mAb AP3 for β3 are widely used. 8. Since anti-LIBS antibodies are added prior lysis, there is no need to add primary antibodies during Western blot.

Acknowledgments This work was supported by National Heart, Lung, and Blood Institute R01 HL098435, HL133497, HL141155, and GM121307 (to A.W.O.), by an American Heart Association Predoctoral Fellowship (19PRE34380751) and Malcolm Feist Cardiovascular Research Endowment Predoctoral Fellowship (to Z.A.Y.).

Integrin Affinity in Adherent Cells

25

References 1. Shattil SJ, Kim C, Ginsberg MH (2010) The final steps of integrin activation: the end game. Nat Rev Mol Cell Biol 11(4):288–300. https://doi.org/10.1038/nrm2871 2. Tzima E, del Pozo MA, Shattil SJ, Chien S, Schwartz MA (2001) Activation of integrins in endothelial cells by fluid shear stress mediates Rho-dependent cytoskeletal alignment. EMBO J 20(17):4639–4647 3. Hynes RO (2002) Integrins: bidirectional, allosteric signaling machines. Cell 110 (6):673–687 4. Pampori N, Hato T, Stupack DG, Aidoudi S, Cheresh DA, Nemerow GR, Shattil SJ (1999) Mechanisms and consequences of affinity modulation of integrin alpha(V)beta(3) detected with a novel patch-engineered monovalent ligand. J Biol Chem 274(31):21609–21616 5. Wickham TJ, Mathias P, Cheresh DA, Nemerow GR (1993) Integrins alpha v beta 3 and alpha v beta 5 promote adenovirus internalization but not virus attachment. Cell 73 (2):309–319. https://doi.org/10.1016/ 0092-8674(93)90231-e 6. Frelinger AL 3rd, Lam SC, Plow EF, Smith MA, Loftus JC, Ginsberg MH (1988) Occupancy of an adhesive glycoprotein receptor modulates expression of an antigenic site involved in cell adhesion. J Biol Chem 263 (25):12397–12402 7. Falke JJ, Koshland DE Jr (1987) Global flexibility in a sensory receptor: a site-directed

cross-linking approach. Science (New York, NY) 237(4822):1596–1600. https://doi. org/10.1126/science.2820061 8. Parise LV, Helgerson SL, Steiner B, Nannizzi L, Phillips DR (1987) Synthetic peptides derived from fibrinogen and fibronectin change the conformation of purified platelet glycoprotein IIb-IIIa. J Biol Chem 262 (26):12597–12602 9. Clark K, Pankov R, Travis MA, Askari JA, Mould AP, Craig SE, Newham P, Yamada KM, Humphries MJ (2005) A specific alpha5beta1-integrin conformation promotes directional integrin translocation and fibronectin matrix formation. J Cell Sci 118 (Pt 2):291–300. https://doi.org/10.1242/ jcs.01623. jcs.01623 [pii] 10. Mould AP, Garratt AN, Askari JA, Akiyama SK, Humphries MJ (1995) Identification of a novel anti-integrin monoclonal antibody that recognises a ligand-induced binding site epitope on the beta 1 subunit. FEBS Lett 363 (1–2):118–122. https://doi.org/10.1016/ 0014-5793(95)00301-o 11. Frelinger AL 3rd, Du XP, Plow EF, Ginsberg MH (1991) Monoclonal antibodies to ligandoccupied conformers of integrin alpha IIb beta 3 (glycoprotein IIb-IIIa) alter receptor affinity, specificity, and function. J Biol Chem 266 (26):17106–17111

Chapter 3 Multiparametric Analysis of Focal Adhesions in Bidimensional Substrates Vanessa C. Talayero and Miguel Vicente-Manzanares Abstract Focal adhesions in planar substrates constitute an excellent cellular resource to evaluate different parameters related to cell morphology, cytoskeletal organization, and adhesive strength. However, their intrinsic heterogeneity in terms of size, molecular composition, orientation, and so on complicates their analysis. Here, we describe a simple and straightforward ImageJ/Fiji-based method to quantify several parameters that describe the morphology and relative composition of focal adhesions. This type of analysis can be implemented in various ways and become useful for drug and shRNA screenings. Key words Focal adhesion, Protrusion, Actin cytoskeleton, ImageJ

1

Introduction Since their initial description in the late 1960s and early 1970s, focal adhesions have fascinated cell biologists [1]. Initially described as meeting points between intracellular and extracellular fibers [2], these structures contain a myriad of proteins and scaffolds [3] that control different cellular processes, including cell division, proliferation, and migration [4, 5]. Whereas focal adhesion significance in vivo and in threedimensional environments is controversial [6–8], their emergence in bidimensional setups coated with integrin ligands is unquestioned, providing excellent opportunities to evaluate multiple parameters in a quantitative manner [9]. For example, focal adhesionrelated parameters can be quantified in response to treatment of the cells with shRNA, drugs or other treatments [10, 11]. Here, we provide an integrated overview of a simple and costefficient protocol that is both flexible and easy to implement, even for novice users, with particular emphasis on the explanation of criteria to eliminate false signals and circumvent high background, which often limits the applicability of this technique.

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_3, © Springer Science+Business Media, LLC, part of Springer Nature 2021

27

28

2

Vanessa C. Talayero and Miguel Vicente-Manzanares

Materials For Subheading 2.1, tissue culture grade is required for all the materials used. For Subheadings 2.2 and 2.3, all the solutions need to be prepared using ultrapure water. Specific instructions for storage and durability are provided for each reagent.

2.1 Cell and Substrate Preparation

1. Cell line of choice. For the description of the protocol, we will use U-2 OS human bone osteosarcoma cells (ATCC HTB-96), but any other mesenchymal cell can be used. Other cell lines we frequently use are CHO.K1 hamster ovary cells (ATCC CCL-61) or NIH/3T3 mouse fibroblasts (ATCC CRL-1658). Culture them as indicated by ATCC or by your cell provider. 2. 1.5H microscopy-grade round coverslips. Any size can be used, but some parameters need to be scaled accordingly. We routinely use 12 mm round coverslips. 1.5 coverslips are also acceptable. 3. Electron microscopy forceps. Two pairs are ideal. Sterilize one of them and keep them in a tissue culture hood. For most applications, ethanol rinsing is sufficient to maintain sterility. 4. Imaging buffer: 20 mM HEPES, 140 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl2 and 1.0 mM MgCl2, pH 7.4 (mOsm ¼ 300). Alternatively, serum-free culture medium can be used to prepare the fibronectin solution. 5. Fibronectin. This can be homemade [12] or purchased from a number of suppliers. Fibronectin stocks are to be preserved at 80  C in small-sized (~100 μl at ~1 mg/ml) aliquots. Prepare a working solution (needs to be titrated for each cell line, but a good starting point is 2 μg/ml) in imaging buffer. 6. 1% bovine serum albumin in PBS. 7. Common tissue culture buffers, including PBS or HBSS, a celldissociation reagent, for example 0.5% trypsin and tissue culture medium, with and without serum as indicated by the cell line provider. Trypsin inhibitor can be used if cells are grown in serum-free medium.

2.2 Fixation and Staining

1. 4% solution of methanol-free paraformaldehyde in PBS. 2. PHEMplus buffer: 60 mM PIPES, 25 mM HEPES, 10 mM EGTA, 2 mM MgCl2 pH 6.9 (PHEM) + 2% BSA + 100 μg/ml human γ-globulins. 3. 0.1% Triton X-100 in PHEMplus buffer. 4. Common laboratory buffers, PBS and TBS (see Note 1). 5. Primary antibodies. A number of antibodies have been validated for focal adhesion staining. The procedures described

Focal Adhesion Parametrization

29

here use a mouse monoclonal antibody against vinculin (Merck V9131). 6. Secondary antibody. A filter-matched secondary antibody is required. To use vinculin as described in item 4, we use an Alexa Fluor 488–conjugated goat anti-mouse (IgM + IgG) antibody. 7. (Optional). Fluorescent phalloidin (see Note 2). 2.3 Mounting and Image Collection

1. Mounting medium. Several types of media can be used. Typically, we use home-made Mowiol [13]. A number of commercially available mounting media also give good results, for example Invitrogen’s Prolong and Vector Lab’s Vectashield. 2. Immunohistochemistry slides. 3. A suitable wide field, confocal or TIRF microscope. Given the number of commercial solutions, we will not provide a full description of the specifications required. The system must have a device capable of recording images in Tiff format, for example a CCD camera coupled to image acquisition software.

2.4 Image Analysis and Parameter Quantification

3

1. ImageJ (NIH) or Fiji. 2. A graphics representation GraphPad, SPSS).

program

(e.g.,

Excel,

Methods

3.1 Coverslip Coating

1. Set up 12 mm coverslips on porcelain racks. 2. Rinse coverslips with running deionized water for 10 min. 3. Immerse the rack overnight in 20% H2SO4 (in deionized water). All the coverslips need to be separated and fully covered. 4. Next morning, rinse them with running deionized water for 10 min. 5. Rinse them once in ultrapure water. 6. Dry the coverslips overnight in a heat stove at 80  C. 7. Set them up in a clean, flat surface (we use a glass frame covered with parafilm) and sterilize them using 254 mm UV light (20 min in a tissue culture hood is sufficient). Using sterile electron microscopy forceps or similar, flip them and repeat the irradiation. 8. Cover them with a sterile cover and set them aside.

30

Vanessa C. Talayero and Miguel Vicente-Manzanares

3.2 Fibronectin Solution Preparation and Coverslip Coating

1. Dilute fibronectin at the desired concentration in imaging medium. Prepare enough for at least 50 μl of solution per coverslip (a 50–100 μl range is acceptable). 2. Form a drop in the center of each coverslip. Make sure the drop is extended uniformly. 3. Incubate at 37  C for 2 h (alternatively, they can be incubated overnight at 4  C). Use within 72 h of coating.

3.3 Coverslip Blocking

1. One hour before adhesion, transfer the coverslips to 24-well tissue culture vessels using sterile forceps. The fibronectin coated side must be face up. 2. Cover the coverslip with 500 μl of a solution containing 1% BSA in PBS pH 7.4. Incubate for at least 1 h (see Note 3).

3.4 Cell Preparation and Adhesion

1. We describe the basic protocol for nontransfected, nonmanipulated cells. Cells can be transfected with fluorescently tagged focal adhesion proteins by diverse methods, including electroporation, lipid-based methods, microinjection, etc. Since these protocols are readily available, readers are encouraged to explore their viability for their own conditions (see Note 4). 2. Dissociate the cells from its tissue culture vessel using trypsin or a similar reagent. This step should be performed in the same way as for routine passage of the culture. 3. When cells are loose, transfer them to a tissue culture tube (15 or 50 ml, depending on the amount of cells required) and double the volume using complete culture medium (i.e., containing the serum concentration used for their culture). If cells are cultured in serum-free conditions, use trypsin inhibitor as recommended by the manufacturer. 4. Centrifuge the cells at 300  g for 5 min at room temperate. 5. Resuspend the cells in imaging medium complemented with 10% FCS. If cells are grown in serum-free conditions, resuspend them in their original medium. Count the cells and adjust their concentration. The number of cells to be used depends on the cell line, because the size of the fully spread cells vary. For U-2 OS cells, deposit 80,000 cells/well in 500 μl, which means adjusting the concentration at 160,000 cells/ml. In general, a concentration of 500–700 cells/mm2 is acceptable. 6. Let them settle at 37  C in a tissue culture incubator. For rapid adhesion formation, proceed to the next step after 2-3 h. For full adhesion formation, wait at least 12 h.

3.5

Fixation

1. There are two protocols for fixation, which depend on the amount of background observed in preliminary experiments using the antibody of choice. Here, we define background as a

Focal Adhesion Parametrization

31

combination of nuclear staining and cytoplasmic, diffuse staining. Obviously, for focal adhesion quantification, a strong adhesive signal with low background is desirable, even at the expense of some intensity. For low background antibodies, go to step 2. For high background antibodies, go to Note 5. 2. Take the coverslip with clean forceps and immerse once in prewarmed (37  C) PBS. 3. Deposit it in a flat surface (a humid chamber with an opaque lid is desirable for next steps), cells facing up and overlay a 100 μl drop of 4% methanol-free paraformaldehyde in PBS on top of the cells. When overlaying, make sure you do not drop the liquid forcefully on top of the cell. A useful tip is to form a drop close to the coverslip and tug it softly toward the drop, which will cover the surface by capillarity. 4. Incubate for 10 min at room temperature. 5. Rinse the coverslip three times with 100 μl TBS. Aspirate paraformaldehyde, replace with TBS drop, aspirate, and repeat the whole procedure twice. See Note 6 for an alternative protocol for rinsing. 3.6

Permeabilization

1. Aspirate TBS from the surface of the coverslip. 2. Overlay a 100 μl drop of 0.1% Triton X-100 dissolved in PHEMplus buffer. 3. Incubate for 10 min at room temperature. 4. Rinse the coverslip three times with 100 μl TBS. Aspirate paraformaldehyde, replace with TBS drop, aspirate, and repeat the whole procedure twice. See Note 6 for an alternative protocol for rinsing.

3.7 Blocking and Staining

1. Aspirate TBS from the surface of the coverslip. 2. Overlay a 100 μl drop of PHEMplus buffer on the surface of the coverslip. Incubate for 30 min at room temperature in a humid chamber (see Subheading 3.5, step 3). 3. Prepare the dilution of the primary antibody. The anti-vinculin antibody is diluted 1:2000 in PHEMplus buffer. Prepare 100 μl of diluted antibody per coverslip. 4. Replace the 100 μl drop of PHEMplus buffer with 100 μl of the diluted antibody. At this point, two protocols can be followed: (1) Incubate in the humid chamber at 37  C for 2 h; (2) incubate overnight at 4  C. 5. After incubation, rinse the coverslip three times with 100 μl TBS. Aspirate paraformaldehyde, replace with TBS drop, aspirate, and repeat the whole procedure twice. An alternative protocol for rinsing is in Note 6.

32

Vanessa C. Talayero and Miguel Vicente-Manzanares

6. In the meantime prepare the solution containing the secondary antibody at the working dilution. Use PHEMplus buffer for dilution. Prepare 120 μl of diluted antibody per coverslip (see Note 2 to use fluorescent phalloidin). 7. Centrifuge (17,000  g, 5 min, room temperature) the secondary antibody solution to remove fluorescent debris. Transfer all but the last 100 μl of the dilution to a clean 1.5 ml tube, leaving the debris behind (it usually is not visible). 8. Replace the 100 μl drop of TBS with 100 μl of the diluted secondary antibody. Incubate in the humid chamber at 37  C for 30 min. 9. Rinse the coverslip three times with 100 μl TBS. 3.8

Mounting

1. Aspirate TBS and, using clean forceps, rinse the coverslips in ultrapure water. 2. Deposit a 15 μl drop of mounting medium (Mowiol, Prolong, Vectashield, or similar) on a clean slide. 3. Lay the coverslip with the side of the cells toward the mounting medium (that is, face down). Tap lightly with the forceps to remove air bubbles. Let it dry in the dark for at least 60 min.

3.9

Image Collection

1. Examine the samples under the microscope, as discussed in Subheading 2.3. We recommend 16-bit TIFF images obtained using a 60–63 objective, of at least 1024  1024 pixels (an adequate resolution is 0.10–0.20 μm/pixel). 2. It is essential to know the resolution of the image. Most microscopes provide this information in the metadata of your image. If not, a calibrated ruler can be used. The resolution must take the form μm/pixel. 3. Lower resolution images can be used, but the results will not be as crisp. A minimum of 512  512 pixels is strongly recommended.

3.10

Image Analysis

1. Open image in ImageJ/Fiji (Fig. 1a, if phalloidin image has been obtained, see Note 7). 2. If the image is not calibrated (if it is, the measurements that appear on the top left corner of the image have the form: aaa  bbb microns (yyy  zzz); if not, only (yyy  zzz) appears), it needs to be calibrated. To do so, please check the resolution of your image, which comes from the microscope/ camera combination used and apply the command Analyze > Set scale, filling in the data in the boxes. 3. Set up the parameters to be measured. Apply the command Analyze > Set measurements. Tick the following boxes: area, shape descriptors, perimeter, and fit ellipse. More parameters

Focal Adhesion Parametrization

33

Fig. 1 An example of image processing, corresponding to Subheading 3.10. (a) Unmodified image. (b) Background-subtracted image. (c) Application of CLAHE algorithm. (d) Application of the Exp algorithm. (e) Negative thresholding of the image. Yellow outline is drawn manually to confine the analysis. (f) Outline map for confirmation of the analysis as compared with (a). (g) Results window

can be selected, but take into account that gray values will not be useful as the image will be subjected to threshold. 4. Apply the command

Process > Subtract background,

Rolling ball radius ¼ 50 pixels, sliding paraboloid. This command reduces the background of the image taking the darkest part as a reference (Fig. 1b). As a result, the whole image will dim. Repeat if a lot of background is still visible.

5. Apply the

command Plugins > CLAHE, block size ¼ 127,

histogram bins ¼ 256, maximum slope ¼ 3.

CLAHE stands for Contrast Limited Adaptive Histogram Equalization, and it sharpens the adhesions, even at the expense of increasing the background (Fig. 1c). It needs to be downloaded separately from https://imagej.nih.gov/ij/plugins/clahe/index.html, then copied into the Plugins folder of ImageJ/Fiji. 4. Apply the command Process > Math > Exp. This increases the difference between background and adhesion, facilitating the thresholding necessary to quantify several adhesive parameters (next step). Sometimes, the image becomes quite dark (Fig. 1d), but this is not important because the next step will take care of it. New generations of microscopes provide novel features that help with this process, for example computational

34

Vanessa C. Talayero and Miguel Vicente-Manzanares

clearing (Thunder, from Leica). If such an acquisition system is available, start this part on step 6. 6. Apply the command Image > Adjust > Threshold. The box labeled dark background needs to be ticked, the rest do not. 7. Slide the top slider in the Threshold box until adhesions are visible in dark over a white background (Fig. 1e). This is the most important step and it requires a skilled observer. An originally dim image will require the top slider to go more to the left for the adhesions to become visible. The survival of obvious background areas (for example, the nucleus) does not eliminate the ability to analyze the image, but it requires additional refinement during the next step. 8. Use the tool Freehand selection to draw a loose perimeter around the whole cell (Fig. 1e, yellow outline). In this manner, we confine the analysis to this region. 9. Apply the command Analyze > Analyze particles. By doing this, the program will measure the dark spots. To avoid measuring large portions of background (which would be mistaken for large adhesions), set Size to 20-Infinity (see Note 8). In this manner, adhesions (or background particles) bigger than 20 contiguous pixels will be excluded. Circularity should be set to 0.00–1.00 to include adhesions of all shapes. Finally, display outlines in the Show menu, and tick display results and clear results from the tick boxes. 10. Two new screens will emerge; one is the outline map (Fig. 1f). This is crucial for the analysis to be accurate. In the end, only a human operator is able to distinguish an adhesion from a nonspecific accumulation of signal. Carefully parse the outline map with the original image. Make sure that all the adhesions you can identify “by eye” appear as numbered outlines. If not, go back to step 10 and repeat the analysis, setting Size to 15-Infinity, or even 25 (see Note 9). The other window is the results list (Fig. 1g). Copy results and paste them in a spreadsheet. 3.11 Data Representation and Statistics

From the spreadsheet, several types of data can be extracted, including: 1. Number of adhesions (from the spreadsheet directly). 2. Percentage of adhesive area (∑(area of all the adhesions)/total area of the cell, see Note 7). 3. Average area of each adhesion. 4. Ellipsoidal coefficient of each adhesion (1/round). 5. Represent data using a graph program.

Focal Adhesion Parametrization

35

6. Regarding some considerations when determining significant differences, see Note 10.

4

Notes 1. Common formulations of PBS and TBS are sufficient as long as they are prepared in ultrapure water. 2. If available, a fluorescent phalloidin can be added to the secondary antibody to counterstain the actin cytoskeleton. If so, you need to make sure to use a fluorescent label that is compatible with that of the secondary antibody used to visualize the focal adhesions. Also, the dilution of the fluorescent phalloidin should be tested independently. This is not necessary. 3. This step is not mandatory, but it reduces nonspecific binding of the cells to the uncoated portions of the coverslip. 4. Some plasmids to visualize focal adhesions are available from Addgene. We recommend: pEGFP Vinculin, plasmid #50513; Paxillin-pEGFP, plasmid #15233; GFP-Talin1, plasmid #26724. For some of there, our lab has single-molecule promoter vectors that enable expression of these proteins with very low background. 5. For high-background antibodies, it is desirable to remove the nuclear/cytoplasmic fraction. To do this, rinse the coverslips in warm PBS (Subheading 3.5, step 2). Then, deposit it in a flat surface with the cells facing up and overlay a 100 μl drop of 0.1% Triton X-100 dissolved in PHEMplus buffer. Incubate for 20–30 s (this time is crucial, longer incubation times will dim the adhesive signal), aspirate the Triton X-100 and overlay a 100 μl drop of 4% methanol-free paraformaldehyde. Go back to Subheading 3.5, step 4, but skip Subheading 3.6 completely. 6. If the porcelain racks used to prepare the coverslips (Subheading 3.1) are available, the coverslips can be set up in them (remember cells are on one side!) and immersed in 250 ml TBS. However, we have observed no difference between rinsing protocols. 7. The phalloidin image makes it extremely simple to calculate the total area of the cell. Open the phalloidin image in ImageJ/Fiji. Using the tool Freehand selection (looks like a lasso), draw the perimeter of the cell and apply the command Analyze > Measure. 8. Size is in square pixels. The value 20-infinity is an arbitrary cutoff, meaning that particles larger than 20 square pixels will be excluded. For very high-resolution images, or cells with very large adhesions, this cutoff may exclude adhesions that need to

36

Vanessa C. Talayero and Miguel Vicente-Manzanares

be analyzed. Use your own criteria after inspecting the outline map (Subheading 3.10, step 10). 9. If some structures (e.g., the nucleus) are still visible, they can be excluded in this step. The outlines are numbered, so you can manually “exclude” the outline corresponding to nonspecific regions. If this is more than 10% of the total adhesions, we recommend to troubleshoot the staining/analysis. 10. When calculating the significance of adhesive parameters among different conditions, you will notice that adhesion data distributions are seldom normal. To verify this, use Shapiro–Wilk test. In most cases, the adhesion population will show a skewed conformation. For this reason, to calculate significance, it is often most convenient to use Mann– Whitney test.

Acknowledgments This work was funded by the Spanish Ministry of Science and Innovation (SAF2017-87408), AECC Seed award (2018–IDEAS18018VICE) and ECRIN_M3-2020 (Accelerator Award) from CRUK/AECC/AIRC. The host institution, IBMCC, is supported by the Programa de Apoyo a Planes Estrate´gicos de Investigacio´n de Estructuras de Investigacio´n de Excelencia of the Ministry of Education of the Castilla–Leo´n Government (CLC–2017–01). References 1. Horwitz AR (2012) The origins of the molecular era of adhesion research. Nat Rev Mol Cell Biol 13(12):805–811. https://doi.org/10. 1038/nrm3473 2. Lazarides E, Burridge K (1975) Alpha-actinin: immunofluorescent localization of a muscle structural protein in nonmuscle cells. Cell 6 (3):289–298 3. Zaidel-Bar R, Itzkovitz S, Ma’ayan A, Iyengar R, Geiger B (2007) Functional atlas of the integrin adhesome. Nat Cell Biol 9 (8):858–867. https://doi.org/10.1038/ ncb0807-858 4. Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, Parsons JT, Horwitz AR (2003) Cell migration: integrating signals from front to back. Science 302 (5651):1704–1709 5. Humphries JD, Paul NR, Humphries MJ, Morgan MR (2015) Emerging properties of adhesion complexes: what are they and what

do they do? Trends Cell Biol 25(7):388–397. https://doi.org/10.1016/j.tcb.2015.02.008 6. Kubow KE, Conrad SK, Horwitz AR (2013) Matrix microarchitecture and myosin II determine adhesion in 3D matrices. Curr Biol 23 (17):1607–1619. https://doi.org/10.1016/j. cub.2013.06.053 7. Kubow KE, Horwitz AR (2011) Reducing background fluorescence reveals adhesions in 3D matrices. Nat Cell Biol 13(1):3–5. author reply 5–7. https://doi.org/10.1038/ ncb0111-3 8. Fraley SI, Feng Y, Krishnamurthy R, Kim DH, Celedon A, Longmore GD, Wirtz D (2010) A distinctive role for focal adhesion proteins in three-dimensional cell motility. Nat Cell Biol 12(6):598–604. https://doi.org/10.1038/ ncb2062 9. Kim DH, Wirtz D (2013) Focal adhesion size uniquely predicts cell migration. FASEB J 27 (4):1351–1361. https://doi.org/10.1096/fj. 12-220160. fj.12-220160 [pii]

Focal Adhesion Parametrization 10. Fokkelman M, Balcioglu HE, Klip JE, Yan K, Verbeek FJ, Danen EH, van de Water B (2016) Cellular adhesome screen identifies critical modulators of focal adhesion dynamics, cellular traction forces and cell migration behaviour. Sci Rep 6:31707. https://doi.org/10.1038/ srep31707 11. Berginski ME, Vitriol EA, Hahn KM, Gomez SM (2011) High-resolution quantification of focal adhesion spatiotemporal dynamics in

37

living cells. PLoS One 6(7):e22025. https:// doi.org/10.1371/journal.pone.0022025 12. Akiyama SK (2013) Purification of fibronectin. Curr Protoc Cell Biol 60:Unit 10.15. https:// doi.org/10.1002/0471143030.cb1005s60 13. Wurm CA, Neumann D, Schmidt R, Egner A, Jakobs S (2010) Sample preparation for STED microscopy. Methods Mol Biol 591:185–199. https://doi.org/10.1007/978-1-60761-4043_11

Chapter 4 Focal Adhesion Isolation Assay Using ECM-Coated Magnetic Beads Wesley Sturgess and Vinay Swaminathan Abstract Focal adhesions are force sensitive structures that dynamically alter their composition, protein-protein interactions, and signaling in response to external mechanical stimuli. These dynamic changes are critical for focal adhesion function and are required for cellular mechanosensing. Here, we describe a simple protocol that allows for isolation of the focal adhesion complex from adherent cells in culture in response to different mechanical stimuli applied at adhesion sites. By combining this assay with approaches such as proteomics or western blot analysis, one can study the force-dependent changes in focal adhesion composition, protein–protein interactions and signaling. Key words Focal adhesion composition, Magnetics, Mechanotransduction

1

Introduction Focal adhesions (FAs) are multimolecular complexes that mediate the cellular response to external mechanical cues, for example the stiffness of the extracellular matrix and physical forces like blood flow and tissue strain. Cellular responses, which include cell migration, differentiation, transcription, and many others, occur downstream of molecular level changes at FA sites that are central to the mechanisms of sensing, transmission, and activation of signaling by the external physical cues [1–3]. These changes at the molecular level include dynamic modulation of FA protein composition, protein conformation, protein–protein interactions, and post-translational modifications leading to activation or inhibition [4]. The multimolecular nature of FAs is central to its ability to be the primary conduit of physical information in a cell. Recent studies have now identified more than 5000 different proteins (called the adhesome) that can localize to FA sites. The function of many of these proteins are still unknown [5–7]. Examples of proteins include cell surface receptors, kinases, phosphatases, GTPases, and many others [8].

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_4, © Springer Science+Business Media, LLC, part of Springer Nature 2021

39

40

Wesley Sturgess and Vinay Swaminathan

To comprehensively understand the mechanisms and role of FAs as the primary cellular sensors of physical cues, it is critical to decipher the identities and force-sensitive properties of its constituent proteins. This requires a robust assay that will allow for application of precise physical forces on adhesion sites and the isolation of the constituents under different conditions for biochemical analysis. Here we present one such protocol that is based on a technique initially developed to isolate adhesion complexes in nonadherent cells for mass-spectrometry analysis [9]. Those experiments originally involved the use of magnetic beads coated with an extracellular matrix protein which we extended to include the application of precise physical forces on FAs via an external magnet. This alteration allowed for the identification of force-sensitive RhoA-GEFs in focal adhesions [10, 11]. Subsequently, this technique has now been used to study a number of force dependent changes in a cell including the integrin pathway, the E-cadherin pathway and even nuclear mechanotransduction pathways [12–14]. By altering the coating on the magnetic beads, this technique can be used to investigate changes in protein composition and signaling of focal adhesions or any other multimolecular complexes.

2 2.1

Materials Cell Culture

1. Mouse embryonic fibroblasts (MEFs). Other cells can be used. Refer to their culture guidelines for specifics. 2. Dulbecco’s modified eagle’s medium (DMEM + GlutaMAX, Gibco), supplemented with 10% fetal bovine serum, +1% penicillin/streptomycin. Cells were kept at 37  C and 5% CO2.

2.2 Bead Functionalization and Coating

1. Tosyl-activated Dynabeads (2.8 or 4.5 μm) (Invitrogen, 14203, 14013). 2. 0.1 M sodium phosphate buffer, pH 7.4 (buffer 1). 3. Ca2+- and Mg2+-free Dulbecco’s phosphate buffered saline (PBS). 4. PBS supplemented with 0.1% bovine serum albumin (BSA) and 2 mM EDTA, pH 7.4 (buffer 2). 5. ECM ligand. In this example, we use fibronectin (SigmaAldrich), but fibronectin can be replaced with another ligand of choice, for example collagen or ligand peptides such as RGD. 6. Magnetic separator, for example DynaMag-2.

2.3 Bead/Cell Incubation and Force Application

1. Serum-free medium. 2. Vortexer and sonicator.

Focal Adhesion Isolation

41

3. Neodymium ring magnets (N42 grade, OD 26.75 mm, ID 16 mm, height 5 mm, approx. strength 11 kg, supermagnete) (see Note 1). 4. DTBP cross-linker (Thermo Fisher). 5. Tris–HCl, pH 8.5. 2.4 Focal Adhesion Isolation

1. Lysis buffer (25 mM Tris–HCl, 0.13 M NaCl, 2 mM MgCl2, 0.1% NP-40 supplemented with protease and phosphatase inhibitor). 2. NP-40. 3. Protease and phosphatase inhibitor, EDTA-free. 4. Benzonase (Sigma-Aldrich).

3

Methods

3.1 Bead Functionalization

1. Vortex bead vial for approximately 45 s before resuspending and transferring the desired volume of beads to a 1.5 ml conical tube. For this description, we functionalize 1 ml (approximately 4  108) beads that can be used in multiple experiments. 2. Place the tube on the magnetic separator for 1 min and remove the supernatant. 3. Remove from magnet and resuspend in 1 ml buffer 1. 4. Repeat steps 2 and 3. 5. Place washed and resuspended beads on the magnet separator and remove supernatant. 6. Remove tube from magnet separator and resuspend in 1 ml buffer 1 containing 200 μg/ml ligand. 7. Incubate for 15 min and add 1 ml 0.1% BSA. 8. Incubate for 16–24 h at 37  C gently rotating the tube during the incubation process. Use the smallest tube possible to avoid the beads drying out. Dried out beads tend to aggregate, making them difficult to separate and use. 9. Place tube on magnetic separator and remove supernatant. 10. Remove tube and resuspend with 1 ml buffer 2. Incubate for 5 min at 4–8  C under gentle agitation (tilting or rotating). 11. Place tube on magnetic separator and remove supernatant. 12. Repeat steps 10 and 11 once. 13. Remove tube and resuspend in 1 ml buffer 2. Store at 4  C. 14. Final concentration should be 4  108 beads per ml (see Note 2).

42

Wesley Sturgess and Vinay Swaminathan

3.2 Cell/Bead Incubation

1. Grow cells in 10 cm dish to 80–90% confluency. 2. Aspirate medium and wash three times with prewarmed PBS. 3. Aspirate PBS and add 2 ml prewarmed serum-free medium (SFM). Incubate for 30 min at 37  C 5% CO2. 4. To prepare beads, vortex and sonicate well. Suspend approximately 2 beads per cell in 1 ml prewarmed SFM. Some ligands will aggregate more than others and will thus require more vigorous separation. Aggregation can be checked in a cell counter prior to adding to cell dish or on a microscope after adding beads to dish. 5. Add bead/SFM solution to dish, gently shake the dish to disperse beads evenly. Incubate for 40 min at 37  C 5% CO2.

3.3

Magnetics

1. Remove lid from cell culture dish and apply magnetic force on the beads by placing a new lid with magnets attached on the lid. Configuration and distance of magnets from cells should be carefully considered. For this protocol we used two layers of four ring magnets held on to dish lids with smaller 19 mm magnets to give a distance of 7 mm from the bottom of the dish. 2. Apply force for the desired time at 37  C. A separate nonsterile incubator is recommended to avoid contamination to sterile cell lab incubators. 3. Remove magnets and replace lid. The following steps are time sensitive and should be done as fast as possible (see Note 3).

3.4 Cross-Linking (Optional)

Given the dynamic nature of focal adhesions, cross-linking the cells before lysing may be advisable. 1. Add 500 μl DTBP cross-linker in SFM to the dish and incubate for 10 min. Mix DTBP so that the final concentration is 3 mM in 3.5 ml SFM. 2. Quench cross-linker with 20 mM Tris–HCl, pH 8.5 for 5 min.

3.5 Cell Lysis and Bead Separation

For the following steps, work on ice. 1. Aspirate medium; hold dish on its side for >30 s to remove as much of the medium as possible. 2. Add 500 μl lysis buffer to dish. 3. Use a cell scraper to lyse cells and move to an Eppendorf tube. Do not leave cells in lysis buffer for more than a minute. 4. Add 0.2 1 μl/ml Benzonase and incubate for 10 min. Beads can get stuck in DNA/RNA making them difficult to isolate. 5. Place tube on magnetic separator for 1 min and remove lysate. 6. Centrifuge lysate for at 10,000  g, 4  C for 10 min.

Focal Adhesion Isolation

43

7. Transfer supernatant from spun-down lysate to a new tube and put on ice for later analysis. 8. Wash beads by removing tube from magnet and resuspending in 500–1000 μl lysis buffer. 9. Place tube on magnetic separator for 1 min and remove supernatant. 10. Repeat steps 8 and 9 twice. 11. Beads should be resuspended in PBS; the volume depends upon your experiment. 12. For confirmation of FA activity, phosphorylation of focal adhesion kinase (FAK) or RhoA GTPase activity or activation of other FA proteins can be assayed. This is highly recommended during optimization of the assay. It is also recommended to Coomassie or Silver stain to ensure protein isolation. We also verify isolation by blotting for FA proteins paxillin and vinculin. 13. For mass spectrometry experiments, beads should be washed three times with PBS and then resuspended in PBS.

4

Notes 1. Neodymium magnets are very strong, so caution is advised when using them, read manufacturer’s instructions for details. 2. Some bead loss is to be expected in the washing and coating processes, so recounting after completing the process is highly recommended. 3. Ring magnets are ideal for uniform magnetic field gradients across the cell culture plate. The magnets can also be attached to a motorized holder to apply dynamic forces on the beads.

References 1. Discher DE, Janmey P, Wang YL (2005) Tissue cells feel and respond to the stiffness of their substrate. Science 310(5751):1139–1143. https://doi.org/10.1126/science.1116995 2. Dupont S, Morsut L, Aragona M, Enzo E, Giulitti S, Cordenonsi M, Zanconato F, Le Digabel J, Forcato M, Bicciato S, Elvassore N, Piccolo S (2011) Role of YAP/TAZ in mechanotransduction. Nature 474 (7350):179–183. https://doi.org/10.1038/ nature10137 3. Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, Parsons JT, Horwitz AR (2003) Cell migration: integrating signals from front to back. Science 302

(5651):1704–1709. https://doi.org/10. 1126/science.1092053 4. Hoffman BD, Grashoff C, Schwartz MA (2011) Dynamic molecular processes mediate cellular mechanotransduction. Nature 475 (7356):316–323. https://doi.org/10.1038/ nature10316 5. Humphries JD, Paul NR, Humphries MJ, Morgan MR (2015) Emerging properties of adhesion complexes: what are they and what do they do? Trends Cell Biol 25(7):388–397. https://doi.org/10.1016/j.tcb.2015.02.008 6. Kuo JC, Han X, Hsiao CT, Yates JR 3rd, Waterman CM (2011) Analysis of the myosinII-responsive focal adhesion proteome reveals a

44

Wesley Sturgess and Vinay Swaminathan

role for beta-Pix in negative regulation of focal adhesion maturation. Nat Cell Biol 13 (4):383–393. https://doi.org/10.1038/ ncb2216 7. Schiller HB, Fassler R (2013) Mechanosensitivity and compositional dynamics of cellmatrix adhesions. EMBO Rep 14 (6):509–519. https://doi.org/10.1038/ embor.2013.49 8. Horton ER, Byron A, Askari JA, Ng DHJ, Millon-Fremillon A, Robertson J, Koper EJ, Paul NR, Warwood S, Knight D, Humphries JD, Humphries MJ (2015) Definition of a consensus integrin adhesome and its dynamics during adhesion complex assembly and disassembly. Nat Cell Biol 17 (12):1577–1587. https://doi.org/10.1038/ ncb3257 9. Byron A Humphries JD Bass MD Knight D Humphries MJ (2011) Proteomic analysis of integrin adhesion complexes. In Science Signaling. https://doi.org/10.1126/scisignal. 2001827 10. Guilluy C, Swaminathan V, Garcia-Mata R, O’Brien ET, Superfine R, Burridge K (2011) The Rho GEFs LARG and GEF-H1 regulate the mechanical response to force on integrins.

Nat Cell Biol 13(6):722–727. https://doi. org/10.1038/ncb2254 11. Marjoram RJ, Guilluy C, Burridge K (2016) Using magnets and magnetic beads to dissect signaling pathways activated by mechanical tension applied to cells. Methods 94:19–26. https://doi.org/10.1016/j.ymeth.2015.09. 025 12. Bays JL, Campbell HK, Heidema C, Sebbagh M, DeMali KA (2017) Linking E-cadherin mechanotransduction to cell metabolism through force-mediated activation of AMPK. Nat Cell Biol 19(6):724–731. https://doi.org/10.1038/ncb3537 13. Collins C, Guilluy C, Welch C, O’Brien ET, Hahn K, Superfine R, Burridge K, Tzima E (2012) Localized tensional forces on PECAM-1 elicit a global mechanotransduction response via the integrin-RhoA pathway. Curr Biol 22(22):2087–2094. https://doi.org/10. 1016/j.cub.2012.08.051 14. Guilluy C, Osborne LD, Van Landeghem L, Sharek L, Superfine R, Garcia-Mata R, Burridge K (2014) Isolated nuclei adapt to force and reveal a mechanotransduction pathway in the nucleus. Nat Cell Biol 16(4):376–381. https://doi.org/10.1038/ncb2927

Part II Proximity and Microscopy-Based Methods to Determine Integrin Interactions

Chapter 5 Functional Integrin Regulation Through Interactions with Tetraspanin CD9 A´lvaro Torres-Go´mez, Beatriz Carden˜es, Ester Dı´ez-Sainz, Esther M. Lafuente, and Carlos Caban˜as Abstract Integrins are adhesion receptors that mediate many intercellular and cell–extracellular matrix interactions with relevance in physiology and pathology. Unlike other cellular receptors, integrins critically require activation for ligand binding. Through interaction in cis with other molecules and the formation of tetraspanin-enriched membrane microdomains (TEMs), the tetraspanin CD9 regulates integrin activity and avidity. Here we present three techniques used to study CD9–integrin interactions and integrin activation. Key words Integrin, Tetraspanin, CD9, Proximity-ligation assay (PLA), Coimmunoprecipitation, Pull-down assay

1

Introduction Tetraspanins are an evolutionary conserved family of small transmembrane proteins present on the plasma membrane and intracellular organelles. Tetraspanins contain four transmembrane domains, with short N- and C-terminal regions both located intracellularly. These delimit two extracellular domains of unequal size, termed SEL (short extracellular loop) or EC1 (extracellular domain 1) and LEL (large extracellular loop) or EC2 (extracellular domain 2), respectively [1–4]. The constant region of the LEL domain is involved in tetraspanin dimerization and oligomerization whilst the variable region dictates lateral interactions with other membrane proteins [4]. These interactions allow the formation of membrane microdomains called Tetraspanin-Enriched Microdomains or TEMs [1, 3]. TEMs compartmentalize signaling of many receptors (tyrosine kinase receptors, G-protein-coupled receptors, MHC, BCR, etc.) and adhesion molecules, with members of the integrin family being prominently found in TEMs [3, 5, 6]. Tetraspanins

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_5, © Springer Science+Business Media, LLC, part of Springer Nature 2021

47

48

A´lvaro Torres-Go´mez et al.

have been demonstrated to regulate clustering of integrins and also have been shown to colocalize with active signaling effectors such as the lipid kinase PI4K, Src family tyrosine kinase Lyn, the Ser/Thr kinase PKC, and metalloproteinases of the ADAM family [1, 3, 7, 8], all of which are capable of regulating integrin activity and function. Tetraspanin CD9 is emerging as a key regulatory molecule of the adhesive function of different integrins [6], through either direct cis interaction with integrins at the cell surface, as has been reported for integrin LFA-1 (αLβ2) [2], or through additional partners such as ADAM17 in the case of integrin α5β1 [9]. CD9 seems to regulate integrin-mediated cell adhesion without altering integrin affinity [2, 9]. Instead, CD9 modulates integrin avidity through regulating integrin inclusion in TEMs and their organization on the cell surface. Here we describe the key methods to study CD9-mediated regulation of integrin functions: coimmunoprecipitation (Co-IP), covalent chemical cross-linking and pull-down assays using CD9 LEL constructs, which determine the composition of protein complexes, and proximity-ligation assays (PLA), which test for “direct” protein–protein interactions (
40 nm apart).

2 2.1

Materials Cell Cultures

1. Peripheral blood mononuclear cells (PBMCs) obtained from density gradient (Biocoll, Biochrom) purification of buffycoats of healthy donors, as described previously [10]. 2. Other cell lines can be used, for example HSB-2 and Jurkat (T cells), JY and Daudi (B cells) and THP-1 and U937 cells (monocytic cells). 3. RPMI-1640 medium supplemented with 10% FBS, 50 U/ml penicillin–streptomycin, and 2 mM glutamine (Lonza). 4. PMA (Sigma-Aldrich), final concentration 20 ng/ml, used to differentiate THP-1 and U937 cells. 5. 100 mm dishes. 6. Poly-L-lysine. 7. 35 mm dishes with 12 mm glass coverslips (MatTek).

2.2 Buffers and Reagents

1. PBS: 137 mM NaCl, 2.7 mM KCI, 8 mM Na2HPO4, 2 mM KH2PO4. 2. Biotin conjugation kit: EZ-LinkR Sulfo-NHS-LC-Biotin (Thermo Scientific).

Assays for CD9-Mediated Integrin Regulation

49

3. Biotinylation buffer: 1 mM CaCl2, 1 mM MgCl2, containing 1 mM EZ-LinkR Sulfo-NHS-LC-Biotin (Thermo Scientific) in PBS. 4. Tris Buffered Saline (TBS): 50 mM Tris–HCl pH 7.5, 50 mM NaCl. 5. Co-IP buffer: 1 mM CaCl2, 1 mM MgCl2 in TBS. 6. Cell lysis buffer: 1% Detergent, 1 mM CaCl2, 1 mM MgCl2, and Protease inhibitor cocktail (Sigma-Aldrich) in TBS (see Note 1). 7. Laemmli loading buffer 5 (nonreducing): 0.25% Bromophenol blue, 50% Glycerol, 10% SDS, Tris–HCl 0.25 M, pH 6.8. 8. Electrophoresis buffer: 25 mM Tris–HCl, pH 8.3, 190 mM glycine, 0.1% (v/v) SDS. 9. Transference buffer: 25 mM Tris–HCl, pH 8.3, 190 mM glycine, 20% (v/v) methanol. 10. Standard SDS-PAGE buffers and protein molecular weight marker. 11. Western-blot washing solution: 0.05% Tween 20 in TBS. 12. Western-blot blocking solution: 0.05% Tween 20 and 3% (w/v) bovine serum albumin (BSA) in TBS. 13. Protein concentration assessment: DC protein® kit (Bio-Rad Laboratories). 14. Proximity ligation assay: Duolink® PLA kit (Sigma-Aldrich). 15. Co-IP resin: G-Sepharose (Sigma-Aldrich). 16. Pull-down resin: GSH-agarose (Sigma-Aldrich). 17. FcR-Blocking buffer: human gammaglobulin (Sigma-Aldrich) (see Note 2). 18. Crosslinking reagent: 3,30 -dithiobis(sulfosuccinimidylpropionate; DTSSP (ThermoFisher Scientific). 2.3

Antibodies

1. Anti-CD9: Mouse monoclonal clones VJ1/20, PAINS10 and PAINS13 (purified from hybridoma cultures) [11], rabbit polyclonal H110 (Santa Cruz Biotechnology). 2. Anti-integrin β2: Mouse monoclonal anti-β2 clones Lia3/ 2 (purified from hybridoma cultures) [12], TS1/18 (BioLegend), MEM-48 (ImmunoTools). 3. Anti-integrin αM: Mouse monoclonal Bear-1 (purified from hybridoma cultures) [13]. 4. Anti-integrin αX: Mouse monoclonal HC1/1 (purified from hybridoma cultures) [14]. 5. Anti-integrin αL: Mouse monoclonal TS1/11 (purified from hybridoma cultures) [15].

50

A´lvaro Torres-Go´mez et al.

6. Activity reporting anti-integrin β2: Mouse monoclonal KIM185 (UCB-Celltech), and m24 (BioLegend). 7. Anti-CD147: Mouse monoclonal VJ1/9 (purified from hybridoma cultures) [11]. 8. Anti-CD59: Monoclonal VJ1/12 (purified from hybridoma cultures) [11]. 9. Anti-GST: Rabbit polyclonal Z-5 (Santa Cruz Biotechnology). 2.4 Instruments and Software

1. Electrophoresis cuvette and other standard Western-blot material. 2. Rotator. 3. Inverted confocal microscope (Leica LSM510). 4. ImageJ image analysis package.

3

Methods

3.1 Specific Coimmunoprecipitation of CD9 and Integrin Molecules from the Cell Surface

Coimmunoprecipitation experiments were performed using intact cells during incubation with relevant antibodies in order to detect protein–protein interactions at the cell surface. 1. Harvest 107 cells and wash three times thoroughly with PBS by centrifuging at 300  g for 5 min. Reserve two aliquots of cells to use as the whole lysate and lysate without antibody (added to the G-Sepharose) to load into gel as controls for each condition. 2. Incubate intact cells for 60 min at 4  C with anti-β2 integrin (TS1/18), anti-CD9 (VJ1/20) or control anti-CD59 antibodies (20 μg/ml) in the Co-IP buffer containing 100 μg/ml human gamma-globulin (see Notes 2 and 3). 3. Wash excess antibody thrice with Co-IP buffer. 4. Lyse cells in lysis buffer for 15 min at 4  C. Spin lysate at 10,000  g 10 min at 4  C. Isolate the supernatant and discard the precipitate. 5. Wash once G-Sepharose beads (40 μl) with Co-IP buffer to equilibrate beads (see Note 4). 6. Incubate supernatant with G-Sepharose beads overnight at 4  C on the rotator. 7. Wash beads thrice with diluted (1/5) Lysis buffer by centrifuging at 2000  g, 1 min at 4  C (see Note 5). 8. Boil (100  C, 10 min) G-Sepharose beads in one volume of nonreducing 2 Laemmli-loading buffer. Samples are now ready to load into cast 8% (for β2 integrin detection) or 12% (for CD9) SDS-PAGE gels.

Assays for CD9-Mediated Integrin Regulation

51

9. Run electrophoresis using at 100 V constant voltage until the bromophenol blue front reaches the bottom of the gel (approximately 1.5 h). 10. Electrotransfer the gels following standard Western blot procedures (300 mA constant current 2 h) onto nitrocellulose membranes (see Notes 6 and 7). 11. Wash membrane to eliminate excess methanol using the Western-blot washing solution and block membranes using the Western-blot blocking solution in constant agitation 1 h at room temperature. 12. Wash the membrane and incubate it with 0.5 μg/ml of biotinconjugated anti-β2 (MEM-48) or anti-CD9 (VJ1/20) antibodies (0.5 μg/ml) overnight at 4  C. Wash thrice and develop with streptavidin-HRP for 1 h (Thermo scientific) and image using luminol-based chemiluminescence (5 min incubation with ECL followed by approximately 1 min of exposure). Quantify image using appropriate software, for example ImageJ (NIH). 3.2 Covalent Chemical Cross-Linking

DTSSP is a water-soluble and membrane impermeable cross-linker that reacts with primary amines (in accessible cell-surface proteins) and is useful in detecting weak or transient protein–protein interactions of cell surface proteins. DTSSP can be used in conjunction with the above mentioned Co-IP protocol. 1. Prior to adding immunoprecipitating antibodies, treat 2  107 cells with 0.25 mM DTSSP for 30 min at 4  C in Co-IP buffer. Set aside a non–cross-linked aliquot to use as a control (see Note 8). 2. Stop reaction by incubating with 10 mM glycine, pH 7.4 for 15 min at room temperature. 3. Wash cells three times and continue with Co-IP protocol by lysing cells.

3.3 Pull-Down Assays

Before the assay, GST fusion proteins containing the LEL region from human wild type CD9, CD81, and CD63 were purified from BL21 codon-plus Escherichia coli transfected with the appropriate cDNA cloned into the pGEX-KG expression plasmid, as described elsewhere [16–18] (see Note 9). 1. Harvest 106 cells and wash three times thoroughly with PBS with 1 mM CaCl2, 1 mM MgCl2 by centrifuging at 300  g for 5 min. 2. Label surface proteins using the Biotin labeling buffer (1 mg/ ml) for 30 min at 4  C.

52

A´lvaro Torres-Go´mez et al.

3. Stop reaction using 100 mM glycine to remove excess biotin (10 min 4  C), wash cells with PBS. Reserve aliquots of nonbiotinylated cells for separate analysis. 4. Lyse cells as per the Co-IP protocol and reserve aliquots for use as whole lysate. 5. Incubate lysates overnight at 4  C with GST-fusion proteins (15 μg). 6. Pull-down with glutathione-agarose for 3 h at 4  C in a rotator (see Note 10). 7. Wash thrice with diluted lysis buffer (1/10) to eliminate nonspecific binding (see Note 5). 8. Boil (100  C, 10 min) in nonreducing 2 Laemmli buffer (25 μl). Proceed with Western-blot as per the Co-IP protocol. 9. Stain blots with the appropriate method. (a) For nonbiotinylated cells, stain pulled-down complexes using biotinylated anti-β2 integrin MEM-48 antibody (1 μg/ml) and with a streptavidin-HRP (0.2 μg/ml). Assess using ECL-chemiluminescence. Use an anti-GST rabbit polyclonal antibody (1 μg/ml) as a loading control. (b) For surface-biotinylated cells, directly detect pulled down proteins with streptavidin-HRP (0.2 μg/ml) and ECL-chemiluminescence. 3.4 Proximity Ligation Assay

In situ proximity ligation assays allow for detection of direct (
40 nm apart) protein–protein interactions in cell samples by fluorescence microscopy. The technique requires the two primary antibodies to be from different species (e.g., Mouse/Rabbit). Secondary oligonucleotide-labeled antibodies (anti-mouse-plus probe and anti-rabbit minus probe) will hybridize. These are then ligated using T4 DNA Ligase. DNA is amplified in a rolling-circle amplification reaction using DNA polymerase and detection is performed through the use of fluorescently labeled oligonucleotide probes (Fig. 1). 1. Harvest 106 cells and wash thrice thoroughly with PBS containing 1 mM CaCl2, 1 mM MgCl2 by centrifuging at 300  g for 5 min (see Note 11). 2. Seed cells onto poly-L-lysine treated microscopy slides (106 cells per slide). Allow adhesion for 1 h at 37  C. Discard the medium. 3. Fix cells with 4% paraformaldehyde for 8 min, as per standard microscopy protocols. Wash thoroughly with TBS. 4. Block Fc Receptors with human gammaglobulin 100 μg/ml for 1 h at 4  C to avoid nonspecific binding of primary antibodies and probes. Wash with PBS.

Assays for CD9-Mediated Integrin Regulation

53

Fig. 1 An example of PLA between CD9 and LFA-1. (a) In Situ Proximity Ligation Assay (PLA) of the molecular association of CD9 with CD147 (negative control) and of CD9 with integrin LFA-1 (positive control) on THP1/ PMA macrophage-like cells. (b) The graph represents the quantification PLA signal from panel a (mean of PLA/dots from different micrographic fields analyzed by unpaired t test. ****p < 0.0001)

5. Incubate samples with mouse primary antibodies (antiβ2 TS1/18, anti-αL TS1/11, anti-CD147 VJ1/9, or antiCD81 5A6) and the anti-CD9 H-110 rabbit primary polyclonal antibody overnight at 4  C (10 μg/ml). 6. Wash dishes twice with room temperature wash Buffer A (Duolink® PLA kit (Sigma-Aldrich) and add the specific oligonucleotide-labeled secondary antibodies (plus and minus probes) (5 μg/ml). Incubate 1 h at 37  C. 7. Wash twice with Buffer A and proceed with DNA-ligation and amplification by incubating with DNA Ligase 1:40 for 30 min at 37  C. 8. Wash twice with Buffer A and incubate with the DNA polymerase 1:80 for 100 min at 37  C. 9. Stop the reaction by washing three times with room temperature Buffer B. 10. Dry slides carefully and mount samples with ProLong® antifade mounting reagent. Avoid trapping air bubbles. Store slides overnight at 4  C and proceed to imaging.

54

A´lvaro Torres-Go´mez et al.

11. Acquire PLA signals using an inverted confocal microscope and analyze images using Fiji Software and quantify dot fluorescence intensity and size (see Note 12).

4

Notes 1. The stability of tetraspanin interactions can be assessed by using 1% of increasingly stronger detergents. We recommend using CHAPS (mild), Brij-97 (intermediate), and NP-40 substitute or Triton X-100 (harsh) for these determinations [2, 19]. 2. Fc Receptor blocking buffer should be freshly prepared each time. 3. To activate integrins, wash cells thrice with 200 μM MnCl2 in TBS and incubate in the same buffer for 10 min. Maintain Mn2 + concentration throughout the assay. 4. When working with Sepharose beads it is advisable to use blue 1 ml pipette tips. These tips should be cut at the narrow end using scissors. Use these tips to resuspend the beads (do not vortex) and to easily pipette the required amount. Alternatively, wide bore tips can be used. 5. When washing, it is important to carefully discard the dilute lysis buffer without pipetting the agarose. We find that leaving a few microliters of the buffer during this stage is acceptable. 6. When performing the electrotransfer it is critical to ensure that all air bubbles between membrane and gel are removed. 7. We find that although less easy to handle than PVDF membranes, nitrocellulose membranes are more suitable and give a lower background signal. 8. DTSSP is moisture-sensitive and needs to be stored desiccated at 4  C. Cross-linkers cannot be stored after reconstitution and need to be prepared immediately before use. When performing the crosslinking assay it is critical to avoid buffers containing primary amines (Tris, and cell media). Therefore, it is necessary to wash thoroughly to eliminate amine compounds that might interfere in the assay. 9. Constructs remain stable for a maximum of 2 months when stored at 80  C. 10. Glutathione reduction state is pH-dependent. Ensure that buffers are adjusted at pH 7.4. 11. We recommend using at least two noninteracting proteins as a negative control. 12. Confocal image acquisition parameters (laser intensity, scan speed, gain, pinhole, etc.) vary and need to be optimized

Assays for CD9-Mediated Integrin Regulation

55

empirically, but then they need to remain constant for all images. The Nyquist criterion should be used to determine the thickness of the confocal slices to prevent oversampling or undersampling. Image analysis also needs to be optimized empirically, maintaining the following criteria in mind: (1) background must appear as dark as possible and signal must be uniform. If the instrument features hi/lo thresholding, it facilitates determining this empirically. Adjustment of the background and signal must be performed using both control and sample images to ensure consistency; (2) adjustments to brightness and contrast must be applied uniformly across all acquired images and only for determining threshold levels in order to extract the regions of interest (PLA dots and cells); (3) only count PLA dots of size larger than 0.3 μm2 when using the analyze particle tool in ImageJ; (4) when quantifying fluorescence intensity for the region of interest, make sure that the image has not been modified in any way. References 1. Charrin S, le Naour F, Silvie O, Milhiet PE, Boucheix C, Rubinstein E (2009) Lateral organization of membrane proteins: tetraspanins spin their web. Biochem J 420(2):133–154 2. Reyes R, Monjas A, Yanez-Mo M, Cardenes B, Morlino G, Gilsanz A, Machado-Pineda Y, Lafuente E, Monk P, Sanchez-Madrid F, Cabanas C (2015) Different states of integrin LFA-1 aggregation are controlled through its association with tetraspanin CD9. Biochim Biophys Acta 1853(10 Pt A):2464–2480. https://doi. org/10.1016/j.bbamcr.2015.05.018 3. Yanez-Mo M, Barreiro O, Gordon-Alonso M, Sala-Valdes M, Sanchez-Madrid F (2009) Tetraspanin-enriched microdomains: a functional unit in cell plasma membranes. Trends Cell Biol 19(9):434–446. https://doi.org/10. 1016/j.tcb.2009.06.004 4. Stipp CS, Kolesnikova TV, Hemler ME (2003) Functional domains in tetraspanin proteins. Trends Biochem Sci 28(2):106–112 5. Cabanas C, Yanez-Mo M, van Spriel AB (2019) Editorial: functional relevance of tetraspanins in the immune system. Front Immunol 10:1714. https://doi.org/10.3389/fimmu. 2019.01714 6. Reyes R, Cardenes B, Machado-Pineda Y, Cabanas C (2018) Tetraspanin CD9: a key regulator of cell adhesion in the immune system. Front Immunol 9:863. https://doi.org/ 10.3389/fimmu.2018.00863 7. Tarrant JM, Robb L, van Spriel AB, Wright MD (2003) Tetraspanins: molecular organisers

of the leukocyte surface. Trends Immunol 24 (11):610–617 8. Yanez-Mo M, Gutierrez-Lopez MD, Cabanas C (2011) Functional interplay between tetraspanins and proteases. Cell Mol Life Sci 68 (20):3323–3335. https://doi.org/10.1007/ s00018-011-0746-y 9. Machado-Pineda Y, Cardenes B, Reyes R, Lopez-Martin S, Toribio V, SanchezOrganero P, Suarez H, Grotzinger J, Lorenzen I, Yanez-Mo M, Cabanas C (2018) CD9 controls integrin alpha5beta1-mediated cell adhesion by modulating its association with the metalloproteinase ADAM17. Front Immunol 9:2474. https://doi.org/10.3389/ fimmu.2018.02474 10. Rodriguez-Fernandez JL, Sanchez-Martin L, Rey M, Vicente-Manzanares M, Narumiya S, Teixido J, Sanchez-Madrid F, Cabanas C (2001) Rho and Rho-associated kinase modulate the tyrosine kinase PYK2 in T-cells through regulation of the activity of the integrin LFA-1. J Biol Chem 276(44):40518–40527 11. Yanez-Mo M, Alfranca A, Cabanas C, Marazuela M, Tejedor R, Ursa MA, Ashman LK, de Landazuri MO, Sanchez-Madrid F (1998) Regulation of endothelial cell motility by complexes of tetraspan molecules CD81/ TAPA-1 and CD151/PETA-3 with alpha3 beta1 integrin localized at endothelial lateral junctions. J Cell Biol 141(3):791–804 12. Campanero MR, del Pozo MA, Arroyo AG, Sanchez-Mateos P, Hernandez-Caselles T,

56

A´lvaro Torres-Go´mez et al.

Craig A, Pulido R, Sanchez-Madrid F (1993) ICAM-3 interacts with LFA-1 and regulates the LFA-1/ICAM-1 cell adhesion pathway. J Cell Biol 123(4):1007–1016 13. Keizer GD, Borst J, Figdor CG, Spits H, Miedema F, Terhorst C, De Vries JE (1985) Biochemical and functional characteristics of the human leukocyte membrane antigen family LFA-1, Mo-1 and p150,95. Eur J Immunol 15 (11):1142–1148 14. Cabanas C, Sanchez-Madrid F, Acevedo A, Bellon T, Fernandez JM, Larraga V, Bernabeu C (1988) Characterization of a CD11c-reactive monoclonal antibody (HC1/1) obtained by immunizing with phorbol ester differentiated U937 cells. Hybridoma 7(2):167–176 15. Sanchez-Madrid F, Nagy JA, Robbins E, Simon P, Springer TA (1983) A human leukocyte differentiation antigen family with distinct alpha-subunits and a common beta-subunit: the lymphocyte function-associated antigen (LFA-1), the C3bi complement receptor (OKM1/Mac-1), and the p150,95 molecule. J Exp Med 158(6):1785–1803 16. Barreiro O, Yanez-Mo M, Sala-Valdes M, Gutierrez-Lopez MD, Ovalle S,

Higginbottom A, Monk PN, Cabanas C, Sanchez-Madrid F (2005) Endothelial tetraspanin microdomains regulate leukocyte firm adhesion during extravasation. Blood 105 (7):2852–2861 17. Gilsanz A, Sanchez-Martin L, Gutierrez-Lopez MD, Ovalle S, Machado-Pineda Y, Reyes R, Swart GW, Figdor CG, Lafuente EM, Cabanas C (2013) ALCAM/CD166 adhesive function is regulated by the tetraspanin CD9. Cell Mol Life Sci 70(3):475–493. https://doi.org/10. 1007/s00018-012-1132-0 18. Higginbottom A, Takahashi Y, Bolling L, Coonrod SA, White JM, Partridge LJ, Monk PN (2003) Structural requirements for the inhibitory action of the CD9 large extracellular domain in sperm/oocyte binding and fusion. Biochem Biophys Res Commun 311 (1):208–214 19. Charrin S, Manie S, Oualid M, Billard M, Boucheix C, Rubinstein E (2002) Differential stability of tetraspanin/tetraspanin interactions: role of palmitoylation. FEBS Lett 516 (1–3):139–144. https://doi.org/10.1016/ s0014-5793(02)02522-x

Chapter 6 Proximity-Dependent Biotinylation (BioID) of Integrin Interaction Partners Satu-Marja Myllym€aki, Xiaonan Liu, Markku Varjosalo, and Aki Manninen Abstract Integrins are heterodimeric adhesion receptors that maintain cell–extracellular matrix (ECM) interactions in diverse tissue microenvironments. They mediate cell adhesion and signaling through the assembly of large cytoplasmic multiprotein complexes that focally connect with the cytoskeleton. Integrin adhesion complexes (IAC) are specialized by the type of integrin-ECM contact and are sensitive to mechanical forces. Thus, they encrypt context-dependent information about the microenvironment in their composition. Signals mediated through IACs modulate many aspects of cell behavior, which allows cells to adapt to their surroundings. To gain insights into their function, IACs have been isolated from cultured cells and explored by proteomics. IACs are insoluble by nature and held together by transient/weak interactions, which makes it challenging to isolate intact IACs. Usually all IACs coupled to a specified ECM, which may employ different integrins, are isolated. Here we describe an alternative method based on proximity-dependent biotin identification (BioID), where specific integrin interaction partners are labeled in live cells and isolated without the need to isolate intact IACs. Key words Integrin, BioID, Integrin adhesion complex, Protein–protein interaction, Cell adhesion

1

Introduction Integrin adhesion complexes (IAC) are both labile and difficult to solubilize, which makes them infeasible for affinity purification using traditional methods. Alternative methods have been developed, where IACs secured by chemical cross-linking are isolated directly on an extracellular matrix (ECM) scaffold to which cells adhere by removing the cell body [1]. These isolated IACs have been extensively characterized by proteomics, which has led to identification of thousands of proteins. Although IACs are no doubt complex, part of the observed complexity stems from copurification of different adhesion complexes bound to the ligand via different integrin or nonintegrin receptors. Therefore, there is a clear need for sorting of information between different adhesion

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_6, © Springer Science+Business Media, LLC, part of Springer Nature 2021

57

58

€ki et al. Satu-Marja Myllyma

complexes using complementary approaches such as proximitydependent biotin identification (BioID). BioID utilizes a promiscuous biotin ligase enzyme tag, which can be fused to the protein of interest and expressed in cells where it biotinylates primary amines in neighboring proteins [2]. The tag was engineered from a bacterial 35 kDa DNA-binding biotin protein ligase BirA, which selectively biotinylates proteins with a specific recognition sequence. By introducing a point mutation (R118G), BirA* was made to prematurely release a highly reactive and labile activated biotin (biotinoyl-50 -AMP), which permanently labels any protein within a 10 nm radius [2, 3]. As BirA* is restricted by the availability of free biotin in standard cell culture conditions, promiscuous labeling can be induced by addition of biotin to the culture media. BioID has recently been adopted also to study IACs [4–7]. It does not rely on a specified ECM and allows any protein to be selected as a bait. Therefore, BioID opens a possibility to examine IACs also from the integrin point of view, to understand the nature of integrin specific signals. BioID labeling takes place in native conditions, which makes it sensitive for low affinity interactions in IACs. Although labeling is restricted to a very short distance, BioID does not, in principle, discriminate between physical interactions and close encounters. Exposure to cytosolic content at random however is limited, given that mature IACs are immobilized to ECM contacts. BioID datasets using integrins and unrelated membrane proteins as baits in Madin–Darby Canine Kidney (MDCK) epithelial cells do not significantly overlap, suggesting that labeling is bait-specific [5, 8–10]. Still, proteins that physically interact with the bait need to be sorted from those that merely colocalize using affinity-based methods. Labeling of biologically relevant candidates requires a fusion protein with localization and conformation similar to the endogenous protein. To study the composition of specialized IACs, integrins are the natural choice of bait (Fig. 1a). The challenge lies in placing the tag so that it does not compromise integrin function, which is based on heterodimerization, ligand-binding and allosteric signaling through conformational changes. Fusions directly to the cytoplasmic C-terminus of integrins are generally better tolerated than to the N-terminus, given the complexity of the ectodomain structure [5, 11]. However, strategically placed intramolecular tags on the ectodomain can be successful [12, 13]. For greater conformational flexibility and labeling radius, linkers of different length can be inserted between integrin and BirA* [14]. Several protocols describing the BioID method have already been published [15– 18]. Here our purpose is to provide additional guidelines on how to apply BioID to IACs, using MDCK cells as an example.

Fig. 1 Design and execution of the BioID experiment with integrins as baits. (a) MDCK cells as an example host, express a subset of integrins, including several β1- and αV-integrins that form different focal adhesions and α6β4-integrins that forms hemidesmosomes. To study specific IACs, subunits that are not shared between different heterodimers, should be selected as baits. For example, BirA* tagged directly to the C-terminus of the β4 subunit will specifically label cytoplasmic proteins in hemidesmosomes. (b) To generate stable clonal BirA*-expressing cell lines, cells are first electroporated with a linearized plasmid encoding for the fusion construct. Forty-eight hour later, transfected cells and nontransfected control cells are selected with Geneticin over a period of 1–2 weeks. During this period, cells are trypsinized and replated several times to help eliminate the negative cells. Once negative cells have been eliminated, selected cells are trypsinized and FACS-sorted into 96-wells with conditional media. Confluent cells are split 1:3 into a fresh 96-well plate and 2:3 to a 24-well plate. The 96-well plate will be processed for western blotting to screen for positive clones. When positive clones are identified, corresponding cells in the 24-well plate can be trypsinized and expanded for further testing. Maturation of integrin α6β4-BirA* heterodimers can be confirmed by immunoprecipitation of surface labeled proteins, whereas targeting of α6β4-BirA* to hemidesmosomes can be studied with TIRF microscopy. It is important to include negative cells and cells expressing the protein without the tag as controls. Biotin ligation is performed in validated clonal BirA* expressing cell lines, including negative controls, by adding biotin to the culture media. Abbreviations: TM transmembrane domain, SS signal sequence, WB western blot, IP immunoprecipitation, LC-MS/MS liquid chromatography tandem mass spectrometry, FACS fluorescence activated cell sorting. This research was originally published in Molecular & Cellular Proteomics (Myllym€aki et al. Mol Cell Proteomics. 2019 Feb;18(2):277–293, © the Authors, ref. 5)

60

2

€ki et al. Satu-Marja Myllyma

Materials

2.1 Generation of Stable BirA* Expressing Cell Lines

1. BirA* fusion construct under the Cytomegalovirus (CMV) enhancer-promoter or other strong promoter in a pcDNA3.1 expression vector with a neomycin resistance gene for selection. 2. MluI restriction enzyme provided with NEB3.1 buffer (New England Biolabs). 3. 25:24:1 phenol–chloroform–isoamyl alcohol. 4. Minimum Essential Media (MEM) included with GlutaMAX (Thermo Fisher) and supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. 5. Trypsin–EDTA. 6. Nucleofector device (Lonza). 7. Ingenio Electroporation Kit (Mirus Bio). 8. G-418 (Geneticin) to select for the Neomycin resistance gene. 9. 10% DMSO in FBS.

2.2 Maturation of BirA Fusion Constructs

1. EZ-Link™ Sulfo-NHS-Biotin (ThermoFisher). Dissolve in PBS just before use. 2. Biotinylation buffer (20 mM HEPES, 130 mM NaCl, 5 mM KCl, 0.8 mM MgCl2 and 1 mM CaCl2 pH 7.5). 3. RIPA buffer (0.15 M NaCl, 0.5% SDS, 1% IGEPAL CA-630, 1% sodium deoxycholate, 10 mM Tris–HCl pH 7.5). 4. Benzonase nuclease (Novagen). 5. Spin-X centrifuge tube filter (Corning). 6. Protein G Dynabeads (ThermoFisher). 7. Laemmli buffer (4% SDS, 20% glycerol, 10% 2-mercaptoethanol, 0.004% bromophenol blue, and 0.125 M Tris–HCl, pH 6.8). 8. Horseradish peroxidase (HRP)-conjugated streptavidin (Jackson ImmunoResearch).

2.3 Localization of BirA Fusion Constructs to IACs

1. Biotin (Sigma Aldrich). 2. 35 mm Ø CELLview glass bottom dish (Greiner). 3. 4% paraformaldehyde (PFA) in PBS.

BioID of Integrin Interactors

61

4. 0.2 M glycine in PBS. 5. 0.1% Triton X-100 in PBS. 6. 1% BSA in PBS. 7. Alexa-conjugated ImmunoResearch).

secondary

antibodies

(Jackson

8. Alexa-conjugated streptavidin (Thermo Fisher Scientific). 2.4 Affinity Purification of Biotinylated Proteins

1. HENN buffer (50 mM NaF, 90 mM NaCl, 5 mM EDTA, 50 mM Hepes pH 8; filter-sterilize and store up to 2 months at +4  C). 2. BioID lysis buffer (0.1% SDS, 0.5% IGEPAL CA-630, 1 mM PMSF, 150 μl of 500 mM Na3VO4, 1 mM DTT, and 1 protease inhibitor mix in HENN buffer; prepare freshly before use). 3. Bio-Rad spin column. 4. IBA Strep-Tactin® beads. 5. Wash buffer (0.5% IGEPAL CA-630, 1 mM PMSF, 150 μl of 500 mM Na3VO4, 1 mM DTT, and 1 protease inhibitor mix in HENN buffer; prepare freshly before use). 6. Elution buffer (0.5 mM biotin in HENN buffer; prepare freshly before use).

3

Methods

3.1 BirA Fusion Constructs and Controls

The original BirA* is available in a pcDNA3.1 mammalian expression plasmid (Addgene #35700 and #36047) [2]. Substantially smaller biotin ligases called BioID2 and miniTurboID have been recently developed that may improve proper folding and targeting of fusion constructs [14, 19]. However, labeling conditions should be optimized individually due to differences in labeling efficiency. A small peptide tag such as Myc should be included in the design to allow detection using antibodies. In addition, antibodies against BirA have been used for detection [14]. The fusion protein can be constructed in a mammalian expression plasmid such as pcDNA3.1 and delivered to cells via stable transfection as described here (Fig. 1b). A successful fusion construct should form functional heterodimers and overlap with IAC markers at the cell surface, which can be confirmed using the following protocols. The ultimate test for the validity of the fusion construct, is to see whether it rescues IAC formation in the absence of the endogenous integrin. To resolve BirA* labeled candidate interactors from falsepositives, negative controls should account for unspecific/chance labeling by BirA*, endogenous biotinylation and unspecific binding to the affinity resin. Cells without BirA* expression can be used to

62

€ki et al. Satu-Marja Myllyma

exclude endogenously biotinylated proteins and unspecific binders from those labeled by BirA*. Proteins may become labeled without having any real connection with the bait if they are particularly abundant or bind unspecifically to the tag. A BirA* control that would share the exact localization of the bait while lacking all interactions would be difficult to achieve. Instead, diffusible cytosolic or membrane-bound GFP tagged with BirA* (or BirA* by itself), which should not form biologically meaningful interactions in mammalian cells, can be used as controls for unspecific/chance labeling [5, 20]. Although these controls partially overlap with BirA* negative cells in their content, having both will give a more accurate assessment of the background [16]. Unrelated BirA*tagged proteins that form independent interactions may not model the background correctly and are therefore not appropriate controls for scoring candidate interactions. 3.2 Generation of Stable BirA* Expressing Cell Lines

The host cells should express the integrin of interest to ensure that specific IAC components are coexpressed with the fusion construct. The endogenous protein may be deleted to make sure IAC assembly is supported by the ectopic form [5]. Electroporation is a suitable method for transfection of MDCK cells that are relatively hard to transfect. Transfection efficiency should be confirmed with a plasmid encoding a fluorescent reporter to visualize expression 24–48 h later. Around ~40–50% of MDCK cells are transfected by the following protocol, but any other protocol that yields similar or higher efficiency, is applicable. Not all transfected cells will have integrated the plasmid DNA in the genome and only a fraction of those express the protein at high levels. Therefore, selection of integrated cells using geneticin (G-418), followed by establishment clonal cell lines with validated fusion construct expression and localization ensures uniformity of biologically relevant material for identification. 1. Linearize 30 μg of pcDNA3.1 plasmid encoding for the BirA* construct with 45 U of MluI in NEB3.1 buffer in a total volume of 300 μl 16 h at +37  C (see Note 1). 2. Purify linearized DNA by Phenol-Chloroform-Isoamyl alcohol extraction and ethanol precipitation or using a commercial kit. Measure DNA concentration and purity with a spectrophotometer (optimal concentration is ~1 μg/μl). 3. Plate MDCK cells 24–48 h previously to reach ~80% confluency upon transfection (MDCK cells should be passaged twice a week at a ratio of 1:10–1:20). 4. Wash cells with 10 ml of PBS and with 5 ml of Trypsin-EDTA at room temperature (RT). Distribute 2 ml of trypsin on the cells and incubate at +37  C for 10–20 min until cells detach. Resuspend into 10 ml of culture media.

BioID of Integrin Interactors

63

5. Aliquot 1  106 cells into a conical tube, fill up to 10 ml with PBS and pellet cells by centrifugation at 600  g for 3 min. Remove supernatant. 6. Add 100 μl of electroporation solution and 3–5 μg of linearized plasmid DNA. Resuspend gently and transfer suspension into a cuvette (see Note 2). 7. Electroporate using Nucleofector device. 8. Add 500 μl of warm media directly into the cuvette and transfer suspension to a well on a 6-well plate containing 2 ml of prewarmed media. Prepare one well with an equal number of nontransfected cells. 9. Change medium 24 h after transfection to wash away dead cells. Add 600 μg/ml of G418 while changing the media again at 48 h (see Note 3). 10. Select cells with G-418 by changing the media every 2–3 days and trypsinize/replate whenever cells appear confluent (see Note 4). 11. When all nontransfected control cells have been eliminated, the colony of resistant cells can be expanded. At this point, an aliquot of cells can be frozen. 12. There are two options for generation of clonal cell lines to screen for fusion protein expression: (1) single cells can be sorted with a fluorescence activated cell sorting (FACS) machine directly into 96-wells containing 200 μl of conditional media (see Note 5); (2) Single cells can be handpicked from a sparse suspension distributed on a 10 cm Ø dish under the microscope into 96-wells containing 250 μl of conditional media. 13. When cells in 96-wells are confluent, trypsinize with 50 μl of trypsin, suspend to 300 μl with medium and divide 1:3 and 2:3 into a 96-well and a 24-well, respectively. 14. Cells in 96-wells can be directly lysed into 30–40 μl of Laemmli sample buffer and samples processed for SDS-PAGE and Western immunoblotting to confirm fusion protein expression (see Note 6). 15. Once positive clones are identified, corresponding cells in 24-wells are expanded for further analysis. 3.3 Maturation and Cell Surface Expression of BirA* Fusion Constructs

Integrins form heterodimers with one or several other integrins that are secreted to the cell surface and bind ECM ligands as one. The interaction between the subunits is relatively stable, which allows the heterodimer to be immunoprecipitated in conditions that disrupt other weaker interactions. Immunoprecipitation can be combined with biotinylation of surface proteins to confirm maturation of integrin heterodimers. Before lysis, cells are briefly

64

€ki et al. Satu-Marja Myllyma

treated on ice with membrane impermeable Sulfo-NHS-Biotin, which attaches itself to lysine residues. The ectopic integrin without the BirA* tag can be immunoprecipitated as a positive control in addition to immunoprecipitation from negative cells. Surface expressed heterodimers should appear as two or more bands (the construct itself and heterodimer partner(s)) on streptavidin-labeled western blots (Fig. 1b). The heterodimers that form can be predicted based on the expression pattern of integrins in the host cells (Fig. 1a). 1. Seed MDCK cells onto 10 Ø cm tissue culture dishes at a density of 4.5  105 to reach a confluency of 80–90% confluency upon 24 h of culture (see Note 7). 2. Treat cells on ice for 30 min with 0.5 mg/ml of Sulfo-NHSBiotin in biotinylation buffer and wash three times with 10 ml of ice-cold PBS. 3. Add 0.5–1 ml of ice-cold RIPA-buffer and collect cells by scraping into an Eppendorf tube. 4. Add 250 U of benzonase nuclease and rotate for 30 min at +4  C to reduce viscosity of the sample. Filter lysate through a Spin-X centrifuge tube filter by centrifugation at +4  C for 2 min. 5. Dilute 300–500 μg of total protein into 1 ml of RIPA-buffer and add 3 μg of antibody selective for the fusion protein (antiMyc or anti-BirA) to immunoprecipitate overnight (o/n) at +4  C on a rotating mixer. 6. Add 60 μg of paramagnetic Protein G Dynabeads and incubate 2–3 h at +4  C on a rotating mixer to bind immunocomplexes (see Note 8). 7. Wash beads four times with 1 ml of cold RIPA-buffer using a magnet. 8. Add 30 μl of Laemmli sample buffer with β-mercaptoethanol and release proteins by cooking at +95  C for 3 min. Collect the sample from the beads. 9. Resolve samples on a 6–7.5% SDS-PAGE gel and perform western immunoblotting to detect surface proteins with HRP-conjugated streptavidin diluted as 1:10,000. 3.4 Localization of BirA* Fusion Constructs and Substrates to IACs

IACs can be visualized as large focal deposits at the cell-ECM interface when stained by fluorescent antibodies. Total Internal Reflection Fluorescence microscopy (TIRF) is well suited for visualization of adhesions, because excitation is restricted to a range of 100–200 nm from the basal cell surface, eliminating out-of-focus light. However, confocal or fluorescence microscopy is recommended in parallel to identify other possible locations in the cell where the construct may be targeted to. TIRF microscopy requires

BioID of Integrin Interactors

65

cells to be plated on a glass surface with a defined thickness of 0.17 μm and immersed in an aqueous imaging media. The fusion construct should be stained together with the endogenous integrin, its heterodimer partner or other specific IAC marker. The ectopic integrin without BirA* fused to it should be included as a positive control, especially if it is of a different species than the host cells. 1. Coat the surface of a glass bottom 35 mm Ø dish with an ECM molecule of choice to allow cell attachment to glass (see Note 9). Most ECM molecules will absorb to glass if applied in PBS for 2 h at +37  C in PBS. 2. Apply media and seed cells at a desired density. 3. (Optional step) To visualize BirA* substrates, cells should be treated with 50 μM biotin for up to 24 h before fixation and untreated cells should be included as negative controls. 4. Fix cells with 4% PFA PBS for 10 min at RT, wash once with PBS. Do not allow the surface to dry while washing and do not pipet directly on top of the cells. 5. Quench for 20 min in 0.2 M glycine PBS. 6. Permeabilize cells with 0.1% TX-100 PBS for 15 min at RT, wash twice with PBS. 7. Block cells with 1% BSA in PBS for 30 min at RT. 8. Add primary antibodies and incubate 1 h at RT or o/n at +4  C. Wash 4  5 min at RT with PBS. 9. Add Alexa-conjugated secondary antibodies (or streptavidin for detection of biotinylation products) and incubate 1 h at RT or o/n at +4  C. Wash cells for 4  5 min at RT with PBS and leave in PBS. 10. Mount glass bottom dish onto a sample holder designed for imaging of 35 mm Ø dishes on a confocal microscope and perform TIRF-imaging with a high-resolution objective. 3.5 Proximal Biotinylation and Sample Collection

Saturated levels of proximal labeling are reached within 24 h of treatment with 50 μM concentration of biotin (2). Labeling conditions can be optimized by observing streptavidin-labeled BirA* substrates by TIRF imaging (see Subheading 2.4) or by western blotting after different labeling periods. If BirA is targeted to the extracellular domain, labeling should be done in sub-confluent conditions to allow diffusion of biotin through the extracellular space especially in MDCK cells that form epithelial barriers. Cells usually secrete several endogenous ligands to a specific integrin heterodimer. If the aim is to restrict IAC assembly to one ligand, labeling should be performed on a ligand-specified surface and limited to a few hours. For reduced labeling periods, the amount of material collected should be scaled up.

66

€ki et al. Satu-Marja Myllyma

1. Plate cells at a density of 4.5  105 onto tissue culture plates and allow cells to settle and adhere. 2. Add 50 μM of biotin, well-dissolved in culture media, and culture up to 24 h. 3. The presence of biotinylation products should be first verified relative to nonbiotinylated controls from cell lysates by Western blotting using HRP-conjugated streptavidin for detection. For this purpose, cells on a 35 mm Ø dish can be lysed into 200 μl of RIPA buffer (see Note 10). 4. To prepare samples for mass spectrometry, three 15 cm Ø dishes of cells, which constitutively and uniformly express the fusion protein at a high level, is sufficient. An equal amount of material should be prepared from negative cells and BirA* control cells. 5. Wash cells three times with 20 ml of cold PBS. 6. Scrape cells into 50 ml conical tubes and centrifuge 5 min 570  g (180  g if cells are sensitive) at +4  C, remove supernatant and snap freeze in liquid nitrogen. 3.6 Affinity Purification of Biotinylated Proteins

1. Add 3 ml of cold BioID lysis buffer and 1 μl of Benzonase nuclease to the cell pellet. Vortex the sample repeatedly 5 s at a time while incubating 5–10 min on ice to lyse the cells. 2. Sonicate lysates on ice three times, 3 min each, with 5 min breaks in between. 3. Divide sample into two 2-ml tubes and centrifuge at 16,000  g for 15 min at +4  C. 4. Collect supernatant and centrifuge again at 16,000  g for 10 min at +4  C. 5. Transfer 400 μl of IBA Strep-Tactin® beads as a 50% slurry into Bio-Rad spin columns and allow the resin to settle for 5 min. 6. Wash the beads once with 1 ml of BioID lysis buffer and apply the cleared lysate to the column on top of the resin without disturbing it. 7. Allow the lysate to flow through the column by gravity completely and wash beads three times with 1 ml of cold wash buffer. 8. Wash four times with 1 ml of HENN buffer while making sure to remove all residual lysis buffer from the column. 9. Secure the end cap to the column and add 300 μl of elution buffer. Seal the top with the snap cap. Invert the column several times and incubate at RT for 5 min. Remove the end cap and place the column on a 2 ml tube to elute proteins from the resin. Elute twice more and combine elutions to obtain maximum yield.

BioID of Integrin Interactors

3.7 Identification of Candidate IAC Proteins

4

67

Identification and label-free quantification of BioID labeled proteins can be done using a liquid chromatography tandem mass spectrometry setup (LC-MS/MS) following similar guidelines as described for affinity-purified proteins [20]. For further details on peptide preparation, LC-MS/MS measurement and protein identification, there is a recent protocol focused on this topic [20]. At least three biological replicates are recommended to test the significance of hits against negative controls. Statistical methods that compensate for multiple comparisons, such as the false discovery rate, can be applied to avoid false positives in complex proteomic data. Identifications with values that are close to the detection limit of the mass spectrometer in all the samples can be excluded from the identified list. Missing values, commonly found in negative controls, complicate numerical operations and can be imputed using a Gaussian distribution [21]. Usually a p-value/fold change -based cutoff is selected to identify high confidence candidate interactions. If the experiment includes multiple negative controls, negative cells can be used for scoring of interactions and BirA* controls as a secondary filtering step [5]. Comparisons with other unrelated BirA* datasets may be useful for demonstrating the specificity of labeling. Public databases such as the CRAPome (https:// www.crapome.org/), which collects proteomic data from negative controls, can be cross-referenced to further test the validity of the filtering criteria. Misfolded constructs may lead to false labeling of ER content, but certain integrins also have a significant steady state pool at the ER. Gene Ontology enrichment analysis with bioinformatics databases such as DAVID (https://david.ncifcrf.gov/) can be utilized to focus on proteins outside of the ER that are labeled by the secreted form and are more likely to be enriched for IAC components. Depending on the context, BioID data on IACs should partially overlap with IACs defined by affinity-based proteomic analyses [22]. However, focal adhesions characterized in fibroblasts and lymphoblasts are likely to be overrepresented in the proteomic adhesome compared to other types of adhesions [1]. Ultimately, the nature of the biological relationship between the bait and the prey should be defined experimentally.

Notes 1. Linearization of the vector improves transfection efficiency in MDCK cells. Any unique restriction site, which leaves open reading frames can be used. 2. The time cells are kept in the electroporation solution should be minimized.

68

€ki et al. Satu-Marja Myllyma

3. The concentration of G-418 should be determined experimentally to determine the optimal elimination of nontransfected cells and survival of resistant cells. 4. Removal of nonresistant cells will be more efficient when cells are kept in low density. Thus, it is necessary to occasionally trypsinize cells and replate to fresh 6-wells. 5. Conditional media can be prepared by collecting media from MDCK cells that have been in culture for 2–3 days. The media should be filtered to remove cell debris. 6. Antibodies against a small peptide tag such as Myc, BirA or integrins can be used for Western blotting (BirA adds approximately 35 kDa). Usually 20–40 MDCK cell clones contain a handful of positives. 7. MDCK cells should be subconfluent for integrins at the basal surface to be efficiently labeled. Alternatively, cells can be surface biotinylated from below, when seeded on Transwell inserts with a porous membrane bottom. 8. Alternatively, 50 μl protein-G agarose beads can be used for binding of immunocomplexes and beads washed by centrifugation for 2–3 min at 2500  g. 9. Alternatively, cells can be seeded onto ECM-coated 10 mm Ø coverslips placed into 24-wells or 35 mm Ø cell culture dishes, picked up for immunofluorescence staining and mounted onto glass slides for confocal imaging. 10. A gradient SDS-PAGE gel is ideal for resolving proteins with a range of different sizes. References 1. Manninen A, Varjosalo M (2017) A proteomics view on integrin-mediated adhesions. Proteomics 17(3–4). https://doi.org/10.1002/ pmic.201600022 2. Roux KJ, Kim DI, Raida M, Burke B (2012) A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells. J Cell Biol 196(6):801–810. https://doi.org/10.1083/jcb.201112098 3. Kim DI, Birendra KC, Zhu W, Motamedchaboki K, Doye V, Roux KJ (2014) Probing nuclear pore complex architecture with proximity-dependent biotinylation. Proc Natl Acad Sci U S A 111(24):E2453–E2461. https://doi.org/10.1073/pnas.1406459111 4. Dong JM, Tay FP, Swa HL, Gunaratne J, Leung T, Burke B, Manser E (2016) Proximity biotinylation provides insight into the molecular composition of focal adhesions at the nanometer scale. Sci Signal 9(432):rs4. https://doi. org/10.1126/scisignal.aaf3572

5. Myllymaki SM, Kamarainen UR, Liu X, Cruz SP, Miettinen S, Vuorela M, Varjosalo M, Manninen A (2019) Assembly of the beta4-integrin interactome based on proximal biotinylation in the presence and absence of heterodimerization. Mol Cell Proteomics 18(2):277–293. https://doi.org/10.1074/mcp.RA118. 001095 6. Rahikainen R, Ohman T, Turkki P, Varjosalo M, Hytonen VP (2019) Talinmediated force transmission and Talin rod domain unfolding independently regulate adhesion signaling. J Cell Sci 132(7): jcs226514. https://doi.org/10.1242/jcs. 226514 7. Wang W, Zuidema A, Te Molder L, Nahidiazar L, Hoekman L, Schmidt T, Coppola S, Sonnenberg A (2020) Hemidesmosomes modulate force generation via focal adhesions. J Cell Biol 219(2). https://doi.org/ 10.1083/jcb.201904137

BioID of Integrin Interactors 8. Van Itallie CM, Aponte A, Tietgens AJ, Gucek M, Fredriksson K, Anderson JM (2013) The N and C termini of ZO-1 are surrounded by distinct proteins and functional protein networks. J Biol Chem 288 (19):13775–13788. https://doi.org/10. 1074/jbc.M113.466193 9. Van Itallie CM, Tietgens AJ, Aponte A, Fredriksson K, Fanning AS, Gucek M, Anderson JM (2014) Biotin ligase tagging identifies proteins proximal to E-cadherin, including lipoma preferred partner, a regulator of epithelial cell-cell and cell-substrate adhesion. J Cell Sci 127(Pt 4):885–895. https://doi.org/10. 1242/jcs.140475 10. Fredriksson K, Van Itallie CM, Aponte A, Gucek M, Tietgens AJ, Anderson JM (2015) Proteomic analysis of proteins surrounding occludin and claudin-4 reveals their proximity to signaling and trafficking networks. PLoS One 10(3):e0117074. https://doi.org/10. 1371/journal.pone.0117074 11. Wehrle-Haller B (2007) Analysis of integrin dynamics by fluorescence recovery after photobleaching. Methods Mol Biol 370:173–202. https://doi.org/10.1007/978-1-59745-3530_13 12. Huet-Calderwood C, Rivera-Molina F, Iwamoto DV, Kromann EB, Toomre D, Calderwood DA (2017) Novel ecto-tagged integrins reveal their trafficking in live cells. Nat Commun 8(1):570. https://doi.org/10.1038/ s41467-017-00646-w 13. Soto-Ribeiro M, Kastberger B, Bachmann M, Azizi L, Fouad K, Jacquier MC, Boettiger D, Bouvard D, Bastmeyer M, Hytonen VP, Wehrle-Haller B (2019) beta1D integrin splice variant stabilizes integrin dynamics and reduces integrin signaling by limiting paxillin recruitment. J Cell Sci 132(8). https://doi.org/10. 1242/jcs.224493 14. Kim DI, Jensen SC, Noble KA, Kc B, Roux KH, Motamedchaboki K, Roux KJ (2016) An improved smaller biotin ligase for BioID proximity labeling. Mol Biol Cell 27 (8):1188–1196. https://doi.org/10.1091/ mbc.E15-12-0844 15. Roux KJ, Kim DI, Burke B (2013) BioID: a screen for protein-protein interactions. Curr

69

Protoc Protein Sci 74:19.23.11–19.23.14. https://doi.org/10.1002/0471140864. ps1923s74 16. Hesketh GG, Youn JY, Samavarchi-Tehrani P, Raught B, Gingras AC (2017) Parallel exploration of interaction space by BioID and affinity purification coupled to mass spectrometry. Methods Mol Biol 1550:115–136. https:// doi.org/10.1007/978-1-4939-6747-6_10 17. Sears RM, May DG, Roux KJ (2019) BioID as a tool for protein-proximity labeling in living cells. Methods Mol Biol 2012:299–313. https://doi.org/10.1007/978-1-4939-95462_15 18. May DG, Roux KJ (2019) BioID: a method to generate a history of protein associations. Methods Mol Biol 2008:83–95. https://doi. org/10.1007/978-1-4939-9537-0_7 19. Branon TC, Bosch JA, Sanchez AD, Udeshi ND, Svinkina T, Carr SA, Feldman JL, Perrimon N, Ting AY (2018) Efficient proximity labeling in living cells and organisms with TurboID. Nat Biotechnol 36(9):880–887. https://doi.org/10.1038/nbt.4201 20. Liu X, Salokas K, Tamene F, Jiu Y, Weldatsadik RG, Ohman T, Varjosalo M (2018) An AP-MS- and BioID-compatible MAC-tag enables comprehensive mapping of protein interactions and subcellular localizations. Nat Commun 9(1):1188. https://doi.org/10. 1038/s41467-018-03523-2 21. Lazar C, Gatto L, Ferro M, Bruley C, Burger T (2016) Accounting for the multiple natures of missing values in label-free quantitative proteomics data sets to compare imputation strategies. J Proteome Res 15(4):1116–1125. https://doi.org/10.1021/acs.jproteome. 5b00981 22. Horton ER, Byron A, Askari JA, Ng DHJ, Millon-Fremillon A, Robertson J, Koper EJ, Paul NR, Warwood S, Knight D, Humphries JD, Humphries MJ (2015) Definition of a consensus integrin adhesome and its dynamics during adhesion complex assembly and disassembly. Nat Cell Biol 17 (12):1577–1587. https://doi.org/10.1038/ ncb3257

Chapter 7 Analyzing the Integrin Adhesome by In Situ Proximity Ligation Assay Brian A. Perrino, Yeming Xie, and Cristina Alexandru Abstract The in situ proximity ligation assay (PLA) is capable of detecting single protein events such as protein protein–interactions and posttranslational modifications (e.g., protein phosphorylation) in tissue and cell samples prepared for analysis by immunofluorescent or immunohistochemical microscopy. The targets are detected using two primary antibodies which must be from different host species. A pair of secondary antibodies (PLA probes) conjugated to complementary oligonucleotides is applied to the sample, and a signal is generated only when the two PLA probes are in close proximity by their binding to the two primary antibodies that have bound to their targets in close proximity. The signal from each pair of PLA probes is visualized as an individual fluorescent spot. These PLA signals can be quantified (counted) using image analysis software (ImageJ), and also assigned to a specific subcellular location based on microscopy image overlays. In principle, in situ PLA offers a relatively simple and sensitive technique to analyze interactions among any proteins for which suitable antibodies are available. Integrin-mediated focal adhesions (FAs) are large multiprotein complexes consisting of more than 150 proteins, also known as the integrin adhesome, which link the extracellular matrix (ECM) to the actin cytoskeleton and regulate the functioning of mechanosignaling pathways. The in situ PLA approach is well suited for examining the spatiotemporal aspects of protein posttranslational modifications and protein interactions occurring in dynamic multiprotein complexes such as integrin mediated focal adhesions. Key words Integrin, Proximity ligation assay, Cancer, Protein–protein interaction, Immunofluorescent microscopy, Immunohistochemistry

1

Introduction Integrins are cell surface receptors composed of noncovalently linked heterodimeric α and β subunits that function to integrate and transduce mechanical signals from both sides of the plasma membrane through large multiprotein signaling complexes known as focal adhesions (FAs) [1]. FAs consist of more than 150 interacting proteins, which collectively form a mechanical linkage between the extracellular matrix (ECM) and the actin cytoskeleton [2]. Each integrin subunit possesses a large extracellular domain, a single transmembrane domain, and a short cytoplasmic tail [3]. The α

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_7, © Springer Science+Business Media, LLC, part of Springer Nature 2021

71

72

Brian A. Perrino et al.

subunits allow integrins to selectively bind to the integrin-binding RGD motifs of the ECM proteins collagen and laminin [4]. The β subunits mediate the interactions between the integrin heterodimer and the numerous FA proteins that connect to the actomyosin cytoskeleton and influence multiple signaling pathways [4]. Twenty-four distinct integrin subtypes have been identified in mammals and are made up of 18 α subunits and eight β subunits [5]. Although ubiquitously expressed, individual tissues and cell types display different patterns of αβ integrin heterodimer expression [6]. As the primary receptors involved in cell–matrix adhesion, integrins are important to the regulation of a wide range of physiological processes including cell survival, proliferation, migration, innate immunity, and the tensile strength and integrity of tissues and organs [6]. The expression of integrins and FA proteins also influences the acquisition and maintenance of the characteristics of transformed cells, including proliferation, survival, migration, invasion, and metastasis [7]. Understanding how integrin expression and function and the interactions of the proteins comprising FAs are regulated during carcinogenesis and tumor progression will enable the development of new therapeutic approaches to inhibit tumorigenesis and suppress their metastatic phenotype. By utilizing and expanding existing immunofluorescence (IF) and immunohistochemistry (IHC) approaches. The in situ proximity ligation assay represents a relatively convenient and sensitive technique to provide spatiotemporal analyses of single protein expression, protein–protein interactions, and protein posttranslational modifications (PTMs) (e.g., phosphorylation) in fixed, intact cells or tissues [8]. The antibody-based proximity ligation assay (PLA) allows in situ detection and localization of endogenous proteins, protein– protein interactions, and protein PTMs, with high specificity and sensitivity in cells and tissues [2]. Typically, two primary antibodies raised in different species are used to detect two unique antigenic targets (Fig. 1). A pair of secondary antibodies conjugated to complementary oligonucleotides (PLA probes) are then used to recognize the primary antibodies from the different host species (Fig. 1a). If the targets of the primary antibodies are within 40 nm of each other, the complementary oligonucleotides on the secondary antibodies will be able to hybridize and join the PLA probes together (Fig. 1b–e). DNA ligase then forms a closed, circular DNA template that is required for rolling-circle amplification (RCA) which then acts as a primer for DNA polymerase, which generates concatemeric sequences during the RCA (Fig. 1f). This allows up to a 1000-fold amplified signal which is still linked to the PLA probe, allowing for localization of the signal (Fig. 1f). Finally, a colorimetric or fluorescent dye conjugated to a complementary oligonucleotide hybridizes to the complementary sequences within the amplicon (Fig. 1g), which are then visualized as discrete spots

PLA of the Integrin Adhesome

73

Fig. 1 Principle of PLA. Two primary antibodies raised in different species are used to detect two unique antigenic targets. A pair of secondary antibodies conjugated to complementary oligonucleotides (PLA probes) are then used to recognize the primary antibodies from the different host species (a). If the targets of the primary antibodies are within 40 nm of each other, the complementary oligonucleotides will be able to hybridize and join the PLA probes together to allow DNA ligase to form a closed, circular DNA template (b, c, d, E) which acts as a primer for DNA polymerase to carry out rolling-circle amplification (RCA) (f), generating concatemeric sequences during the RCA (f). This allows up to a 1000-fold amplified signal which is still linked to the PLA probe, allowing localization of the signal within the sample (f). Finally, a colorimetric or fluorescent dye conjugated to a complementary oligonucleotide hybridizes to the complementary sequences within the amplicon (g), which are then visualized as discrete spots (PLA signals) by microscopy image analysis. (b) Direct protein–protein interaction. (c) Indirect protein–protein interaction. As long as two proteins in a multiprotein complex are within 40 nm of each other, PLA should be able to detect their interaction. (d) Detecting PTMs (phosphorylation) by PLA. (e) Detecting a single protein by PLA. The primary antibodies must be from different host species and also recognize different epitopes on the protein

(PLA signals) by microscopy image analysis. Figure 2 shows a typical example of ITGA8/ITGB1 in situ PLA images obtained in our lab from OCT-embedded slices of human stomach smooth muscle (Fig. 2a) and mucosa (Fig. 2b, c). Quantitation of the spot number is carried out using imaging software such as ImageJ/Fiji [9], allowing for changes in protein abundance, protein– protein interactions, and protein PTMs to be quantified, and changes in their locations to be monitored. In situ PLA can be performed on adherent cells, cytospin preparations, and tissue sections using immunofluorescence (i.e., green, red, far-red, or orange detection) or immunohistochemistry (i.e., peroxidase-catalyzed reaction) [10]. In addition, in situ PLA can be performed on blood or suspension cells for detection by flow cytometry and can be used in multiwell plates (e.g., 96- or 384-well) with a highcontent screening imager for high-throughput analyses [11]. Detecting protein–protein interactions by PLA does not necessarily require a direct interaction between the proteins; as long as the two proteins of interest are within 40 nm of each other in a multiprotein complex, their presence can be detected by PLA [12]. Protein phosphorylation is detected and measured using antibodies against the phosphorylated Ser, Thr, or Tyr, and against the parent protein. Single-protein PLA can be used to detect subcellular compartmentalization of a specific protein, or as an internal standard to measure changes in other protein–protein interactions, protein phosphorylation, or protein expression [13]. The Duolink® PLA reagents are compatible with existing IHC and IF protocols.

74

Brian A. Perrino et al.

Fig. 2 In situ PLA of ITGA8/ITGB1. Human gastric antrum muscle and mucosa samples were prepared for PLA as described [13]. (a) Cross section through the circular muscle layer. Cross section (b) and longitudinal section (c) through the mucosa. The mouse anti-ITGA8 antibody was obtained from Santa Cruz Biotechnology, and the rabbit anti-ITGB1 antibody was obtained from Genetex. Mounting medium with DAPI was used to visualize the nuclei

PLA of the Integrin Adhesome

75

For researchers interested in expanding their current IHC and IF approaches using the in situ PLA approach, it is assumed that the protocols and techniques of sample preparation, fixation, and permeabilization are already in place. These procedures may have to be slightly modified to accommodate the dual or sequential primary antibody incubations, and the additional steps of ligation and RCA reactions in the PLA protocol.

2

Materials Before starting the assay protocol, the sample should have been deposited on a glass slide or coverslip, prepared with respect to the fixation, antigen retrieval, and permeabilization conditions that are compatible with your primary antibodies. All solutions and buffers should be prepared using ultrapure water (18 MΩ) and analytical grade reagents. Bring the wash buffers and mounting medium with DAPI to room temperature before use. Antibodies

A pair of primary antibodies for detecting the protein–protein interaction, phosphoprotein, or single protein. Most antibodies that work in IF can be employed for PLA.

2.2 Duolink® In Situ PLA Reagents

All Duolink® PLA reagents can be obtained from Sigma-Aldrich. Starter kits are available, based on the primary antibody species and desired detection color, and contain all the following necessary components to analyze up to 30 reactions based on a 40 μl reaction volume.

2.1

1. The PLA PLUS or MINUS probe contains an affinity purified, oligonucleotide-conjugated antibody against either mouse, rabbit, or goat IgG (H + L) and each PLA probe is supplied as a 5 concentrated stock. Store the PLA Probes at 4  C. Do not freeze the PLA probes. The PLA Probes and the primary antibodies are diluted using the Antibody Diluent Buffer. 2. Antibody Diluent Buffer is supplied at a 1 concentration. Prepare the appropriate volumes of diluted primary antibodies and the PLA Probes the day of the experiment, do not store diluted primary antibodies and PLA Probes. 3. Blocking Solution is supplied ready to use for incubating the sample prior to the antibody incubations. The Antibody Diluent and Blocking Solution contain salts, blocking agent and detergents to prevent nonspecific binding of the antibodies (see Note 1). 4. The Detection Reagents Kit contains DNA Ligase (1 U/μl), 5 concentrated Ligation Buffer containing the oligonucleotides that hybridize to the PLA probes (see Note 2); DNA Polymerase (10 U/μl), and the 5 concentrated Amplification

76

Brian A. Perrino et al.

Buffer containing all the components needed for rolling circle amplification (RCA), along with the oligonucleotide probes conjugated to the fluorophore that hybridize to the RCA product. The Detection Kit is stored at 20  C. 5. Mounting Medium with DAPI stains the nuclei for visualization, and prevents photobleaching of the PLA signals. The Mounting Medium with DAPI does not solidify but remains a liquid on the slide. For use with oil immersion microscopy or for prolonged storage, permanently seal the edges of the coverslip with nail polish or plastic sealant. 6. Wash Buffer A and Wash Buffer B are supplied in powder form to prepare 1 L of each wash buffer. Here, we provide the recipes. Wash Buffer A: Dissolve 8.8 g NaCl, 1.2 g Tris base, and 0.5 ml Tween 20 in high purity water. Adjust pH to 7.4 using HCl and then bring to the final volume of 1 L with high purity water (final concentrations 0.01 M Tris, 0.15 M NaCl and 0.05% Tween 20). Wash Buffer B: Dissolve 5.84 g NaCl, 4.24 g Tris base and 26.0 g Tris–HCl in high purity water. Adjust pH to 7.5 using HCl. Add high purity water to 1 L (final concentrations 0.2 M Tris and 0.1 M NaCl). Filter each wash buffer through a 0.22 μm filter and store at 4  C (see Note 3). 2.3

Equipment

1. Fluorescence microscope equipped with excitation/emission filters compatible with fluorophore and nuclear stain excitation/emission. 2. Staining jars. 3. Pen or mask for delimitation of reaction area (e.g., grease pen or silicone mask). 4. Shaker. 5. Humidified chamber (moist chamber) (see Note 4). 6. Freeze block for enzymes. 7. Incubator, 37  C. 8. Glass Slides. 9. Coverslips compatible with fluorescence microscopy. 10. Soft tissues (e.g., Kimwipes). 11. Nail polish.

2.4 Software for Image Analysis

The result from a PLA experiment is typically a number of distinct fluorescent dots (“PLA signals”) of ~1 μm size in various locations in the cells or tissue samples. To analyze the result, any image analysis software using suitable algorithms, such as Fiji-ImageJ, or BlobFinder freeware can be used.

PLA of the Integrin Adhesome

3

77

Methods

3.1 Reaction Volumes

Use open droplet reactions. Perform all incubations in a preheated humidity chamber to avoid evaporation of the reactions. Suitable reaction volumes sufficient to cover different reaction areas are shown in Table 1. Never use less than 15 μl of total reaction volume.

3.2

Wash Volumes

Because of the multiple antibody incubations, washing should be performed in a minimum volume of 70 ml with gentle orbital shaking or rocking to minimize any residual soluble, unbound antibody from interfering with the subsequent antibody additions.

3.3

PLA Procedure

3.3.1 Blocking

3.3.2 Primary Antibodies

1. Add the Blocking solution dropwise to each sample. 2. Incubate the samples in a preheated humidified chamber for 30 min at 37  C (see Note 5). Each antibody will require some optimization for the in situ PLA experimental conditions. Different dilutions of each primary antibody may be necessary. In some cases, the primary antibodies can be diluted together, and applied simultaneously to the sample; in other cases it may be necessary to dilute each primary antibody separately, and apply them to the sample sequentially. 1. Dilute the primary antibodies to their suitable concentrations in the required volume of. 2. Antibody Diluent. 3. Tap off the Blocking solution from the samples onto a Kimwipe (see Note 6). Table 1 Suitable reaction volumes for different sample areas Area (cm2)

Total reaction volume (μl)

0.2

15

0.5

25

1

40

2

80

3

120

4

160

6

240

8

320

10

400

78

Brian A. Perrino et al.

4. Immediately add the primary antibody solution to the samples (see Note 7). 5. Incubate the samples in the preheated humidified chamber, using the optimal incubation time and temperature for the primary antibodies. 3.3.3 PLA Probes

1. Dilute each PLA probe 1:5 together in the required volume of Antibody Diluent and vortex. 2. Tap off the primary antibody solution from the samples. 3. Wash the samples in wash buffer A for 5 min, twice. 4. Tap off the wash solution from the samples after the last wash. 5. Add the PLA probe solution. 6. Incubate the samples in the humidified chamber for 1 h at 37  C.

3.3.4 Ligation

1. Dilute the Ligation Buffer 1:5 in the required volume of high purity water and vortex. Add the Ligase at a 1:40 dilution to the Ligation buffer, and vortex, immediately before addition to the samples. Take the addition of Ligase into account when calculating the amount of water added. 2. Tap off the PLA probe solution from the samples. 3. Wash the samples in wash buffer A for 5 min, twice. 4. Tap off the wash solution from the samples after the last wash. 5. Add the Ligation solution to each sample. 6. Incubate the samples in the preheated humidified chamber for 30 min at 37  C.

3.3.5 Amplification

1. Dilute the Amplification solution 1:5 in high purity water and vortex. 2. Add the Polymerase to the Amplification solution, and vortex, at a 1:80 dilution immediately before addition to the samples. Take the addition of Polymerase into account when calculating the amount of water added. 3. Tap off the Ligation solution from the samples. 4. Wash the samples with wash buffer A for 5 min, twice. 5. Tap off the wash solution from the samples after the last wash. 6. Add the Amplification solution to each sample. 7. Incubate the samples in the preheated humidity chamber for 100 min at 37  C (see Note 8).

3.3.6 Final Wash Step

1. Tap off the Amplification solution from the samples. 2. Wash the samples with wash buffer B for 2 min, twice. 3. Place the samples in 0.01 wash buffer B prior to adding mounting medium.

PLA of the Integrin Adhesome 3.3.7 Preparation for Imaging

79

1. Tap off wash buffer from the samples. 2. Use a minimal volume of Duolink Mounting Medium with DAPI (~6 μl per 1 cm2), and apply a coverslip on top of your sample ensuring no air bubbles are trapped under the coverslip. Nail polish can be used to seal the edges (see Note 9). 3. Wait for approximately 15 min before analyzing with a fluorescence or confocal microscope, using at least a 20 objective. After imaging, store the samples at -20  C in the dark. The PLA spot number can be normalized to the number of nuclei, or as we have done, to the cellular cross-sectional areas using ITGB1 IF, or to the spot count from single-protein PLA, after determining that these parameters do not change during in response to the experimental parameters [13].

3.3.8 Biological and Technical Controls for In Situ PLA

4

As in other antibody-based immunohistochemical assays, a set of controls is necessary to ensure the specificity and validity of the antibody signals, and to avoid false-positive signals. Each primary antibody should be tested individually by IF or IHC to determine that it is able to recognize its specific target under the conditions used. After detecting protein–protein interactions by in situ PLA with your particular antibody pair, carry out in situ PLA with an antibody against a protein that is known to not interact or be in close proximity to your interacting proteins of interest, as a negative biological control. Technical controls for in situ PLA include incubating the sample with only one of the primary antibodies, and incubating the sample with only the PLA probes (see Note 10).

Notes 1. The in situ PLA protocol is compatible with a wide variety of antibody dilution and blocking buffers, allowing for previously optimized antibody buffers to be used. 2. The Ligation buffer contains dithiothreitol (DTT) that may precipitate during freezing. Warm to 37  C and vortex to dissolve/homogenize. 3. 1 M Tris–HCl, pH 7.5 can be used to prepare the wash buffers to avoid having to manually adjust the pH. 4. In the PLA protocol there are additional incubation periods for ligation and RCA subsequent to the primary and secondary antibody incubations. Thus, it is critical to ensure that a sufficiently humid environment is provided to prevent evaporation of the reaction solutions. We use a Tupperware container or equivalent which is lined with filter paper soaked with water and with excess water to a depth of around 0.5 cm. An empty pipet tip rack is placed inside the container, and the slides or

80

Brian A. Perrino et al.

coverslips are placed on the pipet tip rack to raise them above the water. In addition, a couple of Kimwipes are soaked with water and placed in the four corners of the container. We warm the water used to humidify the chamber to 37  C prior to use. We also place the empty chamber back into the 37  C incubator while we are washing, adding, or removing solutions from the samples to keep it warm. 5. It is important that all incubations are performed in a humid environment to prevent excessive evaporation. 6. Do not allow the samples to dry before addition of the primary antibodies as this will cause background. Try to obtain equal residual volume for each sample as this will affect the final concentration of the antibody and reproducibility. 7. If adding the primary antibodies sequentially, tap off the first primary antibody solution, and wash the samples in wash buffer A for 5 min, twice. Tap off the wash solution after the last wash prior to adding the next primary antibody. 8. Based on the kinetics of the polymerase, this length of time will typically yield PLA spots of ~1 μm in diameter [12]. 9. Duolink® In Situ Mounting Media with DAPI is aqueous and does not solidify. 10. It has been reported that the number of PLA signals generated can decrease as the kits get older [14]. We did not experience any differences in the PLA results as the kits aged; but we did not use them after their expiration date.

Acknowledgments This work was supported by y a National Institute of Diabetes and Digestive and Kidney Diseases Diabetic Complications Consortium (DiaComp, http://www.diacomp.org) Grant DK076169, and a Takeda Pharmaceuticals USA, Innovation Center Grant to B.A.P., and by a Mick Hitchcock Graduate Student Scholarship to Y.X. References 1. Ginsberg MH (2014) Integrin activation. BMB Rep 47(12):655–659. https://doi.org/ 10.5483/bmbrep.2014.47.12.241 2. Shams H, Hoffman BD, Mofrad MRK (2018) The “stressful” life of cell adhesion molecules: on the mechanosensitivity of integrin adhesome. J Biomech Eng 140(2). https://doi. org/10.1115/1.4038812

3. Campbell ID, Humphries MJ (2011) Integrin structure, activation, and interactions. Cold Spring Harb Perspect Biol 3(3):a004994. https://doi.org/10.1101/cshperspect. a004994 4. Barczyk M, Carracedo S, Gullberg D (2010) Integrins. Cell Tissue Res 339(1):269–280. https://doi.org/10.1007/s00441-009-08346

PLA of the Integrin Adhesome 5. Takada Y, Ye X, Simon S (2007) The integrins. Genome Biol 8(5):215–215. https://doi.org/ 10.1186/gb-2007-8-5-215 6. Seetharaman S, Etienne-Manneville S (2018) Integrin diversity brings specificity in mechanotransduction. Biol Cell 110(3):49–64. https://doi.org/10.1111/boc.201700060 7. Jang I, Beningo KA (2019) Integrins, CAFs and mechanical forces in the progression of cancer. Cancers (Basel) 11(5):721. https:// doi.org/10.3390/cancers11050721 8. Alam MS (2018) Proximity ligation assay (PLA). Curr Protoc Immunol 123(1): e58–e58. https://doi.org/10.1002/cpim.58 9. So¨derberg O, Leuchowius K-J, Gullberg M, Jarvius M, Weibrecht I, Larsson L-G, Landegren U (2008) Characterizing proteins and their interactions in cells and tissues using the in situ proximity ligation assay. Methods 45 (3):227–232. https://doi.org/10.1016/j. ymeth.2008.06.014 10. Rho Jin K, Lee H, Park C-S, Choi C-M, Lee Jae C (2013) Sensitive detection of EML4ALK fusion oncoprotein of lung cancer by in situ proximity ligation assay. Clin Chem Lab

81

Med 51:1843. https://doi.org/10.1515/ cclm-2013-0044 11. Fiebitz A, Nyarsik L, Haendler B, Hu Y-H, Wagner F, Thamm S, Lehrach H, Janitz M, Vanhecke D (2008) High-throughput mammalian two-hybrid screening for proteinprotein interactions using transfected cell arrays. BMC Genomics 9:68–68. https://doi. org/10.1186/1471-2164-9-68 12. Raykova D, Koos B, Asplund A, Gelle´ri M, Ivarsson Y, Danielson UH, So¨derberg O (2016) Let there be light! Proteomes 4(4):36. https://doi.org/10.3390/ proteomes4040036 13. Xie Y, Perrino BA (2019) Quantitative in situ proximity ligation assays examining protein interactions and phosphorylation during smooth muscle contractions. Anal Biochem 577:1–13. https://doi.org/10.1016/j.ab. 2019.04.009 14. Ulke-Lemee A, Turner SR, MacDonald JA (2015) In situ analysis of smoothelin-like 1 and calmodulin interactions in smooth muscle cells by proximity ligation. J Cell Biochem 116(11):2667–2675. https://doi.org/10. 1002/jcb.25215

Part III Biochemical, Proteomics and Computational Methods to Determine Integrin Interactions

Chapter 8 Single-Protein Tracking to Study Protein Interactions During Integrin-Based Migration A. V. Radhakrishnan, Tianchi Chen, Jose Filipe Nunes Vicente, Thomas Orre´, Amine Mehidi, Olivier Rossier, and Gre´gory Giannone Abstract Cell migration is a complex biophysical process which involves the coordination of molecular assemblies including integrin-dependent adhesions, signaling networks and force-generating cytoskeletal structures incorporating both actin polymerization and myosin activity. During the last decades, proteomic studies have generated impressive protein–protein interaction maps, although the subcellular location, duration, strength, sequence, and nature of these interactions are still concealed. In this chapter we describe how recent developments in superresolution microscopy (SRM) and single-protein tracking (SPT) start to unravel protein interactions and actions in subcellular molecular assemblies driving cell migration. Key words Single-protein tracking, Superresolution microscopy, Optogenetics, Supercritical-angle fluorescence emission, Integrin-dependent adhesion, Actin-based lamellipodium

Abbreviations Cas9 CRISPR DONALD dSTORM FBS FRAP FWHM mEos2 MSD PALM SAFe SMLM SPT TIRF WRC

CRISPR associated protein 9 Clustered regularly interspaced short palindromic repeats Direct optical nanoscopy with axially localized detection Direct stochastic optical reconstruction microscopy Fetal bovine serum Fluorescence recovery after photobleaching Full width at half maximum Monomeric Eos2 Mean square displacement Photoactivation localization microscopy Supercritical angle fluorescence emission Single-molecule localization microscopy Single-particle tracking Total internal reflection fluorescence Wave regulatory complex

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_8, © Springer Science+Business Media, LLC, part of Springer Nature 2021

85

86

1

A. V. Radhakrishnan et al.

Introduction Cell migration is an important function involved in various biological processes such as embryo development [1], wound healing [2, 3], and cancer cell metastasis [4]. The movement is usually mediated by protrusive membrane structures like lamellipodia, a flat cellular domain that is composed of a dynamic machinery of actin cytoskeleton, adhesion complexes, and various regulators [5– 7]. At the tip of the lamellipodium, branched actin networks are formed through the nucleator Arp2/3 and nucleation promoting factors such as the SCAR/WAVE complex [8, 9] under the regulation of small GTPases Rho, Rac, and Cdc42 [10, 11]. Actin polymerization induces a retrograde flow of assembled actin that engages and binds to nascent adhesions, while exerting force on the substrate through cell–matrix adhesions to facilitate cell migration [12, 13]. The periodic protrusion and contraction of the lamellipodium associated with nonmuscle myosin II can also exert forces on the adhesions to help their formation and growth [14, 15]. At the rear of the lamellipodium, myosin II pulls on actin filaments and remodels them to form stress fibers that connect to and exert force on the substrate through mature focal adhesions [16–18]. Thus, many important macromolecules undergo motions and transient interactions that are essential to their function. These molecules form nanomachine-like protein complexes that control actin assembly and adhesion formation; thus, they must be tightly regulated at specific locations at precise times. Cell adhesion to the extracellular matrix (ECM) is one of the fundamental processes that are central in many developmental as well as pathophysiological contexts [19–21]. Cell-matrix adhesion is mainly mediated by ubiquitously expressed transmembrane receptors called integrins, which bind to ECM ligands like collagens, laminins or fibronectin (FN). At the same time, they connect to the filamentous F-actin cytoskeleton [19–21]. Integrins function as heterodimers. In mammals, 18 α and 8 β subunits combine in a restricted manner to form 24 specific receptors, most of which exhibit specific ligand binding and display unique mechanical biochemical signaling properties [19]. Each integrin subunit has a large extracellular domain that constitutes the ligand binding domain, a single transmembrane domain and a cytoplasmic tail of variable length. Since the short cytoplasmic tails of integrins lack enzymatic and F-actin–binding activity, they depend on the assembly of adaptor and signaling proteins onto their cytoplasmic tails for signal propagation [22]. The resulting integrin adhesion sites (IAS) can consist of hundreds of different proteins, and each integrin receptor assembles a specialized adhesion complex with distinct molecular organization, life span and mechanochemical signaling potential [23, 24]. Moreover, within the same IAS,

Single-Molecule Integrin Tracking

87

distinct integrin-class receptors having proper biochemical and mechanical signaling, coexist and cooperate to determine the function of the IAS as a whole [25, 26]. A remarkable property of integrins is their tunable affinity, which enables them to reversibly switch from an inactive to an active state after binding of intracellular activators [27, 28]. Interestingly, many proteins compete for a few binding sites on the short cytoplasmic tails of β integrins, implying that regulatory processes control the spatiotemporal reversible binding of β-integrins [22]. For instance, the small intracellular tail of integrins harbors close-by binding sites for regulators controlling integrin activation state. Talin and kindlin may cooperate to activate integrins by binding to different regions of the β integrin tail, while ICAP-1 competes with kindlin to prevent integrin activation [22, 29]. The protein networks in the IAS have been named integrin “adhesomes” and were determined by both mass spectrometry approaches [30–32] and data mining [33, 34]. The components of integrin-based adhesive structures are regrouped in functional families including adhesion receptors, adaptor proteins, actin regulators, kinases/phosphatases, Rho GTPases, lipids, and enzymes of proteolytic activities. A meta-analysis collecting data from several proteomic studies [34] reported that thousands of proteins are enriched in the IAS. Cross-analysis of the different datasets identified a subset of 60 proteins, comprising the “consensus adhesome,” systematically detected in IAS. These proteins could be grouped in four subnetworks of protein interactions: (1) α-actinin-zyxin-VASP, (2) FAK-paxillin, (3) kindlin-ILKPINCH, and (4) talin-vinculin [34]. These IAS components can attain different phosphorylation states, and some are specifically recruited by given integrin subtypes [26, 30, 35, 36]. Proteomic studies have also demonstrated that IAS maturation associated with mechanical tension affect their composition. Myosin-II inhibition decreases the recruitment of most consensus adhesome proteins in the IAS, in particular actin-binding proteins including α-actinin and VASP [34], and LIM-domain proteins such as zyxin and paxillin [26]. Nevertheless, the recruitment of a subset of IAS components could also be increased after myosin-II inhibition, including β-Pix, a Rac1 activator promoting the formation of actin-based lamellipodial protrusions but inhibiting adhesion maturation [32]. However, such interaction maps are not taking into account the subcellular location, duration, strength, order of recruitment, and nature (competitive or cooperative) of these interactions. Furthermore, forces exerted on critical IAS proteins control their binding and enzymatic activities, stabilize and destabilize interactions, reveal domains of interactions, and induce posttranslational modifications [37–40]. Therefore, it is of great interest to develop new advanced imaging methods to detect and measure not only the position and dynamics of cellular components at the nanometer

88

A. V. Radhakrishnan et al.

level but also their interactions and how mechanical forces control those interactions. Surrounded by a complex and dynamic macromolecular environment, the IAS is the nexus of a subcellular system comprising multiple organelles: the plasma membrane, actin and microtubule cytoskeletons, lipid vesicles, and the cytosol composed of soluble proteins. Thus, proteins entering or leaving a specific location within the IAS may use a path involving any of these components, and the available paths may change during IAS maturation. IAS building blocks may be individual proteins or preassembled complexes [41, 42], which reorganize their interactions as the IAS matures. Thus, once inside the IAS, the multivalence and combinatorial diversity of interactions between the constituents enables internal reorganization of the building blocks to support IAS functions. The majority of knowledge about protein interactions during integrin activation is paradoxically derived from studies in which the complexity of the adhesive structures found in adherent cells was reduced or even absent, such as by flow cytometry with suspended cells [43] and by in vitro biochemistry [44], in which cases the intricate and dynamic interactions between binding partners in the IAS are ignored, let alone the complex and diverse in vivo cellular environment that may functionally modify and control these interactions. In order to understand the spatial temporal assembly of IAS, especially at the scale of nanometers and microseconds, where proteins interact and cluster, there is a clear need for techniques that can reveal the fast dynamics and transient interactions of IAS components, illuminate the nanoscale organization, and detect the subcellular location and duration of molecular events. The recent development of single-protein tracking (SPT) and superresolution microscopy (SRM) techniques has provided the possibility to quantitatively study protein motion and location within subcellular macromolecular complexes in their native environment, revealing of their interactions with other proteins and clustering behavior, amid various mechanisms of regulation [25, 45–48]. What information about protein interactions in the IAS can we glean from the SPT trajectories and SRM? (Fig. 1). First, SPT trajectories reveal the diffusion and immobilization events of proteins (Fig. 1b, c). The frequency of integrin immobilization within the IAS is higher than outside, indicating integrinbinding to extracellular ligand or intracellular IAS components, while integrin immobilization also correlates with its activation, as shown by treatment with Mn2+ or mutation [25]. Increase in immobilization has also been observed for small GTPases such as Rac1 in integrin-dependent focal adhesions [49], in neuronal dendritic spines [46] and in lamellipodia of motile cells [48]. Second, SPT shows the location and dynamic nature of protein interactions within the IAS (Fig. 1a), as a substantial fraction of integrins inside IAS can engage in free diffusion, and individual

Fig. 1 Schematic for spt-PALM and superresolution microscopy techniques to study protein interaction. (a, b) Typical information extracted from sptPALM experiments. (a) Superresolved reconstructed image from the single-molecule detection and localization data for mEos2-Kindlin-2-WT in an MEF, obtained from a sptPALM

90

A. V. Radhakrishnan et al.

integrin molecules cycle from diffusion to immobilization states which can last for seconds to tens of seconds [25, 50, 51]. This dynamic interaction, characterized by increased fraction and duration of immobilization, is associated with integrin activation and is controlled by the interaction of integrins with specific activators and inhibitors. These interactions are also compartmentalized in a given area that is regulated by membrane nanotopology, as demonstrated for cancer cells [52]. The apparent diffusion coefficient of integrin is also lower inside the IAS compared to outside [25], potentially due to confinement, but also suggesting transient interactions with IAS components and other effects of local protein crowding. Third, SPT and SRM revealed the 3D nanoscale organization and segregation of proteins into functional nanodomains as well as the path used by crucial integrin activators to reach these domains and activate integrins (Fig. 1e). 3D superresolution microscopy with iPALM has shown the distinct layers of IAS protein axial organization, where an integrin signaling layer near the membrane contains the cytoplasmic tail of integrin, paxillin, and FAK; a force transduction layer in between contains talin and vinculin; and an actin regulatory layer deeper inside the cytosol containing actin, zyxin, VASP, and alpha-actinin [53]. Individual proteins such as talin can span different layers and proteins such as vinculin also shows specific orientation and can shuttle between different layers depending on its activation state [13]. Whereas talin exhibits no membrane free-diffusion and is thus directly recruited to IAS directly from the cytosol [25], kindlin can diffuse on the membrane due to its PH domain [54]. With SPT, we are beginning to understand how proteins travel in different routes to reach designated ä Fig. 1 (continued) acquisition at 50 Hz (duration: 80 s) using an EMCCD camera (scale bar 3 μm). Inset: low-resolution image of Paxillin-GFP, coexpressed for IAS labeling. (b) Superresolved trajectories obtained by reconnecting the detections of mEos2-Kindlin-2-WT, overlaid on IAS. Trajectories are color-coded to show the different diffusion modes: free diffusion (green), confined (yellow), and immobile (red) (top). Distribution of the diffusion coefficient D computed from the trajectories of mEos2-Kindlin-2-WT obtained inside IAS (bottom, left). The gray area including D values inferior to 0.011 μm2/s corresponds to immobilized proteins. Values represent the average of the distributions obtained from different cells. Inset, MSD for trajectories corresponding to freely diffusive, confined, and immobile motion (green, yellow, and red, respectively) (bottom, right). (c–e) Additional information extracted from sptPALM and SAFe experiments. (c) Superresolved trajectories of mEos-Rac1-Q61L in the lamellipodium of a spreading MEF, obtained from a fast sptPALM acquisition with a sCMOS Camera at 333 Hz (duration: 12 s). Trajectories are overlaid on a lamellipodium expressing α-actinin-GFP and color-coded to show the different diffusion modes. (d) Schematic of Rac1 optogenetic activation at the lamellipodium tip with CRY2-CIBN system. (e) 3D superresolution images obtained by SAFe in combination with dSTORM imaging using anti-GFP nanobodies for Kindlin2-WT and Kindlin2-ΔPH. For each pixel, the average axial localization of detected single molecules is color-coded. (Part of this figure is reprinted from ref. 48, with permission from Elsevier)

Single-Molecule Integrin Tracking

91

functional layers. Indeed, deletion of the PH domain of kindlin resulted in its upward shift from the integrin layer localization [54]. In the integrin signaling layer, β3 and β1 integrins which show different force sensitivity and association with downstream signaling partners, suggesting different roles in force generation, sensing, and maintenance [26], are also spatially segregated in 2D into distinct nanodomains within the IAS and exhibit different diffusion and flow dynamics [25]. Such segregation and distinct molecular dynamics have also been shown with SRM for active and inactive β1 integrins [55], and with SPT for different TGFβ receptors in the IAS [56]. Finally, the functional activity and regional specificity of nonstructural signaling proteins that undergo fast cycles of activation and inactivation as well as engage in transient interactions with their substrates can be studied by SPT. Rho GTPases, a family of signaling proteins that can switch between active and inactive states by exchanging GDP to GTP and hydrolyzing GTP to GDP, including RhoA and Rac1, are important regulators of cell migration and IAS formation during lamellipodial protrusion [11, 57–59]. The constitutive mutant Rac1-Q61L, potentially binding more strongly to downstream effectors, exhibits more immobilizations at the lamellipodium tip compared to wild-type Rac1 (Rac1-WT) [48], showing SPT’s ability to detect transient binding events whereas classical fluorescence microscopy could not reveal enrichment of active Rac1 at the lamellipodium tip [58]. SPT can also be combined with newly developed optogenetic techniques that can photoactivate proteins, making it possible to study their functions in live cells such as Rac1 [60, 61]. By recruiting a Rac1 GEF Tiam1 to the plasma membrane through light-induced binding between Tiam1-CRY2 to membrane-anchored CIBN [62], it was shown that Rac1 photoactivation driving protrusion needs to happen close to the tip of the lamellipodium but not a few microns behind, however, the activation of Rac1 did not significantly increase its immobilization, pointing to a model of short-ranged and shortlived Rac1 activation and downstream signaling that depend on transient interactions and rapid inactivation [48]. Therefore SPT/SRM techniques are complementary to biochemical approaches and have become crucial to establish protein dynamics as readout of protein functions and interactions in complex macromolecular assemblies, having seen applications in diverse fields such as EGFR dimerization [63], TCR signaling [64] and AMPR receptor clustering [65] in addition to integrin adhesions. Here we describe how the combination of SRM/SPT with recent technological advances in genome editing, optogenetic control of signaling pathways and three-dimensional imaging can help us unravel the intricate protein interaction dynamics within the IAS.

92

2

A. V. Radhakrishnan et al.

Materials

2.1 Solutions and Reagents

1. Cell culture tested products to be used for all the preparations. 2. Commercial 0.05% trypsin, 0.02% EDTA solution. 3. Cell culture medium: Dulbecco’s modified Eagle medium, high glucose (DMEM), 10% fetal bovine serum (FBS), GlutaMAX supplement, 100 U/ml penicillin–streptomycin, 1 mM sodium pyruvate, 15 mM HEPES. 4. Purified human fibronectin solution (10 μg/ml) in phosphate buffered saline (PBS). 5. Nitric acid solution (65% wt/wt in water). 6. Trypsin inactivation medium: DMEM, 1 mg/ml soybean trypsin inhibitor, GlutaMAX supplement, 100 U/ml penicillin– streptomycin, 1 mM sodium pyruvate, 15 mM HEPES. 7. Ringer solution: 150 mM NaCl, 5 mM KCl, 2 mM CaCl2, 2 mM MgCl2, 10 mM HEPES 11 mM glucose at pH 7.4. 8. 4% PFA–0.2% glutaraldehyde solution in PBS. 9. 150 mM Glycine in PBS. 10. 0.2% Triton X-100 in PBS. 11. 3% BSA in PBS. 12. Anti-GFP nanobodies coupled to AlexaFluor-647. 13. STORM-adapted buffers (Abbelight).

2.2 Other Consumables and Equipment

1. Nucleofector™ transfection kit for MEF-1 and Nucleofector™ 2b device (Amaxa™) (see Note 1). 2. Marienfield #1.5H coverslips matching to imaging chamber dimensions. 3. 75 cm2 flasks for cell culture. 4. Fiducial markers (e.g., Multicolor fluorescent microbeads Tetraspeck, Invitrogen). 5. Incubator at 37
C supplied with humidified air containing 5% CO2.

2.3 Image Acquisition

In general, this is a close to the state-of-the-art protocol that requires the acquisition of superresolution images. We describe the procedure we use, based on a Nikon inverted microscopy with a TIRF/superresolution module from Abbelight. Other techniques may be available, but the requirement is that the technique of choice is SMLM. 1. Nikon inverted motorized microscope (Nikon Ti). 2. Superresolution optimized 100 1.49 NA PL-APO Nikon objective.

Single-Molecule Integrin Tracking

93

3. Perfect Focus System, Nikon. 4. Lasers (continuous wave) depending on the fluorophore to be imaged (e.g., 405, 488, 561, and 643 nm lasers). Those lasers are dedicated for, photoexcitation photoswitching and photoactivation of the fluorescent proteins and organic fluorophores. For instance, 405 nm laser is used for photoswitching of mEostagged proteins and then mEos is imaged upon illumination with 561 nm laser. The latter should be kept at the same intensity throughout all the experimental conditions to conserve the same pointing accuracy. On the other hand, 643 nm laser is used for imaging organic fluorophores for SAFe. 5. iLas2 illumination control system or equivalent design for TIRF illumination. 6. Optical and optomechanical components including mirrors, dichroic filters, and lenses the corresponding wavelengths used. 7. Scientific complementary metal-oxide semiconductor (sCMOS) camera (e.g., Orca Flash4 by Hamamatsu) or low noise highly sensitive electron-multiplying charge-coupled device (EMCCD) camera (e.g., Evolve, Photometrics). 8. Computer with relevant software for image acquisition, image processing, and visualization (see Note 2). 9. Fast shutter (Uniblitz) or AOTF (AA optoelectronic). 10. Ludin sample holder (Life Imaging Services) or equivalent designs. 11. Supercritical-angle fluorescence emission (SAFe) module (Abbelight) (see Note 3). 12. Buffers for d-STORM acquisition (Abbelight).

3

Methods The following procedures describe sample preparation, image acquisition, and data analysis for different single-molecule imaging techniques: sptPALM, fast sptPALM, sptPALM coupled to optogenetics, and 3D supercritical-angle fluorescence emission (SAFe). In all cases, the protein of interest is coupled to a photoactivatable/ convertible fluorophore or to GFP and is often referred to as “protein of interest.”

3.1

Cell Preparation

We describe here a procedure for mouse embryonic fibroblasts (MEF). Actively dividing immortalized MEFs, cultured in DMEM with 10% FBS in 75 cm2 flasks are detached with trypsin– EDTA solution (1.5 ml). Trypsin is inactivated immediately after detachment by adding serum containing DMEM (5 ml). Cells are then counted and cell suspension volume is adjusted to keep 1–2 million cells per tube per experimental condition.

94

A. V. Radhakrishnan et al.

1. Centrifuge cells at 300  g for 5 min. 2. Resuspended the cell pellet in transfection reagent and mix with the DNA plasmids. 3. For PALM experiments, Cells are usually cotransfected with DNAs encoding for the protein of interest (3–5 μg per condition, e.g., mEos2-β1-integrin, see Note 4), and for a GFP-coupled reporter of the structure of interest (1–2 μg per condition, e.g., GFP-paxillin for adhesive structures, GFP-α-actinin for lamellipodia). 4. For 3D SAFe microscopy, cells are transfected with DNA encoding for the fusion protein (2–3 μg per condition, e.g., Kindlin-GFP, see Note 4). 5. For experiments combining sptPALM and optogenetic activation or inhibition of proteins, cells are cotransfected with DNAs encoding for the protein of interest and for the set of proteins enabling optogenetic control of the target protein. In the example we give here, we track mEos2-Rac1 while triggering its activation with light-induced recruitment to the plasma membrane of a cytosolic guanine nucleotide exchange factor (GEF) of Rac1. Membrane recruitment of a Rac1-GEF (e.g., Tiam1) is triggered by interaction of the cryptochrome CRY2 to its membrane-anchored CIBN partner (Fig. 1d). Cells are cotransfected with DNAs encoding for the protein of interest (~6 μg per condition; e.g., mEos2-Rac1), with the Rac1-GEF fused to the CRY2 cryptochrome (~3 μg per condition; e.g., Tiam1-CRY2-IRFP) and for the GFP-coupled CIBN partner of CRY2 fused with a membrane-anchoring domain CAAX (~3 μg per condition; e.g., CIBN-GFP-CAAX). 6. For all cases, cells are then electroporated with the Nucleofector™ 2b Device using the MEF T-020 program (Lonza Nucleofactor protocol) (see Note 1). 7. After electroporation, cells are replated in a 6-well plate (about 0.3 million cells/well) in preheated culture medium and placed in a 37
C incubator with humidified air containing 5% CO2. 3.2 Cleaning of Glass Substrates

Before matrix coating, coverslips need to be cleaned to ensure absence of nonspecific adsorption of fluorescent materials. 1. Coverslips are aligned in ceramic racks and placed in concentrated nitric acid (65% wt/wt) bath in staining glass boxes overnight. 2. The racks are moved in ultrapure water bath in other staining glass boxes to rinse the coverslips. Six changes of ultrapure water bath every 30 min (or more) are required. 3. Coverslips are quickly rinsed in absolute ethanol. 4. Ceramic racks are placed in a glass beaker, covered with aluminum foil and sterilized in an oven at 240
C for 8 h.

Single-Molecule Integrin Tracking

3.3 Matrix Protein Coatings on Coverslips

95

This procedure is usually done a day before, or the same day of the imaging experiment. 1. Cleaned coverslips are covered with a 10 μg/ml fibronectin solution (1 ml per coverslip) and incubated at 37
C for 1–1.5 h. 2. After incubation, fibronectin solution is aspirated and coverslips are washed three times in PBS, and can be stored in PBS medium at 4
C for further use.

3.4 Sample Preparation

1. Cells are washed twice with PBS after culture medium is removal. 2. Incubate with trypsin–EDTA solution (0.3 ml per well) for 1–3 min at 37
C for detaching cells. 3. Trypsin is inactivated with trypsin inactivation medium (1 ml per well) and cells are counted (use any conventional cell counting method). 4. After centrifugation at 300  g for 5 min, cells are resuspended in Ringer medium (1–2 ml). 5. Plate cells according to specific types of experiment: processes given below pertain to specific experiments involving studies on adhesive structures or lamellipodia.

3.5 sptPALM Experiments 3.5.1 Single-Protein Tracking Inside and Outside IAS 3.5.2 Single-Protein Tracking Inside and Outside Lamellipodia (See Note 5)

1. Plate 50,000 cells per coverslip are plated on the fibronectin coated coverslips in Ringer medium. 2. After 3 h, coverslips can be mounted on an open chamber with approximately 1 ml of Ringer medium for imaging.

1. After resuspension, cells are incubated at 37
C and 5% CO2 for 30 min, to allow integrin turnover after trypsin degradation before plating. 2. 10,000–20,000 cells are loaded directly on fibronectin coated coverslips mounted on an open chamber containing 800 μl of Ringer medium. 3. Imaging is started when cells start spreading and forming lamellipodia, typically 5–10 min after loading.

3.5.3 Supercritical Angle Fluorescence and Localization Microscopy

This step is to be followed after Subheading 3.5.1 or 3.5.2, depending on the structures of interest to be studied. 1. Fix cells using 4% PFA–0.2% Glutaraldehyde in PBS for 15–20 min at RT. 2. Quench fix buffer with 150 mM Glycine in PBS for 20 min. 3. Permeabilize with 0.2% Triton X-100 in PBS for 8 min.

96

A. V. Radhakrishnan et al.

4. Block in 3% BSA in PBS for 1 h. 5. Label cells with anti-GFP nanobodies tagged with AlexaFluor647 overnight at 4
C. 6. Mount on a microscope-adapted device using STORMadapted buffers. 3.5.4 Image Acquisition

1. Using a power meter, the intensities of the appropriate lasers are verified at the objective plane and a calibration chart is obtained to adjust to the desired levels. In the case of mEos2 imaging, we typically use a 561 nm beam power ranging from 4 to 8 mW at objective output to allow single fluorophores to emit on several consecutive recorded planes before photobleaching. 2. Acquisitions are to be made in the Total Internal Reflection Fluorescence (TIRF) mode, using an inverted motorized microscope (Nikon Ti) equipped with a 1.49 NA 100 oil immersion objective and a Perfect Focus System, placed in a thermostatic enclosure at 37
C. 3. Different acquisition modes can be adopted based on types of experiment: (a) Classical sptPALM acquisition: upon observation of the GFP-fused reporter of the structure of interest, cells are selected for single-molecule imaging. A subregion of the camera field of view is selected for recording at high frequency (typically 20–50 Hz, see Note 6), in order to study the diffusive behavior of a protein of interest within a given region of the cell. Several sequences of singlemolecule imaging are then continuously recorded (for a total of 5000–20,000 images, see Note 7), interlaced with images of the fluorescent reporter to monitor possible displacement of the structures of interest such as the IAS and the lamellipodium. (b) Fast sptPALM acquisition: fast sptPALM acquisitions are similar to conventional sptPALM except that they are only possible with a sCMOS camera (Fig. 1c). To reach acquisition frequencies of 333 Hz, it is, nevertheless, necessary to adjust certain acquisition parameters. The subregion must necessarily be positioned in the center of the chip with a size of 500  100 pixels (horizontal  vertical). For the acquisitions, 561 nm laser power is kept at 10 mW. (c) sptPALM acquisition coupled with optogenetics: optogenetic activation of Rac1 results from the light-induced recruitment to the plasma membrane of a Rac1-GEF, Tiam1 (Fig. 1d). This recruitment is visualized by the acquisition of Tiam1-CRY2-IRFP images every 20 s. The photoactivation (~500 ms) is done only after a base

Single-Molecule Integrin Tracking

97

line of four Tiam1-CRY2-IRFP images. Observation of the Rac1 behavior during the perturbation is recorded by sptPALM sequences of mEos2-Rac1-WT (10 s, 500 images, 50 Hz) acquired in between each Tiam1CRY2-IRFP images. The Tiam1-CRY2-IRFP and mEos2-Rac1-WT acquisitions are recorded respectively with 643 and 561 nm lasers with TIRF illumination. Photoactivation of CRY2 is done by a FRAP head, (Roper Scientific) to illuminate with the 488 nm laser in defined regions of interest (ROI). During the photoactivation the ROI is also illuminated with 405 laser to photoconvert mEos2-Rac1-WT that may have been activated by Tiam1 (see Note 8). Rac1 activation triggers lamellipodium protrusion when the ROI of optogenetic activation encompasses the lamellipodial tip, but not if the ROI is located 3 μm behind the lamellipodial tip [48] (Fig. 2a, b). (d) Supercritical angle fluorescence and localization microscopy studies: for STORM imaging, we use GFP-tagged proteins recognized by anti-GFP nanobodies labeled with AlexaFluor-647 [54, 66]. To achieve single-molecule regime in dSTORM acquisitions, a dedicated buffer (Smart kit, Abbelight) is used (see Subheading 2.2). The diffraction limited epifluorescence images are acquired at low illumination irradiance (0.15 kW/cm2), while dSTORM images are obtained using a high illumination irradiance (4 kW/cm2) until a sufficient molecule density is obtained (around 1 molecule per μm2), after which the acquisition can be started. The exposure time is set at 50 ms, the optimal timing with the buffer to capture all emitted photons in a single frame. All the acquisitions are performed using the Nemo software (Abbelight). 3.6 Data Treatment and Analysis 3.6.1 Single-Molecule Localization and Generation of Superresolved Images

This step is performed after the experiment, as treatment and analysis of single-molecule images are very demanding in terms of computing power. In the single-molecule images, single-molecule emissions appear as diffracted bright spots. The original images are subjected to decomposition into wavelet maps using custom algorithms. Further, each fluorophore is localized on segmented images by centroid computation. Custom-watershed algorithm are used to separate close molecules [67]. This procedure provides similar localization precision as more conventional Gaussian fitting but with a tenfold decreased computation time. The single-molecule detections can then be plotted on a single image, forming a superresolved reconstruction (SRR) image (Fig. 1a). The pointing accuracy for the setup can be obtained by repetitively imaging purified mEos2 proteins adsorbed on a glass

98

A. V. Radhakrishnan et al.

Fig. 2 sptPALM coupled with optogenetic activation used to study molecular origins of lamellipodial protrusions in the context of integrin-based migrations. (a) Schematic representation of optogenetic Rac1 activation within the lamellipodial tip (blue) or the region at the rear of the lamellipodium (brown) of a migrating cell. (b) Lamellipodium protrusion after photoactivation at the tip (blue or the rear (brown) as a function of time before and during photoactivation (gray area). (c) Superresolution intensity image obtained from a sptPALM acquisition of mEos2-Rac1-WT in the lamellipodium of a spreading MEF after optogenetic membrane recruitment of Tiam1-CRY2-IRFP (left; same cell as (b)). Corresponding trajectories are shown (right). (Part of this figure is reprinted from ref. 48 with permission from Elsevier)

coverslip. As measured with the Full Width at Half Maximum (FWHM) of the obtained distribution of localizations, the typical resolution of our system is around 20–50 nm. 3.6.2 Analysis of Single-Protein Tracking Experiments

Single-particle detection procedure generates multiple localizations that store the center x and y positions of each molecule. A fluorescent molecule emits fluorescence for a short time in consecutive frames before being bleached, enters a dark state or diffuse out of focus. If an optimal density of fluorophores are activated simultaneously and diffusion does not result in path crossing between neighboring frames for nearby fluorophores, the trajectory of a

Single-Molecule Integrin Tracking

99

molecule can be confidently tracked by linking detected localizations for neighboring frames with a simulated annealing algorithm [68, 69]. While different methods have been used to analyze the diffusive behaviors of molecules [70, 71], we present here a simple but powerful way of analysis based on mean square displacement (MSD) computation [25, 46, 49]. Note should be taken when using this analysis, since it assumes that the molecules in question undergoes 2D membrane diffusion or immobilization on a fixed point in space. Other types of movement such as fast and slow directed motion will need to be treated differently. First, the MSD at each time interval n * Δt is calculated for each trajectory based on the following formula:  2 PN n 2 i¼1 ðx iþn  x i Þ þ y iþn  y i MSDðt ¼ n  Δt Þ ¼ ð1Þ N n where N is the number of data points (frames), x(t) and y(t) are x and y coordinates at timepoint t. Plots of the MSD curves against time interval (Fig. 1b) provide information to distinguish between different modes of motion: for freely diffusing molecules MSD increases linearly with time; molecules in confinement shows linear increase of MSD at short time scales and decreasing slopes of MSD at longer time scales; immobile molecules show a flat MSD close to zero with deviation depending on localization precision (see Note 9). For molecules undergoing fast directed movement, the MSD increases with time squared, but for molecules undergoing slow diffusion, the shape of the curve is close to zero at the short time scale plotted here and cannot be distinguished from immobile ones (see Note 10). Therefore, the diffusive behavior of each molecular trajectory can be characterized by fitting the MSD curve to different diffusion models. Because the later time points of the MSD comes from the average of fewer data points and are less accurate, we routinely discard the latter 20–40% of the MSD for the fitting. For short trajectories (10–20 time points), as obtained with fluorescent proteins, 60–80% of the first MSD points are included for fitting. In the following formula:  4r 2  MSDðt Þ ¼ conf 1  et=τ ð2Þ 3 rconf is the measured confinement radius, and the time constant τ ¼ (r2conf/3Dconf). We consider the molecules as immobile if the calculated D < Dthreshold. The value of Dthreshold is obtained by considering the confinement area is defined by the image spatial resolution. We can distinguish between confined and free diffusion by comparing the time constant τ obtained for each trajectory with half the minimum time interval used to compute the MSD. Confined and free-diffusing trajectories were defined as trajectories with a time constant τ respectively inferior and superior to half the time

100

A. V. Radhakrishnan et al.

interval used to compute the MSD. The apparent diffusion coefficient (D) can be calculated by taking of the slope of a linear fit for the first few points (we usually take the first four points) of the MSD, for free and confined diffusion (see Note 11). Analysis can then be filtered by computing MSD for particular subregions of interest of the raw images. This allows to compare the diffusive behavior of the target protein inside and outside a structure of interest, such as the IAS or the lamellipodium. Based on this approach, we could reveal that integrins are freely diffusing outside IASs, but they undergo immobilization inside IASs due to activation via a FN-integrin-talin tripartite interaction [25]. Moreover, we could also show that Rac1 GTPase is transiently immobilized at the lamellipodium tip, which is correlated with protein activation [48]. 3.6.3 Fast sptPALM Measurements

Transient protein-protein interactions are key components in signaling and regulatory networks [72]. These include signaling cascades such as the Epidermal Growth Factor (EGF) pathway or RhoA/Rac1 GTPase-effector interactions at the lamellipodium. For instance, we found that Rac1-WT is less immobilized at the lamellipodium tip compared to constitutively active Rac1-Q61L (see Note 12). This suggests that dwell times of interaction between activated Rac1-WT and effectors are shorter than our acquisition frequency. To bypass this type of limitations, we can increase the acquisition frequency by changing the acquisition device from an EMCCD to sCMOS camera (Fig. 1c). Conventional sptPALM acquisitions with EMCCD cameras are limited by the readout speed of the camera to about 70 full frames per second [73]. Conversely, sCMOS cameras possess a higher quantum efficiency, a larger field-of-view (FOV) and much faster readout speeds compared to EMCCD cameras. For instance, sCMOS cameras are capable of imaging a 2048  2048 pixel FOV at 100 frames per second; EMCCD cameras are limited to a 512  512 FOV at a maximum speed of 56 FPS [74]. With this, we can increase the acquisition frequency of sptPALM experiments from 50 to 333 Hz to more efficiently capture transient interactions between proteins (Fig. 1c). For a fast PALM experiment, changes in frequency will change the total time of each acquisition (from 30 s (1500 frames) to 12 s (4000 frames)), as well as the resolution of the system (49 nm at 333 Hz for sCMOS vs 59 nm at 50 Hz for EMCDD) and the parameters for MSD computing (duration of the minimum trajectory length for analysis) [48].

3.6.4 sptPALM Acquisition Coupled with Optogenetics

One of the main benefits of optogenetic methods consists in achieving spatiotemporal control of protein activation and interaction with high precision. For instance, using the CRY2-CIBN system coupled to Rac1, we can explore the effects of subcellular Rac1

Single-Molecule Integrin Tracking

101

activation on lamellipodium formation [47, 48]. Two sets of data can be extracted from this particular type of experiments. First, by changing the region of illumination between two subcellular areas, for example the lamellipodium tip and the back of the lamellipodium, we can study the effect of protein activation in specific endogenous pathways/processes, that is, cell migration. We found indeed that Rac1 activation leads to cell protrusion and that Rac1 is specifically activated at the lamellipodium tip and rapidly inactivated beyond the illuminated region. Second, by computing again the MSD for particular subcellular regions, we can extract the diffusive behavior of the protein outside and inside the illuminated region. In the case of Rac1, we found no differences of diffusion/immobilization outside and inside the illuminated region. 3.6.5 Analysis of SAFe Measurements

The nanoscale architecture of adhesive structures [53] has previously been elucidated by an implementation combining photoactivated localization microscopy with single-photon, simultaneous multiphase interferometry (known as iPALM, interferometric photoactivated localization microscopy), which provides sub-20nm 3D protein localization [75]. With comparatively simpler instrumentation, a combination of another two techniques, direct stochastic optical reconstruction microscopy (dSTORM) [76–78] and supercritical-angle fluorescence emission (SAFe) detection, known as direct optical nanoscopy with axially localized detection (DONALD), could be used to study single-molecule distribution with an isotropic 3-D localization precision of 15 nm within specimens 200 nm above the coverslip [79]. After describing the working principle of this method, we outline how we can use a commercial implementation of SAFe (Abbelight) to elucidate the 3D organization of proteins in IAS, more specifically, to study the localization of kindlin, a crucial integrin activator [54]. When light enters from a medium of higher refractive index to a lower one at higher angles of incidence than a critical angle θc, Total Internal Reflection (TIR) happens and if the fluorophore is located in the medium at a distance d comparable to its emission wavelength (0 < d < λem), its near field component is converted into light that propagates beyond the critical angle which is known as supercritical-angle fluorescence emission (SAFe). When the fluorophore is in direct contact with the interface (d ¼ 0), the SAFe intensity is potentially equal to as much as half the amount of all fluorescence intensity emitted into the coverslip, (Fig. 3a). The SAFe intensity decreases with increase in separation (d > 0) [80]. This dependence of the near field SAFe intensity on the distance of fluorophores from the coverslip, can used to estimate their axial positions. The number of photons from under-critical angle fluorescence (UAF) (emitted within a cone that is limited by

102

A. V. Radhakrishnan et al.

Fig. 3 Illustration of SAF measurements in IAS revealing the 3D organization of proteins. (a) Principle of separation of Super-critical (SAF) and Under-critical angle fluorescence (UAF) using SAF ring, UAF intensity is nearly constant both for d ¼ 0 and for 0 < d < λemission, whereas the near field SAF intensity in the annular region decreases with depth almost exponentially. (Adapted from ref. 79). (b) Peak value of the height distribution of K2-WT (~50 nm) is significantly lower than the peak value for K2-ΔPH (PH domain entirely deleted). (c) Images with z (height) distribution of K2-WT and K2-ΔPH reconstructed from the DONALD measurements, detections for z < 50 nm represented in cyan, 50 nm < z < 100 nm in magenta and for z > 100 nm in yellow respectively, notice that K2-ΔPH detections are not enriched into IAS defined by paxillin fluorescence. (d) Schematic to illustrate the result that Kindlin-WT is at the membrane proximal integrin layer, whereas deletion of PH domain causes alteration in the axial distribution peaking at higher z-values, indicating that the PH domain is necessary for Kindlin localization to integrin nanolayer

angle θc), NUAF remains nearly constant as a function of the interface fluorophore distance d. But the number of SAF photons (NSAF) decreases approximately exponentially with d [81]. Hence, the simultaneous measurement of these photon numbers and computation of the fluorophore SAF ratio, NSAF/NUAF for each detected fluorophore can be used to determine the absolute axial position of the fluorophore [79, 82]. Validation of this imaging principle has been also carried out using Origami tetrahedral [82].

Single-Molecule Integrin Tracking

103

With the SAFe module, d-STORM data is acquired and processed with NEMO software (Abbelight). After removing the background signal, molecules were detected and the numbers of SAF and UAF photons were measured to extract the corresponding axial positions as described [54, 79] (Fig. 3a). Lateral drifts are corrected from the localized data using a cross-correlation based algorithm. DONALD is free of any axial drift, as the supercritical emission allows one to extract the absolute axial position of the fluorophore with reference to the coverslip–sample interface [54]. For image display, molecules detected 200 nm above the surface are discarded to improve the contrast in the IAS/plasma membrane layer (Fig. 3c). On the superresolution reconstructed image, each pixel value corresponds to the average axial localization of single molecules detected in this pixel (size: 15 nm) (Fig. 3c). For ease of observation, the obtained images are smoothed using a xy mean filter with a 5  5 kernel. For the curve of occurrence, all detections are included and Z distributions are plotted (Fig. 3b) [54]. Kindlin-2 potentially binds proteins on multiple IAS nanolayers (i.e., integrin, ILK, actin, paxillin) [83–85]. To test whether kindlin-2 function could rely on its shifting from one layer to another in IAS, we used a commercial SAFe module (Abbelight) and the ratiometric approach, for investigating its 3D localization within the nanolayers of IAS (see Note 13). With GFP-paxillin defining the upper bound of the integrin layer (z-peak: 58 nm), GFP-kindlin-2-WT was found concentrated in the integrin layer at the vicinity of the plasma membrane (z-peak: 48.5 nm) (Fig. 3c). Deletion of the PH domain resulted in its upward shift from the integrin layer localization (z-peak: 66 nm) (Fig. 3c, d). Therefore, the PH domain of kindlin is required to target kindlin to the proper functional nanolayer where it will act, namely the integrin layer. 3.6.6 Using Mutants of Integrin/Regulators

In the highly complex and crowed environment of IAS, integrin regulation and activation (see Note 14) are highly dependent on interactions with multiple binding partners, which can be revealed by sptPALM and SRM [23, 86]. To further probe these chains of interactions, we can employ a series of known integrins and regulators mutations which prevent or stabilize interactions with their binding partners. These strategies can be combined with genetic knockout cell models for the protein of interest (see Note 4). Below we give several examples illustrating how this strategy could be used to decipher the sequence of molecular events/interactions leading to integrin activations in mature IAS. 1. Protein mutants can provide additional information on the immobilization and diffusion of proteins linked to different interactions. According to our previous findings, integrin activation is correlated with immobilization of integrin and associated activators, while inhibition is characterized by free-

104

A. V. Radhakrishnan et al.

diffusion. Therefore, diffusive properties are one of the readouts to probe FN/integrins/regulators/F-actin connections. We showed that a constitutively active mutant of β3-integrin (β3-N305T) increased integrin immobilization outside IASs linked to extracellular matrix binding [25]. Moreover, mutants of β3-integrin that prevent either binding to fibronectin (β3D119Y) or talin (β3-Y747A) lead to decreased fractions of immobilization of β3-integrin inside IAS, as well as shorter immobilization times within IASs [25] (see Note 15). Hence, we demonstrated that a FN–integrin–talin tripartite interaction is crucial for a full long-lived integrin immobilization. 2. Protein mutants can help tracing the path used by crucial integrin regulators to reach functional nanodomains and activate integrins. We showed that Talin, for instance, displays almost no membrane free-diffusion inside and outside IASs. Due to the acquisition frequency and TIRF illumination of a sptPALM experiment, it is impossible to detect cytosolic free diffusion of a protein. Therefore, talin is mostly cytosolic and it is not codiffusing with integrins outside or inside IAS. To further dissect the mechanism, we employed two short forms of Talin; the C-terminal THATCH mutant, comprising an actin and vinculin-binding site, did not display membrane freediffusion and has an immobile fraction similar to full-length talin [25].On the other hand, talin head alone exhibited prominent membrane diffusion outside and inside IAS, but no immobilization inside IAS. These differences reflect the autoinhibition of talin in the cytosol, which will likely be relieved inside the IAS. Therefore, the results suggest that talin is recruited to IASs directly from a cytosolic pool, a process mediated by actin or vinculin. A similar rationale was applied to kindlin-2, which can associate to the plasma membrane through multiple phospholipid motifs found on its F0, F1 and PH domain [87]. We found that kindlin, similar to talin, is immobile and enriched in IAS; however, unlike talin, kindlin displays membrane free diffusion outside and inside the IAS [54]. These results suggest that kindlin is recruited inside IAS by membrane diffusion. 3. Protein mutants are useful tools to study signaling proteins that undergo transient interactions with modulators and effectors. For instance, Rho GTPases are crucial for cytoskeleton reorganization in the context of cell migration [88]. Using gain- or loss-of-function mutants, we demonstrated that there is a strong correlation between protein dynamics and protein activation and/or binding states [25, 48]. For example, using sptPALM with loss- or gain-of-function mutants of Rac1, we revealed that Rac1 activation is correlated with its interaction with effectors at the tip of protrusive structures, at remote locations from sites of Rac1 activation (e.g., IAS) [48]. We

Single-Molecule Integrin Tracking

105

showed that a constitutively active Rac1-Q61L mutant exhibits more immobilizations and slower free diffusions at the lamellipodium tip compared to Rac1-WT, while a fast cycling Rac1F28L mutant has a more similar behavior to Rac1-WT. Conversely, an inactive Rac1-T17N mutant displays predominantly free diffusions and no selective immobilizations. We then sought to capture more efficiently these transient immobilizations by performing fastPALM acquisitions at 333 Hz. This resulted in an enhanced difference of diffusive behavior between WT and active Rac1 mutants (Q61L and F28L). Indeed, fast PALM revealed larger fractions of immobilization for the fast cycling F28L mutant, which were not captured at 50 Hz acquisition frequencies. Nonetheless, even a frequency of 333 Hz was too slow to detect the bulk of Rac1-WT immobilizations, suggesting that most Rac1-effector interactions at the lamellipodium tip are transient and less than a few dozen milliseconds long. This correlation between Rac1 activation and immobilization was also shown to occur in IAS, with constitutively active Rac1 mutant G12V displaying increased immobilization in IAS, while inactive mutant T17N was mostly mobile [49].

4

Notes 1. Select the appropriate program in the Amaxa equipment suitable to the cell lines indicated. Also use transfection reagents suitable to the cell lines used. 2. We used Metamorph (Molecular Devices) for 2D acquisition, image processing and visualization. Wavetracer commercial software [89] or free alternatives are available as well. ThunderSTORM [90] for superresolution reconstruction, ICY [91] with spot detector and spot tracker plugins for single-particle detection and tracking [92] are also frequently used. Kymotoolbox ImageJ plug-in is available from Fabrice Cordelie`res [93]. Nemo software (Abbelight) is used for 3D-SAF acquisition and processing. 3. Alternatively, for the optical setup for 3D imaging using DO NALD, we used an Olympus IX83 inverted microscope with an autofocus system. The excitation path was composed of three laser lines: 637, 532 and 405 nm (Errol lasers) and a TIRF module (Errol lasers) used in combination with a matched 390/482/532/640 multiband filter (LF405/488/532/ 635-A-000, Semrock). The fluorescence was collected through an Olympus 100 1.49 NA oil immersion objective lens. The detection path was composed of a SAF module (Abbelight) and a Flash 4 v3 (Hamamatsu).

106

A. V. Radhakrishnan et al.

4. The majority of SPT experiments presented here were performed in cellular systems in which the wild-type form of the studied protein is already expressed at the endogenous level. In that case, transfection with the WT tagged protein (e.g., mEos2, mEos3.2) will lead to overexpression of the target protein. Although this ensures the labeling density necessary for high effective resolution, it can induce functional consequences, besides the inherent experimental variations associated to protein transfection. With the development of genomic editing, it is now possible to engineer cellular systems with CRISPR-Cas9 to endogenously express a specific protein labeled with a tag compatible with single-protein tracking and superresolution microscopy [94]. This produces systems with stable expression patterns of tagged protein expression and removes the variability associated to plasmid transfection. In the context of genetic modifications, studies on protein function can also be performed on a background where the endogenous protein has been deleted. This is of particular interest when employing different mutants to further characterize protein function and organization. Knockout cell models, either from animal cell lines or obtained using CRISPR/Cas-9, allow for linking diffusive behavior and protein function. The ability to rescue or not a particular phenotype caused by endogenous protein deletion can reveal the importance of specific dynamics and domains present in IAS proteins. For instance, fibroblasts double KO for kindlin-1 and kindlin-2 are unable to spread and to form IAS. By studying the diffusive behavior of kindlin2 mutants and their ability to trigger cell spreading or the formation of IAS in kindlin-1,2 double KO (see Note 13), we demonstrated that kindlin membrane free-diffusion is key to trigger integrin activation [54]. KO cell models can be applied for other dynamic structures in cell migration, such as the lamellipodium. Indeed, we have also characterized the interaction between Rac1 and the WAVE complex in cells where Sra1 and PIR121 had been genetically disrupted by CRISPR/Cas9 (Sra1/PIR121-KO) [95]; Sra1 is an element of WAVE protein complex. By rescuing these cell lines with constitutively activated Sra1 mutants, we could modulate WAVE complex activation and interactions with Rac1 [48]. Consistent with other results, we showed that immobilizations of activated Rac1 depend on interactions with WAVE complex at the lamellipodium tip. 5. To test that we were not performing experiments on cells having aberrant protrusive behaviors, we also measured rates of lamellipodium protrusion for the cells used for sptPALM analysis. The rates of protrusion for the cells analyzed in different conditions were not dramatically different among another.

Single-Molecule Integrin Tracking

107

This suggests that, in our acquisition conditions, levels of expression of distinct mEos2-fused proteins were not triggering dramatic effects on lamellipodium protrusion. 6. Fast diffusion of proteins requires tracking with a high temporal resolution. We routinely use 50 Hz imaging frequency on a center quadrant of 512-by-512 pixel EMCCD camera. While EMCCD provides high signal to noise ratio with weak single fluorophores, faster imaging speed can be achieved by reducing the area of imaging on the camera chip, which is a limiting factor for the speed of data transfer. Alternatively, sCMOS cameras can be used, as further elaborated in the Subheading 3.6.3. 7. For a typical PALM acquisition, 20,000 or more tracks can be collected from a cell, in batches of 4000 frames of recording. While longer acquisition time can generate more trajectories, phototoxicity may change the behavior of cells and protein dynamics, the experimenter should control the laser intensity and imaging time to optimize according to cell type and protein studied. 8. Using the 405 laser only in the ROI prevents unwanted Tiam1 photoactivation outside the ROI, since CRY2 is sensitive to blue light (405–488 nm). 9. Molecules may transition between different states of diffusion and immobilization. Longer trajectories may include mixed diffusion states in the same trajectory, thus underestimating the diffusion coefficient when calculating it from the MSD curve. The experimenter should examine the individual trajectories and determine whether the protein of interest undergoes frequent transitions between states. If this is the case, acquiring shorter tracks or breaking the trajectories into short pieces can help mitigate the issue, albeit increasing the variation of MSD curve. Alternatively, a rolling MSD analysis can help identify different diffusion states and segment accordingly. 10. A single fluorophore can undergo “blinking” for tens of seconds. Therefore, it is possible to track a slowly moving fluorophore for a prolonged duration by reconnecting emissions from different time points. Directed movements of proteins can thereby be analyzed by tracking the position of individual proteins over time. In addition, to understand slow dynamics, kymographs (position over time) can be created in a region of interest by a straight line or a segmented line is typically drawn directly on the superresolution time-lapse movie. By merging several superresolved reconstruction images (e.g., 25 images) obtained from 50 Hz single-molecule imaging, an SR time lapse movie (e.g., 2 Hz) is generated. From this, different

108

A. V. Radhakrishnan et al.

parameters such as the speed, direction, and dwell time of proteins are extracted. 11. To make sure that the observed dynamics are specific to the protein of interest and not arising from any protein density distributions, we also analyse trajectories of mEOS2 fused to the trans-membrane domain of the PDGF receptor (mEOS2– TM) or anchored to inner leaflet lipids (CAAX–mEOS2; (C; cysteine; A, aliphatic residue; X, any amino acid). 12. Since high levels of expression of Rho GTPase mutants might affect lamellipodia formation, dynamics and morphologies [96, 97], acquisitions are to be performed only on cells able to spread and polarize, and in the absence of dramatic phenotypes such as (1) being unable to spread but forming membrane tubules (high levels of RhoA-Q63L expression), (2) bearing numerous lamellipodia (high levels of Rac1Q61L). Analyze cells displaying an active, protrusive lamellipodium, in phase 2 of spreading (according to [14, 15]). 13. Kindlin-1, Kindlin-2 double knockout cells (KindKo) cell line was provided by Reinhard F€assler (Max Planck Institute of Biochemistry, Martinsried) and are described elsewhere [83]. Absence of mycoplasma contamination was assessed using the MycoAlert detection kit (Lonza). 14. Integrin activation can be induced by replacing the cell media (Ringer+glucose) with a Ringer+glucose solution with MnCl2 at 5 mM, at least 5 min before acquisition. 15. Differences in the dynamic behavior of WT and mutant β3integrins were not correlated with their surface expression levels measured by fluorescence-activated cell sorting (FACS) or using the density of mEOS2 detections. Dynamics of WT and mutant β3-integrins were not affected by the probes used to localize IAS (GFP–paxillin or GFP–VASP), or by the presence of endogenous β3-integrins, as indicated by experiments performed in β3-integrin/ MEFs. References 1. Montell DJ, Rorth P, Spradling AC (1992) Slow border cells, a locus required for a developmentally regulated cell migration during oogenesis, encodes Drosophila C/EBP. Cell 71(1):51–62. https://doi.org/10.1016/ 0092-8674(92)90265-e 2. Grinnell F (1992) Wound repair, keratinocyte activation and integrin modulation. J Cell Sci 101(Pt 1):1–5 3. Abreu-Blanco MT, Verboon JM, Liu R, Watts JJ, Parkhurst SM (2012) Drosophila embryos close epithelial wounds using a combination of

cellular protrusions and an actomyosin purse string. J Cell Sci 125(Pt 24):5984–5997. https://doi.org/10.1242/jcs.109066 4. Friedl P, Locker J, Sahai E, Segall JE (2012) Classifying collective cancer cell invasion. Nat Cell Biol 14(8):777–783. https://doi.org/10. 1038/ncb2548 5. Pollard TD, Borisy GG (2003) Cellular motility driven by assembly and disassembly of actin filaments. Cell 112(4):453–465. https://doi. org/10.1016/s0092-8674(03)00120-x

Single-Molecule Integrin Tracking 6. Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, Parsons JT, Horwitz AR (2003) Cell migration: integrating signals from front to back. Science 302 (5651):1704–1709. https://doi.org/10. 1126/science.1092053 7. Ridley AJ (2011) Life at the leading edge. Cell 145(7):1012–1022. https://doi.org/10. 1016/j.cell.2011.06.010 8. Svitkina TM, Borisy GG (1999) Arp2/3 complex and actin depolymerizing factor/cofilin in dendritic organization and treadmilling of actin filament array in lamellipodia. J Cell Biol 145(5):1009–1026. https://doi.org/10. 1083/jcb.145.5.1009 9. Chen Z, Borek D, Padrick SB, Gomez TS, Metlagel Z, Ismail AM, Umetani J, Billadeau DD, Otwinowski Z, Rosen MK (2010) Structure and control of the actin regulatory WAVE complex. Nature 468(7323):533–538. https://doi.org/10.1038/nature09623 10. Ridley AJ, Paterson HF, Johnston CL, Diekmann D, Hall A (1992) The small GTP-binding protein rac regulates growth factor-induced membrane ruffling. Cell 70 (3):401–410. https://doi.org/10.1016/ 0092-8674(92)90164-8 11. Machacek M, Hodgson L, Welch C, Elliott H, Pertz O, Nalbant P, Abell A, Johnson GL, Hahn KM, Danuser G (2009) Coordination of Rho GTPase activities during cell protrusion. Nature 461(7260):99–103. https://doi. org/10.1038/nature08242 12. Ponti A, Machacek M, Gupton SL, WatermanStorer CM, Danuser G (2004) Two distinct actin networks drive the protrusion of migrating cells. Science 305(5691):1782–1786. https://doi.org/10.1126/science.1100533 13. Case LB, Baird MA, Shtengel G, Campbell SL, Hess HF, Davidson MW, Waterman CM (2015) Molecular mechanism of vinculin activation and nanoscale spatial organization in focal adhesions. Nat Cell Biol 17(7):880–892. https://doi.org/10.1038/ncb3180 14. Giannone G, Dubin-Thaler BJ, Dobereiner HG, Kieffer N, Bresnick AR, Sheetz MP (2004) Periodic lamellipodial contractions correlate with rearward actin waves. Cell 116 (3):431–443. https://doi.org/10.1016/ s0092-8674(04)00058-3 15. Giannone G, Dubin-Thaler BJ, Rossier O, Cai Y, Chaga O, Jiang G, Beaver W, Dobereiner HG, Freund Y, Borisy G, Sheetz MP (2007) Lamellipodial actin mechanically links myosin activity with adhesion-site formation. Cell 128(3):561–575. https://doi.org/10. 1016/j.cell.2006.12.039

109

16. Hotulainen P, Lappalainen P (2006) Stress fibers are generated by two distinct actin assembly mechanisms in motile cells. J Cell Biol 173 (3):383–394. https://doi.org/10.1083/jcb. 200511093 17. Koestler SA, Auinger S, Vinzenz M, Rottner K, Small JV (2008) Differentially oriented populations of actin filaments generated in lamellipodia collaborate in pushing and pausing at the cell front. Nat Cell Biol 10(3):306–313. https://doi.org/10.1038/ncb1692 18. Vicente-Manzanares M, Ma X, Adelstein RS, Horwitz AR (2009) Non-muscle myosin II takes centre stage in cell adhesion and migration. Nat Rev Mol Cell Biol 10(11):778–790. https://doi.org/10.1038/nrm2786 19. Hynes RO (2002) Integrins: bidirectional, allosteric signaling machines. Cell 110 (6):673–687. https://doi.org/10.1016/ s0092-8674(02)00971-6 20. Geiger B, Spatz JP, Bershadsky AD (2009) Environmental sensing through focal adhesions. Nat Rev Mol Cell Biol 10(1):21–33. https://doi.org/10.1038/nrm2593 21. Humphrey JD, Dufresne ER, Schwartz MA (2014) Mechanotransduction and extracellular matrix homeostasis. Nat Rev Mol Cell Biol 15 (12):802–812. https://doi.org/10.1038/ nrm3896 22. Legate KR, Fassler R (2009) Mechanisms that regulate adaptor binding to beta-integrin cytoplasmic tails. J Cell Sci 122(Pt 2):187–198. https://doi.org/10.1242/jcs.041624 23. Orre T, Rossier O, Giannone G (2019) The inner life of integrin adhesion sites: from single molecules to functional macromolecular complexes. Exp Cell Res 379(2):235–244. https:// doi.org/10.1016/j.yexcr.2019.03.036 24. Sun Z, Guo SS, Fassler R (2016) Integrinmediated mechanotransduction. J Cell Biol 215(4):445–456. https://doi.org/10.1083/ jcb.201609037 25. Rossier O, Octeau V, Sibarita JB, Leduc C, Tessier B, Nair D, Gatterdam V, Destaing O, Albiges-Rizo C, Tampe R, Cognet L, Choquet D, Lounis B, Giannone G (2012) Integrins beta1 and beta3 exhibit distinct dynamic nanoscale organizations inside focal adhesions. Nat Cell Biol 14(10):1057–1067. https://doi.org/10.1038/ncb2588 26. Schiller HB, Hermann MR, Polleux J, Vignaud T, Zanivan S, Friedel CC, Sun Z, Raducanu A, Gottschalk KE, Thery M, Mann M, Fassler R (2013) β1- and αv-class integrins cooperate to regulate myosin II during rigidity sensing of fibronectin-based

110

A. V. Radhakrishnan et al.

microenvironments. Nat Cell Biol 15 (6):625–636. https://doi.org/10.1038/ ncb2747 27. Moser M, Legate KR, Zent R, Fassler R (2009) The tail of integrins, talin, and kindlins. Science 324(5929):895–899. https://doi.org/10. 1126/science.1163865 28. Shattil SJ, Kim C, Ginsberg MH (2010) The final steps of integrin activation: the end game. Nat Rev Mol Cell Biol 11(4):288–300. https://doi.org/10.1038/nrm2871 29. Bouvard D, Pouwels J, De Franceschi N, Ivaska J (2013) Integrin inactivators: balancing cellular functions in vitro and in vivo. Nat Rev Mol Cell Biol 14(7):430–442. https://doi.org/10. 1038/nrm3599 30. Humphries JD, Byron A, Bass MD, Craig SE, Pinney JW, Knight D, Humphries MJ (2009) Proteomic analysis of integrin-associated complexes identifies RCC2 as a dual regulator of Rac1 and Arf6. Sci Signal 2(87):ra51. https:// doi.org/10.1126/scisignal.2000396 31. Schiller HB, Friedel CC, Boulegue C, Fassler R (2011) Quantitative proteomics of the integrin adhesome show a myosin II-dependent recruitment of LIM domain proteins. EMBO Rep 12 (3):259–266. https://doi.org/10.1038/ embor.2011.5 32. Kuo JC, Han X, Hsiao CT, Yates JR 3rd, Waterman CM (2011) Analysis of the myosinII-responsive focal adhesion proteome reveals a role for beta-Pix in negative regulation of focal adhesion maturation. Nat Cell Biol 13 (4):383–393. https://doi.org/10.1038/ ncb2216 33. Zaidel-Bar R, Itzkovitz S, Ma’ayan A, Iyengar R, Geiger B (2007) Functional atlas of the integrin adhesome. Nat Cell Biol 9 (8):858–867. https://doi.org/10.1038/ ncb0807-858 34. Horton ER, Byron A, Askari JA, Ng DHJ, Millon-Fremillon A, Robertson J, Koper EJ, Paul NR, Warwood S, Knight D, Humphries JD, Humphries MJ (2015) Definition of a consensus integrin adhesome and its dynamics during adhesion complex assembly and disassembly. Nat Cell Biol 17 (12):1577–1587. https://doi.org/10.1038/ ncb3257 35. Byron A, Humphries JD, Craig SE, Knight D, Humphries MJ (2012) Proteomic analysis of alpha4beta1 integrin adhesion complexes reveals alpha-subunit-dependent protein recruitment. Proteomics 12(13):2107–2114. https://doi.org/10.1002/pmic.201100487 36. Robertson J, Jacquemet G, Byron A, Jones MC, Warwood S, Selley JN, Knight D,

Humphries JD, Humphries MJ (2015) Defining the phospho-adhesome through the phosphoproteomic analysis of integrin signalling. Nat Commun 6:6265. https://doi.org/10. 1038/ncomms7265 37. Kechagia JZ, Ivaska J, Roca-Cusachs P (2019) Integrins as biomechanical sensors of the microenvironment. Nat Rev Mol Cell Biol 20 (8):457–473. https://doi.org/10.1038/ s41580-019-0134-2 38. Evans EA, Calderwood DA (2007) Forces and bond dynamics in cell adhesion. Science 316 (5828):1148–1153. https://doi.org/10. 1126/science.1137592 39. Moore SW, Roca-Cusachs P, Sheetz MP (2010) Stretchy proteins on stretchy substrates: the important elements of integrinmediated rigidity sensing. Dev Cell 19 (2):194–206. https://doi.org/10.1016/j. devcel.2010.07.018 40. Sun Z, Costell M, Fassler R (2019) Integrin activation by talin, kindlin and mechanical forces. Nat Cell Biol 21(1):25–31. https:// doi.org/10.1038/s41556-018-0234-9 41. Hoffmann JE, Fermin Y, Stricker RL, Ickstadt K, Zamir E (2014) Symmetric exchange of multi-protein building blocks between stationary focal adhesions and the cytosol. Elife 3:e02257. https://doi.org/10. 7554/eLife.02257 42. Han SJ, Dean KM, Whitewood J, Bachir A, Guttierrez E, Groisman A, Horwitz AR, Goult BT, Danuser G (2019) Formation of talin-vinculin pre-complexes dictates maturation of nascent adhesions by accelerated force transmission and vinculin recruitment. bioRxiv:735183. https://doi.org/10.1101/ 735183 43. Tadokoro S, Shattil SJ, Eto K, Tai V, Liddington RC, de Pereda JM, Ginsberg MH, Calderwood DA (2003) Talin binding to integrin beta tails: a final common step in integrin activation. Science 302(5642):103–106. https:// doi.org/10.1126/science.1086652 44. Woodside DG, Obergfell A, Talapatra A, Calderwood DA, Shattil SJ, Ginsberg MH (2002) The N-terminal SH2 domains of Syk and ZAP-70 mediate phosphotyrosineindependent binding to integrin beta cytoplasmic domains. J Biol Chem 277 (42):39401–39408. https://doi.org/10. 1074/jbc.M207657200 45. Sungkaworn T, Jobin ML, Burnecki K, Weron A, Lohse MJ, Calebiro D (2017) Single-molecule imaging reveals receptor-G protein interactions at cell surface hot spots. Nature 550(7677):543–547. https://doi. org/10.1038/nature24264

Single-Molecule Integrin Tracking 46. Chazeau A, Mehidi A, Nair D, Gautier JJ, Leduc C, Chamma I, Kage F, Kechkar A, Thoumine O, Rottner K, Choquet D, Gautreau A, Sibarita JB, Giannone G (2014) Nanoscale segregation of actin nucleation and elongation factors determines dendritic spine protrusion. EMBO J 33(23):2745–2764. https://doi.org/10.15252/embj.201488837 47. Remorino A, De Beco S, Cayrac F, Di Federico F, Cornilleau G, Gautreau A, Parrini MC, Masson JB, Dahan M, Coppey M (2017) Gradients of Rac1 nanoclusters support spatial patterns of Rac1 signaling. Cell Rep 21 (7):1922–1935. https://doi.org/10.1016/j. celrep.2017.10.069 48. Mehidi A, Rossier O, Schaks M, Chazeau A, Biname F, Remorino A, Coppey M, Karatas Z, Sibarita JB, Rottner K, Moreau V, Giannone G (2019) Transient activations of Rac1 at the lamellipodium tip trigger membrane protrusion. Curr Biol 29(17):2852–2866. e2855. https://doi.org/10.1016/j.cub.2019.07.035 49. Shibata AC, Chen LH, Nagai R, Ishidate F, Chadda R, Miwa Y, Naruse K, Shirai YM, Fujiwara TK, Kusumi A (2013) Rac1 recruitment to the archipelago structure of the focal adhesion through the fluid membrane as revealed by single-molecule analysis. Cytoskeleton 70 (3):161–177. https://doi.org/10.1002/cm. 21097 50. Tsunoyama TA, Watanabe Y, Goto J, Naito K, Kasai RS, Suzuki KGN, Fujiwara TK, Kusumi A (2018) Super-long single-molecule tracking reveals dynamic-anchorage-induced integrin function. Nat Chem Biol 14(5):497–506. https://doi.org/10.1038/s41589-018-00325 51. Leduc C, Si S, Gautier J, Soto-Ribeiro M, Wehrle-Haller B, Gautreau A, Giannone G, Cognet L, Lounis B (2013) A highly specific gold nanoprobe for live-cell single-molecule imaging. Nano Lett 13(4):1489–1494. https://doi.org/10.1021/nl304561g 52. Paszek MJ, DuFort CC, Rossier O, Bainer R, Mouw JK, Godula K, Hudak JE, Lakins JN, Wijekoon AC, Cassereau L, Rubashkin MG, Magbanua MJ, Thorn KS, Davidson MW, Rugo HS, Park JW, Hammer DA, Giannone G, Bertozzi CR, Weaver VM (2014) The cancer glycocalyx mechanically primes integrin-mediated growth and survival. Nature 511(7509):319–325. https://doi.org/ 10.1038/nature13535 53. Kanchanawong P, Shtengel G, Pasapera AM, Ramko EB, Davidson MW, Hess HF, Waterman CM (2010) Nanoscale architecture of integrin-based cell adhesions. Nature 468

111

(7323):580–584. https://doi.org/10.1038/ nature09621 54. Orre, T. et al. (2020) ‘Molecular motion and tridimensional nanoscale localization of kindlin control integrin activation in focal adhesions’, bioRxiv, p. 2020.08.06.239731. doi: 10.1101/2020.08.06.239731 55. Spiess M, Hernandez-Varas P, Oddone A, Olofsson H, Blom H, Waithe D, Lock JG, Lakadamyali M, Stromblad S (2018) Active and inactive beta1 integrins segregate into distinct nanoclusters in focal adhesions. J Cell Biol 217(6):1929–1940. https://doi.org/10. 1083/jcb.201707075 56. Rys JP, DuFort CC, Monteiro DA, Baird MA, Oses-Prieto JA, Chand S, Burlingame AL, Davidson MW, Alliston TN (2015) Discrete spatial organization of TGFbeta receptors couples receptor multimerization and signaling to cellular tension. Elife 4:e09300. https://doi. org/10.7554/eLife.09300 57. Kiosses WB, Shattil SJ, Pampori N, Schwartz MA (2001) Rac recruits high-affinity integrin alphavbeta3 to lamellipodia in endothelial cell migration. Nat Cell Biol 3(3):316–320. https://doi.org/10.1038/35060120 58. Del Pozo MA, Kiosses WB, Alderson NB, Meller N, Hahn KM, Schwartz MA (2002) Integrins regulate GTP-Rac localized effector interactions through dissociation of Rho-GDI. Nat Cell Biol 4(3):232–239. https://doi.org/ 10.1038/ncb759 59. Pertz O, Hodgson L, Klemke RL, Hahn KM (2006) Spatiotemporal dynamics of RhoA activity in migrating cells. Nature 440 (7087):1069–1072. https://doi.org/10. 1038/nature04665 60. Wu YI, Frey D, Lungu OI, Jaehrig A, Schlichting I, Kuhlman B, Hahn KM (2009) A genetically encoded photoactivatable Rac controls the motility of living cells. Nature 461(7260):104–108. https://doi.org/10. 1038/nature08241 61. Yazawa M, Sadaghiani AM, Hsueh B, Dolmetsch RE (2009) Induction of proteinprotein interactions in live cells using light. Nat Biotechnol 27(10):941–945. https://doi. org/10.1038/nbt.1569 62. Valon L, Etoc F, Remorino A, di Pietro F, Morin X, Dahan M, Coppey M (2015) Predictive spatiotemporal manipulation of signaling perturbations using optogenetics. Biophys J 109(9):1785–1797. https://doi.org/10. 1016/j.bpj.2015.08.042 63. Kasai RS, Suzuki KG, Prossnitz ER, KoyamaHonda I, Nakada C, Fujiwara TK, Kusumi A (2011) Full characterization of GPCR

112

A. V. Radhakrishnan et al.

monomer-dimer dynamic equilibrium by single molecule imaging. J Cell Biol 192 (3):463–480. https://doi.org/10.1083/jcb. 201009128 64. Dustin ML, Depoil D (2011) New insights into the T cell synapse from single molecule techniques. Nat Rev Immunol 11 (10):672–684. https://doi.org/10.1038/ nri3066 65. Opazo P, Sainlos M, Choquet D (2012) Regulation of AMPA receptor surface diffusion by PSD-95 slots. Curr Opin Neurobiol 22 (3):453–460. https://doi.org/10.1016/j. conb.2011.10.010 66. Chamma I, Rossier O, Giannone G, Thoumine O, Sainlos M (2017) Optimized labeling of membrane proteins for applications to super-resolution imaging in confined cellular environments using monomeric streptavidin. Nat Protoc 12(4):748–763. https://doi. org/10.1038/nprot.2017.010 67. Izeddin I, Boulanger J, Racine V, Specht CG, Kechkar A, Nair D, Triller A, Choquet D, Dahan M, Sibarita JB (2012) Wavelet analysis for single molecule localization microscopy. Opt Express 20(3):2081–2095. https://doi. org/10.1364/OE.20.002081 68. Racine V, Hertzog A, Jouanneau J, Salamero J, Kervrann C, Sibarita J (2006) Multiple-target tracking of 3D fluorescent objects based on simulated annealing. In: Third IEEE international symposium on biomedical imaging: nano to macro, 6–9 Apr 2006, pp 1020–1023. https://doi.org/10.1109/ISBI. 2006.1625094 69. Racine V, Sachse M, Salamero J, Fraisier V, Trubuil A, Sibarita JB (2007) Visualization and quantification of vesicle trafficking on a three-dimensional cytoskeleton network in living cells. J Microsc 225(Pt 3):214–228. https://doi.org/10.1111/j.1365-2818.2007. 01723.x 70. Das S, Yin T, Yang Q, Zhang J, Wu YI, Yu J (2015) Single-molecule tracking of small GTPase Rac1 uncovers spatial regulation of membrane translocation and mechanism for polarized signaling. Proc Natl Acad Sci U S A 112(3):E267–E276. https://doi.org/10. 1073/pnas.1409667112 71. El Beheiry M, Dahan M, Masson JB (2015) InferenceMAP: mapping of single-molecule dynamics with Bayesian inference. Nat Methods 12(7):594–595. https://doi.org/10. 1038/nmeth.3441 72. De Keersmaecker H, Camacho R, Rantasa DM, Fron E, Uji IH, Mizuno H, Rocha S (2018) Mapping transient protein interactions at the nanoscale in living mammalian cells. ACS

Nano 12(10):9842–9854. https://doi.org/ 10.1021/acsnano.8b01227 73. Huang F, Hartwich TM, Rivera-Molina FE, Lin Y, Duim WC, Long JJ, Uchil PD, Myers JR, Baird MA, Mothes W, Davidson MW, Toomre D, Bewersdorf J (2013) Video-rate nanoscopy using sCMOS camera-specific single-molecule localization algorithms. Nat Methods 10(7):653–658. https://doi.org/ 10.1038/nmeth.2488 74. Almada P, Culley S, Henriques R (2015) PALM and STORM: into large fields and high-throughput microscopy with sCMOS detectors. Methods 88:109–121. https://doi. org/10.1016/j.ymeth.2015.06.004 75. Shtengel G, Galbraith JA, Galbraith CG, Lippincott-Schwartz J, Gillette JM, Manley S, Sougrat R, Waterman CM, Kanchanawong P, Davidson MW, Fetter RD, Hess HF (2009) Interferometric fluorescent super-resolution microscopy resolves 3D cellular ultrastructure. Proc Natl Acad Sci U S A 106(9):3125–3130. https://doi.org/10.1073/pnas.0813131106 76. Rust MJ, Bates M, Zhuang X (2006) Subdiffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat Methods 3(10):793–795. https://doi.org/ 10.1038/nmeth929 77. van de Linde S, Loschberger A, Klein T, Heidbreder M, Wolter S, Heilemann M, Sauer M (2011) Direct stochastic optical reconstruction microscopy with standard fluorescent probes. Nat Protoc 6(7):991–1009. https:// doi.org/10.1038/nprot.2011.336 78. Huang B, Wang W, Bates M, Zhuang X (2008) Three-dimensional super-resolution imaging by stochastic optical reconstruction microscopy. Science 319(5864):810–813. https:// doi.org/10.1126/science.1153529 79. Bourg N, Mayet C, Dupuis G, Barroca T, Bon P, Le´cart S, Fort E, Le´veˆque-Fort S (2015) Direct optical nanoscopy with axially localized detection. Nat Photonics 9 (9):587–593. https://doi.org/10.1038/ nphoton.2015.132 80. Fort E, Gre´sillon S (2007) Surface enhanced fluorescence. J Phys D Appl Phys 41 (1):013001. https://doi.org/10.1088/00223727/41/1/013001 81. Ruckstuhl T, Rankl M, Seeger S (2003) Highly sensitive biosensing using a supercritical angle fluorescence (SAF) instrument. Biosens Bioelectron 18(9):1193–1199. https://doi.org/ 10.1016/s0956-5663(02)00239-7 82. Deschamps J, Mund M, Ries J (2014) 3D superresolution microscopy by supercritical angle detection. Opt Express 22

Single-Molecule Integrin Tracking (23):29081–29091. https://doi.org/10. 1364/OE.22.029081 83. Theodosiou M, Widmaier M, Bottcher RT, Rognoni E, Veelders M, Bharadwaj M, Lambacher A, Austen K, Muller DJ, Zent R, Fassler R (2016) Kindlin-2 cooperates with talin to activate integrins and induces cell spreading by directly binding paxillin. Elife 5: e10130. https://doi.org/10.7554/eLife. 10130 84. Hirbawi J, Bialkowska K, Bledzka KM, Liu J, Fukuda K, Qin J, Plow EF (2017) The extreme C-terminal region of kindlin-2 is critical to its regulation of integrin activation. J Biol Chem 292(34):14258–14269. https://doi.org/10. 1074/jbc.M117.776195 85. Shi X, Ma YQ, Tu Y, Chen K, Wu S, Fukuda K, Qin J, Plow EF, Wu C (2007) The MIG-2/ integrin interaction strengthens cell-matrix adhesion and modulates cell motility. J Biol Chem 282(28):20455–20466. https://doi. org/10.1074/jbc.M611680200 86. Rossier O, Giannone G (2016) The journey of integrins and partners in a complex interactions landscape studied by super-resolution microscopy and single protein tracking. Exp Cell Res 343(1):28–34. https://doi.org/10.1016/j. yexcr.2015.11.004 87. Calderwood DA, Campbell ID, Critchley DR (2013) Talins and kindlins: partners in integrin-mediated adhesion. Nat Rev Mol Cell Biol 14(8):503–517. https://doi.org/10. 1038/nrm3624 88. Heasman SJ, Ridley AJ (2008) Mammalian Rho GTPases: new insights into their functions from in vivo studies. Nat Rev Mol Cell Biol 9 (9):690–701. https://doi.org/10.1038/ nrm2476 89. Kechkar A, Nair D, Heilemann M, Choquet D, Sibarita JB (2013) Real-time analysis and visualization for single-molecule based superresolution microscopy. PLoS One 8(4): e62918. https://doi.org/10.1371/journal. pone.0062918 90. Ovesny M, Krizek P, Borkovec J, Svindrych Z, Hagen GM (2014) ThunderSTORM: a

113

comprehensive ImageJ plug-in for PALM and STORM data analysis and super-resolution imaging. Bioinformatics 30(16):2389–2390. https://doi.org/10.1093/bioinformatics/ btu202 91. de Chaumont F, Dallongeville S, Chenouard N, Herve N, Pop S, Provoost T, Meas-Yedid V, Pankajakshan P, Lecomte T, Le Montagner Y, Lagache T, Dufour A, OlivoMarin JC (2012) Icy: an open bioimage informatics platform for extended reproducible research. Nat Methods 9(7):690–696. https://doi.org/10.1038/nmeth.2075 92. Chenouard N, Bloch I, Olivo-Marin JC (2013) Multiple hypothesis tracking for cluttered biological image sequences. IEEE Trans Pattern Anal Mach Intell 35(11):2736–3750. https://doi.org/10.1109/TPAMI.2013.97 93. Zala D, Hinckelmann MV, Yu H, Lyra da Cunha MM, Liot G, Cordelieres FP, Marco S, Saudou F (2013) Vesicular glycolysis provides on-board energy for fast axonal transport. Cell 152(3):479–491. https://doi.org/10.1016/j. cell.2012.12.029 94. Khan AO, Simms VA, Pike JA, Thomas SG, Morgan NV (2017) CRISPR -Cas9 mediated labelling allows for single molecule imaging and resolution. Sci Rep 7(1):8450. https:// doi.org/10.1038/s41598-017-08493-x 95. Schaks M, Singh SP, Kage F, Thomason P, Klunemann T, Steffen A, Blankenfeldt W, Stradal TE, Insall RH, Rottner K (2018) Distinct interaction sites of Rac GTPase with WAVE regulatory complex have non-redundant functions in vivo. Curr Biol 28(22):3674–3684. e3676. https://doi.org/10.1016/j.cub.2018. 10.002 96. Pertz O (2010) Spatio-temporal Rho GTPase signaling – where are we now? J Cell Sci 123 (Pt 11):1841–1850. https://doi.org/10. 1242/jcs.064345 97. Cherfils J, Zeghouf M (2013) Regulation of small GTPases by GEFs, GAPs, and GDIs. Physiol Rev 93(1):269–309. https://doi.org/ 10.1152/physrev.00003.2012

Chapter 9 Biochemical Characterization of the Integrin Interactome Rejina B. Khan, Lorena Varela, Alana R. Cowell, and Benjamin T. Goult Abstract More than 250 proteins are associated with the formation of integrin adhesion complexes involving a vast number of complex interactions between them. These interactions enable adhesions to serve as dynamic and diverse mechanosignaling centers. Our laboratory focuses on the biochemical and structural study of these interactions to help unpick this complex network. Here, we describe the general pipeline of biochemical assays and methods we use. The chapter is split into two sections: (1) protein production and characterization and (2) biochemical assays for the characterization of binding between full-length proteins and/or specific regions of proteins with other proteins, peptides, and phospholipids. The suite of assays we use routinely includes circular dichroism (CD) and nuclear magnetic resonance (NMR) spectroscopy for sample quality assessment, prior to biochemical analysis using NMR, fluorescence polarization (FP), microscale thermophoresis (MST), size-exclusion chromatography multiangle light scattering (SEC-MALS), and pulldown/cosedimentation-based approaches. The results of our analysis feed into in vivo studies that allow for the elucidation of the biological role of each interaction. Key words Integrin, Adhesome, Interaction, Biochemistry

1

Introduction Cell–matrix adhesions are large dynamic macromolecule complexes that assemble on the cytoplasmic face of the integrin receptors. Over 250 different proteins have been detected in these integrin adhesion complexes (IACs) and new components are continually being discovered. Implicit in the formation of these large IACs are many interactions that together enable adhesions to serve as dynamic and diverse mechanosignaling centers. The goal of our laboratory is to understand these interactions and disrupt each interaction in turn in order to define approaches that enable us to unpick this complexity with exquisite detail. Biochemical assays are analytical, in vitro procedures that are used to characterize biological molecules and their interactions.

Rejina B. Khan and Lorena Varela contributed equally. Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_9, © Springer Science+Business Media, LLC, part of Springer Nature 2021

115

116

Rejina B. Khan et al.

There are many biochemical assays that have been developed and each assay has its own advantages and disadvantages. The aim of this chapter is not to cover them all but instead to describe the biochemical pipeline we use in our laboratory for expression, purification and biochemical and biophysical investigation of integrin adhesion complex-associated proteins. The chapter starts with a brief overview of our common methods of protein production, including cloning and mutagenesis to generate the expression plasmids for recombinant protein expression. The generation of highquality protein samples is key to biochemical studies. Circular dichroism (CD) and nuclear magnetic resonance (NMR) are powerful approaches for assessing protein quality and ensuring correct folding. We then introduce our “go-to” suite of biochemical assays for characterization of protein interactions which include CD, NMR, fluorescence polarization (FP), microscale thermophoresis (MST), size-exclusion chromatography multi-angle light scattering (SEC-MALS) and the pulldown/cosedimentation-based approaches including GST-pulldowns, actin cosedimentation and lipid cosedimentation. We use these methods to characterize binding between fulllength proteins and/or specific regions of proteins. One of the ultimate goals is to identify the specific residues involved in mediating interactions so as to design mutations that abolish binding. Our work goes hand-in-hand with in vivo studies; biochemical insight allows mapping of specific sites of interest which can then be tested in cells or animal models to validate the findings. Utilizing mutations which disrupt the interactions for in vivo studies allows for observation of a phenotype, ultimately allowing identification of the role for that specific interaction.

2

Materials

2.1 Molecular Cloning

1. Thermocycler.

2.1.1 Restriction Enzyme Method

3. dNTPs.

2. Template DNA which contains the insert of interest. 4. Oligonucleotides (primers) designed for your insert. 5. Polymerase (e.g., Promega GoTaq® DNA Polymerase) and associated buffer. 6. DNA clean up kit (e.g., Extraction Kit).

QIAGEN

QIAquick Gel

7. Vector DNA. 8. Restriction enzymes (based upon the designed oligonucleotides) and appropriate reaction buffer. 9. 1% agarose gels for gel electrophoresis analysis and gel extraction.

Biochemical Assessment of the Adhesome

117

10. Appropriate ligase for your vector of choice (e.g., Promega T4 DNA Ligase) and associated buffer. 11. E. coli DH5α/DH10β or other appropriate bacterial strain. 12. Plasmid DNA miniprep kit (e.g., QIAGEN QIAprep Spin Miniprep Kit). 13. Appropriate sequencing oligonucleotides. 2.1.2 Gibson Assembly

1. Thermocycler. 2. Template DNA of vector and of insert. 3. dNTPs. 4. Oligonucleotides (primers). 5. Polymerase (e.g., Promega Pfu® DNA Polymerase) and associated buffer. 6. DNA clean up kit (e.g., Extraction Kit).

QIAGEN

QIAquick Gel

7. Gibson Master Mix (e.g., NEB Master Mix). 8. E. coli DH5α/DH10β or other appropriate bacterial strain. 9. Plasmid DNA miniprep kit (e.g., QIAGEN QIAprep Spin Miniprep Kit). 10. Appropriate sequencing oligonucleotides. 2.1.3 Site-Directed Mutagenesis

1. Thermocycler. 2. Template DNA. 3. dNTPs. 4. Oligonucleotides (primers). 5. DNA polymerase (e.g., Promega Pfu® DNA Polymerase) and associated buffer. 6. DpnI endonuclease (e.g., NEB DpnI). 7. DNA clean up kit (e.g., QIAGEN QIAquick Gel Extraction Kit). 8. E. coli DH5α/DH10β or other appropriate bacterial strain. 9. Plasmid DNA miniprep kit (e.g., QIAGEN QIAprep Spin Miniprep Kit). 10. Appropriate sequencing oligonucleotides.

2.2 Protein Expression 2.2.1 Buffers

1. LB media (per liter) 5 g tryptone, 10 g yeast extract, and 5 g sodium chloride in 1 L. distilled/deionized water. The final pH should be 7.0–7.5. Autoclave. 2. 2M9 minimal media (per liter) is comprised of two parts, solution A and B. Solution A: 12.5 g/L Na2HPO4∙2H2O, 7.5 g/L KH2PO4. The final pH should be 7.0–7.5. Autoclave.

118

Rejina B. Khan et al.

Solution B: 4.0 g glucose (D-Glucose-13C6 (>99% 13C) for carbon labeling), 10.0 mL BME Vitamins (e.g., Sigma BioReagent BME vitamins 100
), 2.0 mL MgSO4 (from autoclaved 1 M stock), 0.1 mL CaCl2 (from autoclaved 1 M stock), 1.0 g NH4Cl (15N-labeled NH4Cl for nitrogen labeling) diluted in 10.0 mL water. Use a 0.2 μm filter to sterilize and add to autoclaved solution A. 2.2.2 General Equipment/Reagents

1. Appropriate antibiotic(s): e.g., 100 mg/mL ampicillin and 50 mg/mL kanamycin for 1000
stock solutions. 2. Refrigerated shaking incubator (for 37 incubation).



C and 18



C

3. His-tagged protein cloned into expression vector of choice (e.g., pET-151 TOPO) or GST-tagged protein cloned into expression vector of choice (e.g., pGEX-TEV). 4. E. coli BL21(DE3) or other appropriate bacterial strain. 5. Centrifuge. 6. Buffer in which to resuspend the harvested cell pellet (resuspend the pellet in the buffer of the first purification step, for example, for His-tagged protein resuspend pellet in Nickel buffer A (20 mM Tris pH 8.0, 500 mM NaCl, 20 mM imidazole). 7. Protease inhibitor (e.g., Roche Complete® Protease Inhibitor Tablets). 2.3 Protein Purification

All buffers should be filtered and degassed.

2.3.1 Buffers Protein Purification for His-Tagged Proteins Using Immobilized Nickel-Affinity Chromatography

1. HisTrap HP column, 5 mL (GE Healthcare).

Protein Purification by Batch Method for His-Tagged Proteins

1. Agarose beads charged with nickel (e.g., HisPur™ Ni-NTA Superflow agarose).

2. Nickel buffer A: 20 mM Tris pH 8.0, 500 mM NaCl, 20 mM imidazole. 3. Nickel buffer B: 20 mM Tris pH 8.0, 500 mM NaCl, 500 mM imidazole.

2. Nickel buffer A: 20 mM Tris pH 8.0, 500 mM NaCl, 20 mM imidazole. 3. Wash buffer: 20 mM Tris pH 8.0, 500 mM NaCl, 50 mM imidazole. 4. Nickel buffer B: 20 mM Tris pH 8.0, 500 mM NaCl, 500 mM imidazole.

Biochemical Assessment of the Adhesome

119

Protein Purification by Batch Method for GST-Tagged Proteins

1. Glutathione beads (e.g., Pierce™ Glutathione Superflow agarose).

Ion-Exchange Chromatography

1. Anion exchange (Q HP) column, 5 mL (GE Healthcare).

2. PBS: 137 mM NaCl, 27 mM KCl, 100 mM Na2HPO4, 18 mM KH2PO4, pH 7.4.

2. Q buffer A (20 mM Tris pH 8.0, 50 mM NaCl). 3. Q buffer B (20 mM Tris pH 8.0, 1 M NaCl). 4. Cation exchange (SP HP) column, 5 mL (GE Healthcare). 5. S buffer A (20 mM phosphate pH 6.5, 50 mM NaCl). 6. S buffer B (20 mM phosphate pH 6.5, 1 M NaCl).

2.3.2 General Equipment/Apparatus

1. Sonicator, French press, or cell disruptor. 2. Centrifuge (for low speed 2831
g (4000 rpm in a Beckman JA-10 rotor), and 48,384
g (20,000 rpm in a Beckman JA-25.5 rotor) at 4  C). 3. Dialysis tubing (e.g., SnakeSkin™, ThermoFisher) with appropriate molecular weight cut off (usually 3, or 10 kDa MWCO) or PD-10 desalting column (GE Healthcare). 4. SDS-PAGE gels at a percentage appropriate to visualize proteins of interest. ¨ KTA™ purification system (GE Healthcare). 5. A 6. Spectrophotometer (e.g., Implen NanoPhotometer).

2.4

Peptides

1. Peptide resuspended in water or PBS to an appropriate concentration. 2. Thiol-reactive fluorescent dye (e.g., fluorescein-5-maleimide or BODIPY™ TMR C5-Maleimide, both from ThermoFisher), solubilized in DMSO according to manufacturer’s instructions. 3. TCEP (tris(2-carboxyethyl)phosphine). 4. Triton X-100. 5. PD-10 desalting column (GE Healthcare).

2.5 Biochemical Assays 2.5.1 Circular Dichroism (CD)

1. JASCO J-715 spectropolarimeter with JASCO power supply PS-150J equipped with JASCO Peltier thermoelectric cooler PTC-423S/15. 2. Quartz cuvette. 3. Protein (0.2–0.5 mg/mL).

2.5.2 Nuclear Magnetic Resonance (NMR)

1. NMR spectrometer (e.g., a 600 MHz five channel Bruker Avance III spectrometer with a CryoProbe).

120

Rejina B. Khan et al.

2. Unlabeled or isotopically labeled protein (150 μM concentration is ideal for HSQC titrations). 3. Deuterated water. 4. NMR tubes: 3 or 5 mm tubes (Wilmad) or 5 mm susceptibility matched tube and plunger (Shigemi). 5. NMR software CcpNmr Analysis for spectra analysis. 2.5.3 Fluorescence Polarization (FP)

1. Black 96-well plate (e.g., Nunc™ F96 MicroWell™ Polystyrene Plate from ThermoFisher). 2. Microplate reader with appropriate filters for fluorophore (e.g., BMG CLARIOstar equipped with polarizing filters). 3. Fluorescently labeled peptide (see “coupling peptides”). 4. Protein at appropriate concentration (determined by affinity of the interaction). 5. PBS (100 mM Na2HPO4, 18 mM KH2PO4, pH 7.4, 137 mM NaCl, 27 mM KCl). 6. Curve fitting software (e.g., GraphPad Prism).

2.5.4 Microscale Thermophoresis (MST)

1. Microscale thermophoresis instrument (e.g., NanoTemper Monolith NT.115). 2. Monolith His-tag labeling kit (e.g., RED-Tris-NTA—this kit includes His-tag dye, buffers, capillaries and tubes). 3. His-tagged protein (target). 4. Ligand to study interaction with (typical ligands include protein without His-tag, peptide, nucleotide, macromolecule, lipid vesicle, or metal ion). 5. Benchtop centrifuge (up to 16,200
g), refrigerated. 6. MO.Control software for analysis.

2.5.5 Size-Exclusion ChromatographyMultiangle Light Scattering (SEC-MALS)

¨ KTA™ purification system (GE Healthcare). 1. A 2. Appropriate SEC columns (e.g., Superdex 200, Sephacryl 200, or Sephacryl 300). 3. MALS system (e.g., Malvern MALS system Viscotek SEC-MALS 9 with Modular RI detector Viscotek VE 3580). 4. OmniSEC software for analysis.

2.5.6 GST-Pulldowns

1. GST-tagged protein bound to Glutathione beads (e.g., Pierce™ Glutathione Superflow agarose). 2. SDS-PAGE gels at a percentage appropriate to visualize your proteins (e.g., NuPAGE™ 4–12% Bis-Tris Protein Gels). 3. ImageJ software for analysis.

Biochemical Assessment of the Adhesome 2.5.7 Actin Cosedimentation Assay

121

1. Polymerized F-actin at a stock concentration above the critical concentration (we use ~100 μM). 2. Actin co-sed buffer (10 mM Tris pH 7.0, 2 mM MgCl2, 0.2 mM ATP, 1 mM DTT, 50 mM NaCl). 3. Ultracentrifuge (up to 100,000
g, for example a Beckman TLA-100 rotor) for high-speed actin-binding assay. 4. Benchtop centrifuge (up to 16,200
g) for low-speed actinbundling assay. 5. SDS-PAGE gels at a percentage appropriate to visualize your proteins (e.g., NuPAGE™ 4–12% Bis-Tris Protein Gels). 6. ImageJ software for analysis.

2.5.8 Lipid Cosedimentation Assay

1. Chloroform for dissolving lipids to make lipid films. 2. Nitrogen cylinder. 3. Lipids: phosphatidylcholine (PC), phosphatidylserine (PS) from Sigma-Aldrich and Phosphatidylinositol 4,5-bisphosphate (PIP2) from Avanti Polar Lipids (Alabaster, AL). 4. Protein buffer (20 mM Tris pH 7.4, 0.1 mM EDTA, 15 mM β-mercaptoethanol). 5. Lipid co-sed buffer (20 mM HEPES pH 7.4, 0.2 mM EGTA). 6. Benchtop centrifuge (up to 16,200
g). 7. SDS-PAGE gels at a percentage appropriate to visualize your proteins (e.g., NuPAGE™ 4–12% Bis-Tris Protein Gels). 8. ImageJ software for analysis.

3

Methods

3.1 Molecular Cloning

For biochemical and structural analysis, we predominantly use two expression vectors: (1) pET-151, which contains a 6
His-tag followed by a TEV cleavage site, and the base vector contains a multiple cloning site with BamHI, NotI, EcoRI, and XhoI restriction sites; (2) pGEX-TEV (a modified version of pGEX-4T), which contains a glutathione S-transferase (GST)-tag followed by a TEV cleavage site, and the base vector contains a broader multiple cloning site with BamHI, NheI, EcoRI, SalI, XhoI, and NotI restriction sites. Both of these vectors encode ampicillin resistance. The cloning methods described here are applicable to any vector but are written in this chapter as used for pET-151 and pGEX-TEV specifically.

3.1.1 Restriction Enzyme Digest and Ligation

Restriction enzyme digest followed by ligation of the generated sticky ends is a classic approach to molecular cloning. We use this

122

Rejina B. Khan et al.

method primarily for subcloning regions of interest, often from mammalian vectors or genomic DNA, into our expression vectors listed above. Here, we describe our general approach for this process and the reagents we use. 1. Design forward and reverse oligonucleotides using the sequence of the insert. Each oligonucleotide should: consist of ~15–20 bases from your insert, introduce required restriction sites, contain ~60–70% guanine/cytosine content, and terminate with a guanine or cytosine base at the 50 end. 2. Perform PCR according to manufacturer’s instructions for your polymerase of choice and purify the product. To purify, use ethanol precipitation method or commercial kits such as QIAGEN QIAquick Gel Extraction Kit. 3. Carry out a restriction digest, usually at 37  C for 2–4 h (check manufacturer’s instructions for your restriction enzymes of choice), for your purified PCR product and for your vector of choice. Each reaction should be performed with identical/ compatible restriction enzymes to generate compatible sticky ends. Use agarose gel extraction to purify the digested products. 4. Set up a ligation reaction. For pET-151 and pGEX-TEV vectors, T4 DNA ligase is a suitable enzyme. Follow manufacturer’s instructions for your enzyme of choice. 5. Transform the ligation reaction into DH5α/DH10β E. coli. Perform minipreps for individual colonies and perform a test digest, if suitable. 6. Sequence the DNA to completion using appropriate sequencing oligonucleotides to ensure that the insert is present. 3.1.2 Gibson Assembly

Gibson assembly methodology is a flexible cloning strategy developed by Daniel G. Gibson, of the J. Craig Venter Institute [1]. This is a robust exonuclease-based method to assemble DNA seamlessly and in the correct order. The reaction is carried out under isothermal conditions using three enzymatic activities: a 50 exonuclease generates long overhangs, a polymerase fills in the gaps of the annealed single strand regions, and a DNA ligase seals the nicks of the annealed strands and fills in the gaps. If deletions or insertions in the same plasmid are required, it is only necessary to design two primers (forward and reverse) that introduce the desired variation in the original template DNA. If a change of plasmid is required, two sets of forward and reverse primers are required, one set for the plasmid and another set for the insert DNA fragments, to introduce the desired variations. A high-fidelity DNA polymerase and a thermocycler are required to perform separate PCR reactions to amplify the different noncircularized fragments with overlapping ends.

Biochemical Assessment of the Adhesome

123

Table 1 PCR program used for Gibson assembly and for site-directed mutagenesis Process

Time

Temperature ( C)

Initial DNA denaturation

30 s

95

DNA denaturation Primers annealing DNA chain extension

30 s 95 30 s T ¼ Tm  5 Depends on polymerase used Depends on polymerase used and plasmid or fragment size

Cycles 1 30

DNA elongation completion 5 min

Depends on polymerase used

1

Storage

4

1

End

1. Design the required primers. First, define the desired final assembled product and then select primer sequences for both sides of the joints with sufficient binding to both the templates and the overlapping region of the other fragment to allow the Gibson assembly reaction to procced. Starting with either fragment, select a region of sequence that starts at the joint with a Tm of around 60  C, making sure to include a G/C anchor at the 50 end. This is the binding region of the primer. Next, add the overlapping region by selecting some bp of the other fragment. In general, an overlap of 40 bp yields a sufficient Tm for the Gibson reaction, so if the primer is extended by 20 bp that will give 40 bp of overlap. This process needs to be repeated for each joint. 2. Perform PCR reactions to amplify each fragment following manufacturer’s instructions for your polymerase of choice. The PCR program we normally use is described in Table 1. 3. Purify the PCR products using ethanol precipitation method or commercial kits (we use QIAGEN QIAquick Gel Extraction Kit). 4. Perform the Gibson assembly reaction by adding the Gibson Master Mix (we use the NEB Gibson Assembly Master Mix) to the recommended concentrations of vector and insert as described in the manufacturer’s instructions. Perform the reaction in a thermocycler at 50  C for 15 min. 5. Transform the now circularized vector into high insert stability competent cells (we use DH10β E. coli). 6. Perform minipreps for individual colonies and sequence the DNA to completion using appropriate sequencing oligonucleotides to confirm the correct plasmid assembly. Once checked, insert the plasmid into BL21(DE3) E. coli cells to overexpress the proteins.

124

Rejina B. Khan et al.

3.1.3 Site-Directed Mutagenesis

Site-directed mutagenesis (SDM) is used to make changes to DNA to modify the protein of interest. Such modifications we introduce include targeted mutation of key residues to disrupt binding, disease-causing mutations, small insertions or small deletions, introduction of stop codons to make shorter proteins, etc. SDM uses PCR with overlapping forward and reverse primers designed such that they introduce the desired variations into the sequence. A high-fidelity DNA polymerase and a thermocycler are required to amplify the mutated DNA plasmid. 1. Design forward and reverse oligonucleotides using the sequence of your insert. 2. The primers should be approximately 30 bp in length with the mutated site as close to the center as possible, with a minimum of 12 bp either side. 3. Perform PCR according to manufacturer’s instructions for your polymerase of choice and purify the product. The PCR program used is described in Table 1. 4. Following temperature cycling, treat the product with DpnI. The DpnI endonuclease is used to digest the parental DNA template and to select for mutation-containing synthesized DNA. Purify the PCR products using ethanol precipitation method or commercial kits (we use QIAGEN QIAquick Gel Extraction Kit). 5. Transform the vector containing the desired mutations into high insert stability competent cells (we use DH10β E. coli). 6. Perform minipreps for individual colonies and sequence the DNA to completion using appropriate sequencing oligonucleotides to confirm the presence of the mutation. Once checked, transform the plasmids into BL21 (DE3) E. coli cells to overexpress the proteins.

3.2 Protein Expression

Here, we describe the most common expression conditions for the talin constructs that we work with. These conditions give good expression for many proteins and are a good starting point. However, it is desirable to optimize the conditions for each protein to attain maximum yields. Expression trials at different temperatures, different IPTG concentrations and different induction times will enable the identification of optimal conditions for expression. For proteins that cannot be expressed in bacteria, or require chaperones, cofactors, or posttranslation modification, it can be necessary to use insect or mammalian cell expression systems, but these are not discussed here.

3.2.1 Unlabeled Protein Expression

Unlabeled proteins are expressed recombinantly in bacterial cultures using standard LB media. The required antibiotic will be dependent on the plasmid and strain used. Our usual strategy

Biochemical Assessment of the Adhesome

125

uses the pET-151 vectors transformed into BL21 (DE3)* E. coli which requires only ampicillin. 1. Transform BL21(DE3)* E. coli with the plasmid encoding the protein of interest and streak onto LB agar plate containing the appropriate antibiotic(s). 2. Inoculate 10 mL of LB media containing the appropriate antibiotics with BL21 (DE3)* E. coli culture containing the desired protein construct. Grow this starter culture overnight at 37  C in a rotary shaker. 3. Use 5 mL of the overnight culture to inoculate 1 L of LB with the appropriate antibiotic and continue to grow the culture at 37  C to an approximate OD600 of 0.6–0.8. At this point, induce the cell culture with 1 mM IPTG for 3 h at 37  C or 200 μM IPTG for overnight expression at 18  C. 4. Harvest the bacteria by centrifugation at 2831
g (4000 rpm in a Beckman JA-10 rotor) at 4  C for 10 min and resuspend them in 30 mL of Nickel buffer A on ice. Add protease inhibitor tablet. At this point, the resuspended pellet can be frozen for further purification in the future or it can be immediately further purified. 3.2.2 Isotopically Labeled Protein Expression

For multidimensional NMR experiments, proteins need to be labeled with 15N and/or 13C. For isotopic labeling, 2M9 minimal media needs to be used. 1. Inoculate 10 mL of 2M9 minimal media containing the appropriate antibiotic(s) with BL21(DE3)* E. coli culture containing the desired protein construct. Grow this starter culture overnight at 37  C in a rotary shaker. 2. Use 5 mL of the overnight culture to inoculate 1 L of 2M9 minimal media with appropriate antibiotic and continue to grow the culture at 37  C to an OD600 of 0.6–0.8. At this point, induce the cell cultures with IPTG (200 μM final concentration) for 15–20 h at 18  C. 3. Harvest the bacteria by centrifugation, as previously described, and resuspend in 30 mL of Nickel buffer A on ice. Add protease inhibitor tablet. At this point, the resuspended pellet can be frozen for further purification in the future or it can be immediately further purified.

3.2.3 Condensation Method of Isotopically Labeled Protein Expression

In order to boost the rate of cell growth and simultaneously minimize the required cost (by reducing the amount of isotopic labeling), the 4:1 condensation method can be used [2]. Here, the cells are grown to high density in LB media before being transferred into the labeled media for expression.

126

Rejina B. Khan et al.

1. Inoculate LB media containing the appropriate antibiotic (four times the volume of the 2M9 minimal media that will be used later) with 5 mL of starter culture per L. 2. Grow cells at 37  C to an OD600 of 0.6–0.8. 3. Next, wash the cells to remove the rest of LB media by centrifugation (3000
g, 5 min at 4  C) and resuspend them in 200 mL of 2M9 solution A and repeating the centrifugation. 4. Resuspend the cells in 2M9 minimal media final desired volume (solution A + solution B) with appropriate antibiotic(s) and incubate for 1 h at 18  C to allow them to adapt to the new growth conditions. At this point, induce protein expression with 200 μM IPTG at 18  C for 15–20 h. 5. Harvest the bacteria by centrifugation, as previously described, and resuspend in 30 mL of Nickel buffer A on ice. Add protease inhibitor tablet. At this point, the resuspended pellet can be frozen for further purification in the future or it can be immediately further purified. 3.3 Protein Purification

There are many approaches and variations of methodology for purifying recombinant proteins. Here we describe our common strategies which are optimized for speed and purified protein quality. Collect samples for SDS-PAGE analysis at each step of the purification for monitoring the purification process. 1. If resuspended bacterial pellet is frozen, thaw it at room temperature. Next, lyse the cells by sonication for 5 cycles of 30 s on followed by 30 s off at ~35% amplitude. Ensure to keep the sample on ice throughout the sonicating process and thereafter. 2. Centrifuge the lysed cells at ~48,000
g at 4  C for 30 min. Discard the pelleted cell membranes. Retain the supernatant and filter using a 0.45 μm filter to remove any remaining cell debris. Prepare a sample of the supernatant and pellet for SDS-PAGE analysis—this will allow validation that the protein of interest is soluble.

3.3.1 Protein Purification for His-Tagged Proteins Using Immobilized Nickel-Affinity Chromatography

1. Equilibrate a 5 mL HisTrap HP (see Note 1) column using Nickel buffer A and Nickel buffer B. At a flowrate of ~4 mL/ min, wash through with buffer: 3 column volumes (cv) of 100% A followed by 3 cv of 100% B and finally 6 cv of 100% A. 2. Ensure that the peristaltic sample loading pump has been equilibrated in Nickel buffer A. Load the filtered supernatant onto the column at a flowrate of ~2.5 mL/min. 3. Following sample application, use the sample pump to wash through ~20 mL of Nickel buffer A. This ensures that the entirety of the sample is loaded onto the column and reequilibrates the sample loading pump for the next run.

Biochemical Assessment of the Adhesome

127

4. Wash out unbound protein by washing the column with 5–8 cv of Nickel buffer A at 5 mL/min. 5. Use gradient elution at 5 mL/min—this strategy increases the amount of Nickel buffer B (high imidazole buffer) in a linear gradient. Set the starting %B concentration to 0% and the target %B concentration to 100% over 50 mL. Collect 1–5 mL fractions. 6. Finish by reequilibrating the column and the system using 5 cv of Nickel buffer A at 5 mL/min. The next sample can be loaded at this point. 7. Once complete, store the column and the system in 20% ethanol by flowing at 1 mL/min for 2 cv. 8. Prepare samples for SDS-PAGE analysis for fractions corresponding to a peak to confirm presence of protein of interest and to assess purity. Pool suitable fractions. 3.3.2 Protein Purification for His-Tagged Proteins by Batch Method

Purification of His-tagged proteins can also be performed by the batch method, a method useful for purification of proteins which are more sensitive or susceptible to aggregation/degradation. The batch method involves the use of centrifugation to separate the supernatant from the nickel-NTA beads to allow purification. This method is an effective alternative if purifying multiple proteins in ¨ KTA™ purification parallel and/or there is limited access to an A system. Collect samples for SDS-PAGE analysis at each step of the purification for monitoring the purification process. 1. Use ~1 mL Ni-NTA agarose slurry (e.g., Qiagen HisPur™ Ni-NTA Superflow) per liter of bacterial culture and wash the beads by adding 2
30 mL Nickel buffer A and mixing gently. Centrifuge at 700
g (2510 rpm in a Hettich Rotanta 460 R centrifuge) for 2 min at 4  C. Discard supernatant. 2. Add the filtered cell lysate to the washed beads and agitate at room temperature for 30 min. For some proteins, particularly those which are more unstable, it may be preferable to perform this step at 4  C. 3. Centrifuge at 700
g (2510 rpm in a Hettich Rotanta 460 R centrifuge) for 2 min at 4  C. Remove supernatant. 4. Wash the beads with 40 mL Nickel buffer A and mix gently. Centrifuge at 700
g (2510 rpm in a Hettich Rotanta 460 R centrifuge) for 2 min at 4  C. Remove supernatant. Repeat this wash step 5–6 times total. 5. Perform the final wash step using 40 mL of Wash buffer. 6. After the final wash step, resuspend the beads in 5 mL of buffer and pour into an empty gravity flow column (Bio-Rad).

128

Rejina B. Khan et al.

7. Elute the His-tagged protein with Nickel buffer B in 1 mL fractions. Some proteins precipitate in high imidazole, so it may be preferable to dilute the eluted protein fractions with Nickel buffer A immediately following elution. 8. Prepare a sample of each fraction for SDS-PAGE analysis to confirm presence of protein of interest and to assess purity. Pool suitable fractions. Following either method for purification of His-tagged proteins, the pooled protein sample should be 90–95% pure. If the His-tag is to be cleaved, add TEV protease at this stage. Exchange the pooled fractions (including TEV protease, if applicable) into an appropriate buffer for cleavage. If ion exchange chromatography will be performed, use dialysis to exchange the sample into S buffer A or Q buffer A, depending on whether cation or anion exchange chromatography are to be performed. 3.3.3 Protein Purification by Batch Method for GST-Tagged Proteins

1. Take ~1 mL glutathione agarose slurry (e.g., Pierce™ Glutathione Superflow) and wash the beads by adding 20 mL PBS and mixing gently. Centrifuge at 700
g for 3 min at 4  C. Remove supernatant. 2. Add the filtered cell lysate to the washed beads and agitate at room temperature for 1–2 h. For some proteins, particularly those which are more unstable, it may be more suitable to perform this step at 4  C. 3. Centrifuge at 700
g for 3 min at 4  C. Remove supernatant. 4. Wash the beads with 30 mL PBS and mix gently. Centrifuge at 700
g for 3 min at 4  C. Remove supernatant. Repeat this wash step 5–6 times. 5. After final wash step, remove the supernatant and resuspend the beads in a suitable buffer. Prepare a sample of this beadbound protein slurry for SDS-PAGE analysis to ensure that the protein of interest is present. The protein is now ready to be used for GST-pulldown experiments, if applicable. 6. If the protein is to be cleaved from its GST-tag and thus the beads, add TEV protease and agitate overnight at room temperature or 4  C as required. 7. Transfer to a column and collect the flow through. This fraction will contain the protein of interest, now cleaved from the GST-tag. Confirm this by preparing a sample for SDS-PAGE analysis.

3.3.4 Ion-Exchange Chromatography

We use ion-exchange chromatography to remove TEV protease and cleaved His-tags from our purified protein. Even if TEV cleavage is not performed, this is a useful step to increase the purity of the final protein sample generated. Use the isoelectric point of your

Biochemical Assessment of the Adhesome

129

protein of interest to deduce whether a Q (anion exchange) or S (cation exchange) column is required. With the buffers listed, an isoelectric point of ~7 and below is generally suitable for a Q column; an isoelectric point above ~8 is suitable for a S column. For any proteins with an isoelectric point within/close to this range, the pH of the buffers could be altered (ensuring to remain within the buffering capacity) or the HisTrap HP column could be used again instead. If the latter is used, the cleaved protein of interest will be in the column flow through, separated from the TEV protease and cleaved His-tag which will both instead bind to the column. 1. Wash either a 5 mL HiTrap Q HP or a 5 mL HiTrap SP HP column (both from GE Healthcare, see Note 2) using the appropriate buffers, that is (1) Q or S buffer A and (2) Q or S buffer B. At a flowrate of ~4 mL/min, wash through with buffer: 3 cv of 100% A followed by 3 cv of 100% B and finally 6 cv of 100% A. This allows removal of any contaminants and then prepares the column for use. 2. Ensure that the sample pump has been equilibrated in Q/S buffer A. At a flowrate of ~2.5 mL/min, load the filtered sample onto the column. 3. Following sample application, use the sample pump to load ~20 mL of Q/S buffer A. This ensures that the entirety of the sample loaded has reached the column. 4. Wash out unbound protein by washing column with 5–8 column volumes of Q/S buffer A at 5 mL/min. 5. Use gradient elution at 5 mL/min. Set start %B concentration at 0 and target %B concentration to 100. Fractionate using settings of your choice (Fig. 1a). Stepwise elution can also be used. 6. Finish by reequilibrating the column using 4 column volumes of Q/S buffer A at 5 mL/min. The next sample can be loaded at this point and the process repeated. 7. Once complete, store the column and the system in 20% ethanol by flowing at 1 mL/min for 2 cv. 8. Prepare samples for SDS-PAGE analysis of fractions corresponding to a peak (Fig. 1b). Once fractions containing protein of interest and minimal contaminants have been identified, pool these together. 9. Use a PD-10 desalting column or dialysis to exchange the purified protein into a suitable buffer for the experiments to be performed. Measure the concentration using the absorbance at 280 nm (we use Implen NanoPhotometer to measure this) and the molar extinction coefficient (calculated using

130

Rejina B. Khan et al.

Fig. 1 (a) Example elution profile using anion-exchange chromatography (Q column) of talin1 R7-R8 G1404L mutant. The peak in A280 absorbance (blue) corresponds to the elution of the protein of interest at a given concentration of elution buffer (gray). (b) SDS-PAGE analysis of elution fractions corresponding to the peak in A280 absorbance

ProtParam https://web.expasy.org/protparam/) and concentrate or dilute as required. 10. Aliquot and flash freeze the final purified protein using liquid nitrogen. Note: the effects of flash freezing may vary between proteins—while most proteins we have worked with function identically before freezing and after thawing, some proteins are susceptible to precipitation, degradation or differences in activity.

Biochemical Assessment of the Adhesome

3.4

Peptides

3.4.1 Synthetic Peptide Design

131

There are a number of factors to consider when designing synthetic peptides to use for biochemical and biophysical experiments. These will vary depending for each experiment, but we typically design peptides as summarized briefly here. 1. Identify the region of the protein to synthesize as a peptide. We use a combination of deletion mapping to identify the region of the protein involved in the interaction, secondary structure prediction (using PSIPRED [3]), amino acid conservation (regions of functional importance are highly conserved across species), and motif identification (i.e., if it is a vinculin binding site, or an LD-motif then both of these have well defined consensus sequences). 2. If the identified region is part of a secondary structure element such as a helix then it is important that the peptide spans the entire helix and has a couple of nonhelical residues either side to enable the secondary structure to form. 3. The solubility of each peptide is dependent on the specific sequence and the net charge of the peptide. These factors can be calculated using various online servers for predicting peptide properties. If poor solubility is predicted then these can be dissolved in DMSO, but it can also be useful to adjust the sequence to increase the percentage of charged residues. 4. If fluorescence polarization experiments will be carried out (or any experiments where coupling to a thiol-reactive moiety will be required), the peptide design should include a terminal cysteine residue. The choice of N- or C-terminus can depend on whether the putative interacting region is located close to the termini. If so, add the terminal cysteine residue to the opposite end. 5. If the target peptide already contains a cysteine in its sequence then maleimide coupling is not possible. Ordering peptides precoupled to the required dye may be more suitable. Ordering synthetic peptides up to ~35 amino acids is reasonably cost effective. Above 10 mg, the cost tends to get prohibitive and it might be better to consider making them recombinantly. We recommend to always order the peptide in 2 mg aliquots.

3.4.2 Coupling Peptides

For fluorescence polarization experiments (see following section), synthetic peptides must be coupled to a fluorescent dye. While the peptides can be synthesized with these, we generally couple them in-house. We use fluorescein-5-maleimide or BODIPY ™ TMR C5-Maleimide diluted according to manufacturer’s instructions. An important consideration is to choose the thiol-reactive fluorophore so that it is compatible with the equipped polarizing filters in your plate reader.

132

Rejina B. Khan et al.

1. Mix 100 μM peptide (resuspended in water or PBS), 0.1% Triton X-100, 5 mM TCEP, and 25 μL of fluorescent dye. Make up to 1 mL using PBS. 2. Protect from light and leave stirring at room temperature for 2 h. 3. Using a PD-10 desalting column equilibrated with PBS, remove excess dye. Elute using 2.5 mL PBS and collect the entirety of the flow through. The coupled peptide will elute after ~1 mL. 4. Prepare 50 μL aliquots, flash-freeze using liquid nitrogen and store at 20  C until needed. 3.5 Biochemical Assays 3.5.1 Circular Dichroism (CD)

3.5.2 Measurement of Far-UV CD Spectra

CD is an excellent tool for rapid determination of the secondary structure and folding and binding properties of proteins. CD also provides insight early on in the project to the behavior of the protein of interest including its thermal stability. This method is a critical control test for validating point mutations and ensuring they do not have off-target effects such as protein destabilization. CD works by measuring the wavelength dependence of the differential absorption of the left and right-handed circularly polarized light. The left- and right-handed components of a polarized beam of light interact differently with the chiral centers of an optically active chromophore (present in the amides of the polypeptide backbone of proteins). Far-UV CD spectroscopy has high sensitivity to changes in the different secondary structure elements in proteins including α-helices, β-sheets, β-turns and random coils, whereas near-UV CD spectroscopy gives information about changes in the tertiary structure of the protein. Melting curves can also be acquired by measuring the change of CD signal at a fixed wavelength over a temperature gradient. Figure 2 shows the typical melting curve data set obtained using this method. In this experiment (taken from ref. 4), the melting curve of the talin rod domain 3 (R3) was measured for the wild type (WT) and a stabilizing mutant (IVVI). 1. Prepare a protein sample at 20–50 μM. 2. Place the sample in a quartz cuvette using an appropriate path length depending on the concentration of the sample. 3. Collect the far-UV spectra between 260 and 190 nm. We normally average 4–8 scans at 100 nm/min, 0.5 nm step resolution, 1 s of response and 0.5 nm of band width (see Note 3).

3.5.3 Measurements of Melting Curves

1. To monitor the time course of unfolding, prepare a protein sample at 20–50 μM.

Biochemical Assessment of the Adhesome

133

Fig. 2 Example of a dataset collected using the CD assay. The melting curve for wildtype talin R3 domain (black) and a stabilizing IVVI mutant (red) were measured as described in the text by monitoring the change in CD at 222 nm with increasing temperature. The melting temperature (Tm) is the point halfway between the transition of the folded and unfolded state; for WT it is 69.6  C, the IVVI mutant >90  C. (Reproduced from ref. 4 under a Creative Commons CC-BYNC license)

2. Place the sample in the quartz cuvette and ensure a lid is in place to avoid evaporation of the sample. 3. Set the temperature ramp while keeping a constant wavelength. We normally measure the CD signal at 222 nm (maximal signal for an α-helix) as a function of temperature between 25 and 90  C with 20 s step resolution, 4 s of response and 1.0 nm of band width. 4. Far-UV CD spectra can be collected at different temperatures during the melting curve course (see Note 3). 3.5.4 Nuclear Magnetic Resonance (NMR)

NMR spectroscopy is a powerful technique to obtain structural and dynamic information on proteins and protein complexes. NMR provides insight into the behavior of the protein and how homogeneous the sample is. It has many uses and is a central part of our repertoire: we use it for structural studies, small molecule screening, protein–protein interactions, protein–ligand interactions, assessing the effects of truncations and mutations on domain folding, and many other contexts. We have discussed its role in structural determination in our previous review [5]. Our focus here is on the use of NMR as a biochemical assay for studying interactions. During an NMR experiment, the sample is placed in a strong magnetic field and irradiated with pulses of radio frequency electromagnetic radiation which cause the NMR-active nuclei to resonate at characteristic frequencies. Chemical shifts are very sensitive to the chemical environment constituting a very sensitive probe of

134

Rejina B. Khan et al.

Fig. 3 Example of a 2D 15N HSQC NMR experiment showing binding of unlabeled Rap1b to 50 μM talin F0. (a) The chemical shift assignments of the F0 spectra are shown; each peak pertains to a specific residue. On addition of Rap1, the residues in close proximity to the binding site experience an altered chemical environment which results in the peak moving. (b) The peaks that move can be mapped onto the protein structure to define the binding site. (Reproduced from ref. 18 under a license from the Company of Biologists (https://jcs.biologists.org/content/131/24/jcs225144))

the conformation of the protein in solution and its interactions with ligands. Proteins can be unlabeled for 1D NMR, or they need to be singly labeled (15N) or doubly labeled (15N and 13C) in order to make those nuclei active under the magnetic field. Each amide group (NH) of the amino acids (with the exception of proline) of a protein will have a particular chemical shift that will be represented as a single peak with a specific location in an NMR spectrum. These chemical shifts can be easily measured using 2D 1H-15N HSQC (Heteronuclear Single-Quantum Correlation) based experiments [6]. For fast time-resolved measurements, 2D 1 H-15N SOFAST-HMQC (band-Selected Optimized Flip-Angle Short Transient Heteronuclear Multiple-Quantum Correlation) spectra [7] can be measured in a few minutes providing similar information as the standard 2D HSQC. Binding of ligands to proteins result in localized alterations to the chemical environment of the residues in close proximity to the binding site. These chemical shift displacements produced by ligand binding allow the biochemical and structural characterization of protein interactions. Figure 3 shows part of the HSQC spectra of 15N-labeled talin F0 domain in the absence (teal) and presence (green) of the small GTPase Rap1b. To follow biochemical features of the protein (folding, aggregation, binding, etc.) by two-dimensional NMR:

Biochemical Assessment of the Adhesome

135

1. Prepare a sample of 15N-labeled protein at approximately 0.1–1 mM in 500 μL of buffer for standard 5 mm NMR tubes or 350 μL for Shigemi tubes, containing 10% v/v D2O. 2. Place the sample into the NMR tube. Take care to avoid bubbles. 3. Set the probe to the desired temperature (usually 298 K). 4. Once the spectrometer is set up (tuned, locked, shimmed, and pulse lengths determined) in accordance with the Bruker manual, collect 1D proton and 2D 1H-15N HSQC spectra. 5. Process spectra (we use NMRPipe [8] and/or Bruker Topspin). 6. Analyze processed spectra using appropriate NMR software. In our laboratory, we use CcpNmr Analysis [9]. We have described our CCPN pipeline for talin NMR previously [5]. 7. If more specific structural information is required, the assignment of the 1H-15N cross peaks of the HSQC spectrum of the protein needs to be performed. This identifies the residue that each amide peak corresponds to. To easily assign a protein, double- labeling is recommended (15N and 13C) in order to be able to use three-dimensional, triple resonance NMR experiments. A normal set of 3D NMR experiments to assign a protein includes HNCA, HN(CO)CA, CBCANH , CBCA (CO)NH, HNCO, and HN(CA)CO. 8. NMR chemical shift assignments for each assigned protein are deposited into the Biological Magnetic Resonance Data Bank (BMRB) repository [10]. 3.5.5 Fluorescence Polarization (FP)

We use FP experiments to investigate putative interactions between proteins and fluorescently labeled peptides. When a fluorescent molecule is excited by polarized light, it will emit light that is polarized. This assay utilizes the rapid tumbling that occurs when fluorescent peptides are in solution; as the molecule tumbles, it will lead to loss of polarization. If an interaction occurs between the fluorescently labeled peptide and the protein, the fluorescently labeled peptide will tumble more slowly with increasing levels of protein, causing an increase in the polarization of the light in one direction. This change in polarization as a function of protein concentration can be quantified, allowing determination of binding affinity. Figure 4 shows a typical data set obtained using this method. In this experiment, synthetic peptides of the talin binding site of KANK proteins were labeled with the fluorophore BODIPYTMR [11]. This peptide was then added to a serial dilution of purified talin R7-R8 domains as follows.

136

Rejina B. Khan et al.

Fig. 4 Example of a dataset collected with the FP assay. Binding of BODIPYlabeled KANK1 (30–60), KANK2 (31–61), and KANK1-4A peptides to talin1 R7–R8. Dissociation constants  SE (μM) for the interactions calculated with the one-site binding equation are indicated in the legend. ND not determined. (Reproduced from ref. 11 under a Creative Commons CC-BY license)

1. Set up an FP program with appropriate parameters for your filters as described in the plate reader software manual. We generally perform experiments at 25  C. 2. Each experiment uses one lane of a 96-well Nunc plate. Add 50 μL PBS into wells 1–11. 3. Add 50 μL of protein to well 12 (talin1 R7-R8 stock solution was 60 μM). 4. Add 50 μL of protein to well 11 and mix. Take 50 μL of this and add to well 10. Mix and transfer 50 μL of this into well 9. Repeat until well 2 and discard the final 50 μL. 5. Make 750 μL stock of 2 μM fluorescently labeled peptide. Add 50 μL of the peptide stock to wells 1–12 and mix well. Note: the fluorescently labeled peptide is susceptible to photobleaching so it is advised to protect from light. 6. Place the plate into the plate reader and run the appropriate FP program for the filters to be used. 7. Repeat in triplicate. 8. To calculate the binding constant, we use nonlinear curve fitting and a one site total binding model on GraphPad Prism. FP assays can be used for multiple experiments, including competition assays where a labeled peptide is displaced from a complex using a competitor ligand.

Biochemical Assessment of the Adhesome 3.5.6 Microscale Thermophoresis (MST)

137

MST is a flexible assay that allows for the analysis of interactions of proteins with ligands of any size and is well suited for binding analysis of proteins which can only be expressed in small amounts. We use MST to measure protein–protein interactions as, unlike the FP assay, it does not require a difference in size between the two species. This also makes the technique ideal for measuring affinities of protein dimerization. The MST assay measures temperatureinduced changes in directed movement of a fluorescently labeled component against increasing concentrations of a putative ligand. A laser is used to create a temperature gradient (typically a few degrees), which induces the movement of particles due to a phenomenon called thermophoresis. Thermophoresis results in movement of the fluorescent moiety which can be quantified. The fluorescently labeled protein is titrated with an unlabeled ligand which upon binding alters the amount of movement in a concentration-dependent manner. This can be quantified using this assay, allowing determination of binding affinity. Figure 5 shows a typical data set obtained using this method. In this experiment, the dimerization of Talin rod domain containing protein 1 (TLNRD1) was measured using MST [12]. Any fluorescent labeling strategy will work for this technique, but we find that using NTA-dyes that bind with high affinity to the His-tag provides a simple, rapid, and efficient method for studying protein interactions. Sample preparation includes the following steps: 1. Mix 1.5 μL of His-tag dye (diluted according to manufacturer’s instructions) with 200 nM of His-tagged protein. Make up to 150 μL total volume using PBS. Leave incubating at room temperature for 30 min and then centrifuge at 16,200
g

Fig. 5 Example of a dataset collected with the MST assay. Unlabeled TLNRD1 was titrated into a fixed concentration (50 nM) of fluorescently labeled TLNRD1 (black). The fitted curve yielded a dimerization Kd of 80  0.59 nM. A monomeric mutant, F250D (red), resulted in a Kd not determined. (Reproduced with permission from ref. 12)

138

Rejina B. Khan et al.

(13,000 rpm in a standard benchtop centrifuge) for 10 min at 4  C to remove any aggregates. Note: protein concentrations may need adjusting according to affinities being measured. 2. Set up 14
200 μL tubes. Add 10 μL of PBS-Tween to tubes 2–14. 3. Add 10 μL ligand (unlabeled protein) to tube 1. 4. Add 10 μL ligand to tube 2 and mix. Transfer 10 μL of this to tube 3. Mix and take 10 μL of this and add to tube 4. Repeat until well 14 and discard the final 10 μL. Thus, tube 1 contains the highest concentration of ligand and tube 14 contains the lowest. 5. Add 10 μL of protein–dye solution to each tube and mix uniformly. 6. Place a capillary into each tube. This will draw up the solution. Place each capillary into the corresponding space on the stage within the NanoTemper Monolith NT.115. Steps to set up the titration protocol. 1. Using MO.Control software, open a new session using NanoRED settings (a different option may need to be selected according to the dye used). 2. Select “binding affinity.” 3. The “target” is the His-tagged protein which has been coupled to the His-dye. Select the His-tag option and input information for the other parameters, for example stock concentrations for each component. 4. For MST power and LED power, these can be set to autodetect if unsure. As a starting point, we generally use 50 and 40, respectively. 5. Ensure that a temperature has been set and keep this consistent for all experiments. We generally use 25  C. 6. Start cap scan and MST. The cap scan should be consistent for each sample. If any potential issues such as sample inhomogeneity or concentration-dependent aggregation are suspected, the software will highlight this. 7. Generate a triplicate dataset. 3.5.7 Size-Exclusion Chromatography Multiangle Light Scattering (SEC-MALS)

SEC-MALS is a technique that combines size exclusion chromatography (SEC) with multiangle light scattering (MALS) analysis. A MALS detector is a form of static light scattering detector which allows the absolute molecular weight (MW) and potentially the radius of gyration (Rg) of a sample to be measured. In SEC-MALS instruments, a MALS detector is connected to an

Biochemical Assessment of the Adhesome

139

Fig. 6 Example of a dataset collected with the SEC-MALS binding assay. Figure shows the SEC-MALS analysis of talin R13-DD (dimerization domain) which constitutes actin-binding site 3 (ABS3). The WT protein elutes as a dimer whereas the L2509P mutant elutes as a monomer. The molecular weight of each peak obtained from the MALS is given. (Reproduced from ref. 13 under a Creative Commons CC-BY-NC-ND license)

HPLC/FPLC SEC column and a concentration, refractive index or UV detector is also connected. The SEC-MALS detector measures the light scattered by the protein sample at different angles and a Debye plot is created. SEC-MALS can be used to acquired information about your protein size and also about aggregation state and binding by determining the absolute molecular weight, independent of the protein conformation, size and elution position. Figure 6 shows a typical data set obtained using this method. In this experiment, the SEC-MALS elution profiles of talin WT R13-DD (dimerization domain) and the L2509P mutant show the difference in molecular weight between them indicating dimerization in the wild type and not in the mutant [13]. To perform a SEC-MALS experiment: 1. Chose the appropriate SEC column for your proteins’ size and equilibrate with the protein buffer until the MALS signal is stable. Most of the experiments we perform in our laboratory use either a Superdex-75 or Superdex-200 column (GE Healthcare). 2. Load the filtered protein sample, ideally 100 μL of protein at 100–150 μM.

140

Rejina B. Khan et al.

3. Run method appropriate for column size and type. Do not change the flow at any point to keep a stable signal. We normally use a flow rate of 0.75 mL/min. 4. The eluted peaks from the SEC column are measured by the MALS machine (we use a Viscotek Sec-Mals 9 and Viscotek RI detector VE3580 (Malvern Panalytical)) which is placed in line downstream of the column. 5. The elution profile is then analyzed to determine the absolute molecular weight of the protein/proteins of interest (we use the OmniSEC software from Malvern Panalytical). 3.5.8 GST Pulldowns

GST-tagged proteins bound to glutathione beads can be an excellent tool for pulldown experiments. Due to the difficulty in measuring the concentration of a purified GST-tagged protein while it is bound to glutathione beads, it may be useful to prepare samples of this and of the protein you are wishing to perform the pulldown experiment against for SDS-PAGE analysis. An image processing program such as ImageJ [14] can be used to quantify the bands on the gel for each protein, allowing an estimation of the concentration of the glutathione beads-bound protein. 1. Use the purified glutathione bead-bound GST-tagged protein (from step 5 of Subheading 3.3.3). Mix 50 μL of this with the putative ligand at the concentration required, to a final volume of 200 μL. Using a 1:1 ratio of the GST-tagged protein to putative ligand is standard but it may be useful to have a different ratio depending on the particular interaction being investigated. Incubate for 1 h at room temperature. 2. After incubation, apply to a gravity flow column. It is useful to collect the flow through and prepare a sample for SDS-PAGE analysis. 3. Wash the beads with 1 mL of buffer and repeat this for a total of 2–3 washes. Allow the buffer to pass through the column. 4. Resuspend the beads in 500 μL buffer. Transfer to a tube and centrifuge at 700
g (2510 rpm in a Hettich Rotanta 460 R centrifuge) for 3 min at 4  C. Remove the supernatant. 5. Prepare a sample of the supernatant and of the pellet for SDS-PAGE analysis. If an interaction occurs, the bead-bound GST-tagged protein and the putative ligand should both be present in the pellet. 6. Important controls include: (1) Washed glutathione beads without the GST-tagged protein present and (2) glutathione beads with the GST-tag present but the protein removed via cleavage with TEV protease. These eliminate false-positive results generated from the putative ligand interacting with the beads or with the GST-tag.

Biochemical Assessment of the Adhesome

141

7. Optimization of conditions may be required. These could include concentrations of the GST-tagged protein or the putative ligand, incubation temperatures, amount loaded onto SDS-PAGE gels, among others. 8. Generate a triplicate dataset. An image processing program such as ImageJ [14] can be used to quantify the bands on the gel. 3.5.9 Actin Cosedimentation Assay

Many adhesion proteins engage the actin cytoskeleton either directly or indirectly as part of actin-binding complexes. As a result, the biochemical characterization of protein interactions with actin is important for understanding these cytoskeletal linkages and can be achieved using an actin cosedimentation assay. By varying the speed at which the samples are spun it is possible to observe actin binding (high speed actin-binding assay) and actin bundling (low speed actin-bundling assay).

Actin Polymerization

All actin was isolated from rabbit muscle acetone powder (ours was kindly gifted by Prof Mike Geeves) using a previously described protocol [15]. Purified G-actin (monomeric) is stored at 10 mg/ mL in G-buffer (5 mM Tris pH 7.5, 0.2 mM CaCl2, 0.2 mM ATP) at 80  C until required. Actin is then polymerized by diluting to ~50 μM with actin polymerization buffer (50 mM KCl, 2 mM MgCl2, 0.2 mM ATP, 1 mM DTT, 1 mM NaN3, pH 7), which is then stored at 4  C and used for up to 1 month after polymerization.

High-Speed Actin-Binding Assay

The interaction between a protein of interest and actin can be measured using a simple high-speed actin cosedimentation assay. When spun at high speed such as 100,000
g, polymerized actin filaments will form a pellet. If the protein of interest binds to the F-actin, it will go into the pellet with the actin whereas, if no interaction occurs, the protein will remain in the supernatant. 1. Dilute prepolymerized actin to 20 μM in polymerization buffer and incubate with 20 μM of protein for 1 h at room temperature. Protein–actin ratios can be altered according to experiment requirements. Optimal final volume is 100 μL. 2. Spin samples at 100,000
g for 20 min at 4  C using an ultracentrifuge. It helps to mark the outside facing edge of the centrifugation tube to identify pellet position. 3. Transfer 50 μL of supernatant to a clean tube and add 10 μL of 6
gel loading buffer. Discard the remaining supernatant, taking care not to disturb the pellet. 4. Gently wash the pellet with 50 μL polymerization buffer and resuspend in 60 μL 1
gel loading buffer (Co-sed buffer plus 6
gel loading buffer).

142

Rejina B. Khan et al.

Fig. 7 Example of a dataset collected with the high speed actin-binding assay. This experiment shows TLNRD1 binding to actin. Left panel: Control experiment of the high-speed spin of the actin and TLNRD1 protein alone. Actin is seen in the Pellet fraction (P) whereas the TLNRD1 protein remains in the Supernatant fraction (S/N). Right panel: When TLNRD1 is incubated with actin TLNRD1 protein is seen in the Pellet fraction. (Reproduced with permission from ref. 12)

5. Load equal volumes of pellet and supernatant onto SDS-PAGE gels. We analyze the band densities using ImageJ software [14]. Figure 7 shows a typical data set obtained using this method. When spun at high speed, F-actin filaments will pellet. If proteins interact with the actin, they will cosediment and be detected in the pellet fraction. In this experiment, TLNRD1 was incubated with F-actin and shown to be an actin-binding protein, as seen by the TLNRD1 band in the pellet fraction only in the presence of F-actin [13]. Low-Speed Actin-Bundling Assay

Some proteins are able to bind to more than one actin filament simultaneously and so can induce actin-bundling. This actinbundling functionality can be tested using a similar approach but using a low-speed actin cosedimentation assay. At low speed centrifugation (e.g., 16,200
g), F-actin does not pellet and remains in the supernatant. If the protein of interest induces bundling of the actin filaments, the F-actin will form a pellet. By comparing the amount of actin in the pellet vs. supernatant, whether the protein is bundling actin filaments can be determined. Figure 8 illustrates how a low speed centrifugation can identify proteins with actinbundling capabilities. When spun at low speed, single F-actin

Biochemical Assessment of the Adhesome

143

Fig. 8 Example of a dataset collected with the low speed actin-bundling assay. This experiment shows TLNRD1 bundling actin. In the absence of TLNRD1 the actin is mostly in the solution. As the amount of actinbundling protein is increased the amount of bundled actin in the pellet increases. (Reproduced with permission from ref. 12)

filaments do not pellet while bundled F-actin filaments (many filaments crosslinked together) do. In this assay, actin is spun at low speed; by itself (right side of gel) only a small amount of actin is found in the pellet. However, as the actin-bundling protein, TLNRD1 is added the amount of actin in the pellet increases as the filaments are bundled together. 1. Prepare samples using the same approach as the high-speed binding assay. Protein–actin ratios can be altered according to experiment requirements. 2. Spin samples at 16,200
g (13,000 rpm in a standard benchtop centrifuge) for 20 min at 4  C using a benchtop centrifuge. 3. As previously, transfer 50 μL of supernatant to a clean tube and add 10 μL of 6
gel loading buffer. Carefully remove and discard the remaining supernatant. 4. Resuspend the pellet in 60 μL 1
gel loading buffer (Co-sed buffer plus 6
gel loading buffer). 5. Analyze supernatant and pellets on SDS-PAGE using the same approach as the high-speed assay. 3.5.10 Lipid Cosedimentation Assays

Integrin adhesion complexes assemble on the cytoplasmic face of the integrin. As a result, many proteins are in close proximity to, and engage with, the plasma membrane. The biochemical characterization of protein interactions with lipids is important for understanding the role of the membrane in mediating these interactions and can be probed using lipid cosedimentation assays. In these experiments, we prepare phospholipid vesicles comprised of different combinations of lipids. Figure 9 shows a typical data set obtained using this method. When spun at low speed, lipid vesicles

144

Rejina B. Khan et al.

Fig. 9 Example of a dataset collected with the lipid cosedimentation assay. This experiment shows the talin F2–F3 domain (0.15 mg/mL) binding to lipid vesicles. Either WT or a non–membrane binding mutant (4E) F2–F3 is mixed with vesicles (0.5 mg/mL) consisting of phosphatidylcholine (PC), phosphatidylserine (PS), or a 4:1 ratio of PC:PS. The Supernatant (S) and Pellet (P) lanes are shown. The vesicles are found in the Pellet (P) fractions and proteins that bind the vesicles are also found in the pellet. (Reproduced from ref. 16 under license from John Wiley and Sons (https://www.embopress.org/doi/full/10.1038/emboj.2009.287))

will pellet. If proteins interact with the lipids, they will cosediment and be detected in the pellet fraction. In this experiment, the talin head domains F2–F3 were incubated with different lipid vesicles [16]. The talin head cosediments with vesicles with negatively charged head groups as seen by the F2–F3 band in the pellet fraction only in the presence phosphatidyl serine. Preparation of Large Multilamellar Vesicles

We prepare large multilamellar vesicles essentially as described previously [17]. 1. Dissolve lipids in chloroform and evaporate under nitrogen to produce films of dried phospholipids. 2. Swell films of dried phospholipids at 5 mg/mL in 20 mM HEPES pH 7.4, 0.2 mM EGTA at 42  C for 3 h. 3. Centrifuge at 16,200
g (13,000 rpm in a standard benchtop centrifuge) for 20 min at 4  C. 4. Resuspend pellet at 5 mg/mL in 20 mM HEPES pH 7.4, 0.2 mM EGTA. 5. For large multilamellar vesicles (LMV) the resuspended lipids can be used as is. 6. To generate small unilamellar vesicles (SUV), the resuspended lipids can be passed 11
through a lipid extruder (we use Avanti Mini-Extruder). The extrusion process makes the vesicles smaller and more uniform in size.

Biochemical Assessment of the Adhesome Proteins

145

1. Dilute protein into 20 mM Tris pH 7.4, 0.1 mM EDTA, 15 mM β-mercaptoethanol at 1 mg/mL. 2. Centrifuge at 16,200
g (13,000 rpm in a standard benchtop centrifuge) for 20 min at 4  C. 3. Transfer and keep supernatant.

Interaction Experiment

1. Incubate protein sample at 0.15 mg/mL for 30 min at 25  C in the absence or presence of phospholipid vesicles (0.5 mg/mL) in 200 μL total volume. 2. Make up the following mixture: 20 μL lipid (final conc. 0.5 mg/mL). 30 μL protein (final conc. 0.15 mg/mL). 150 μL 20 mM Tris pH 7.4, 0.1 mM EDTA, 15 mM β-mercaptoethanol. 3. Incubate at 25  C for 30 min. 4. Centrifuge at 16,200
g for 20 min at 4  C. 5. Remove supernatant and add 40 μL of 5
sample buffer. 6. Resuspend pellet in 120 μL of 1
sample buffer. 7. For optimal visualization using SDS-PAGE analysis, we load 14 μL for supernatant fractions and 7 μL for pellet fractions. 8. Stain with Coomassie blue. 9. The percentage of protein bound (protein in pellet/total protein) can now be calculated by measuring band density in ImageJ [14].

3.5.11

4

Structural Studies

A key part of our pipeline is to elucidate the structure of the proteins and protein complexes. Investigating proteins and their interactions at the atomic level provides insight into their biological function and enables the elucidation of the molecular determinants of their interactions. This information allows the design of targeted mutations to disrupt the interactions which, following validation of the mutations’ efficacy using the biochemical pipeline outlined here, allows the interaction to be studied in fine detail in cell biology experiments. We use NMR, small-angle X-ray scattering (SAXS), X-ray crystallography, and cryo-electron microscopy (cryoEM) to explore the structural aspects of integrin adhesion complexes. A detailed discussion of these structural biology approaches is beyond the scope of this chapter and will be the subject of a subsequent chapter.

Notes 1. HisTrap HP columns (5 mL) have a pressure limit of 0.3 mPa. For each step of this process, ensure that a high-pressure alarm

146

Rejina B. Khan et al.

is set to warn if the pressure is reaching/exceeding this limit. The maximum flow rate for this column is 5 mL/min. 2. Q HP and SP HP columns (5 mL) have a pressure limit of 0.5 mPa. For each step of this process, ensure that a highpressure alarm is set to warn if the pressure is reaching/exceeding this limit. The maximum flow rate for this column is 5 mL/ min. 3. A nitrogen flow of 9 L/min must be used to purge and refrigerate the system to protect the optics.

Acknowledgments We thank members of the Goult laboratory past and present for help in the development of these assays. We also thank Alex Gingras and Neil Bate for helping develop the protocols for protein expression and purification. The Goult lab is funded by BBSRC grants (BB/N007336/1 and BB/S007245/1), and HFSP grant (RGP00001/2016). R.K. is funded by a University of Kent studentship. References 1. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA 3rd, Smith HO (2009) Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 6 (5):343–345. https://doi.org/10.1038/ nmeth.1318 2. Marley J, Lu M, Bracken C (2001) A method for efficient isotopic labeling of recombinant proteins. J Biomol NMR 20(1):71–75. https://doi.org/10.1023/a:1011254402785 3. McGuffin LJ, Bryson K, Jones DT (2000) The PSIPRED protein structure prediction server. Bioinformatics 16(4):404–405. https://doi. org/10.1093/bioinformatics/16.4.404 4. Goult BT, Zacharchenko T, Bate N, Tsang R, Hey F, Gingras AR, Elliott PR, Roberts GC, Ballestrem C, Critchley DR, Barsukov IL (2013) RIAM and vinculin binding to talin are mutually exclusive and regulate adhesion assembly and turnover. J Biol Chem 288 (12):8238–8249. https://doi.org/10.1074/ jbc.M112.438119 5. Skinner SP, Goult BT, Fogh RH, Boucher W, Stevens TJ, Laue ED, Vuister GW (2015) Structure calculation, refinement and validation using CcpNmr analysis. Acta Crystallogr D Biol Crystallogr 71(Pt 1):154–161. https:// doi.org/10.1107/S1399004714026662

6. Bax A, Morris GA (1981) An improved method for heteronuclear chemical shift correlation by two-dimensional NMR. J Magn Reson (1969) 42(3):501–505. https://doi. org/10.1016/0022-2364(81)90272-9 7. Schanda P, Kupce E, Brutscher B (2005) SOF AST -HMQC experiments for recording two-dimensional heteronuclear correlation spectra of proteins within a few seconds. J Biomol NMR 33(4):199–211. https://doi.org/ 10.1007/s10858-005-4425-x 8. Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6(3):277–293. https://doi.org/10.1007/bf00197809 9. Vranken WF, Boucher W, Stevens TJ, Fogh RH, Pajon A, Llinas M, Ulrich EL, Markley JL, Ionides J, Laue ED (2005) The CCPN data model for NMR spectroscopy: development of a software pipeline. Proteins 59 (4):687–696. https://doi.org/10.1002/prot. 20449 10. Ulrich EL, Akutsu H, Doreleijers JF, Harano Y, Ioannidis YE, Lin J, Livny M, Mading S, Maziuk D, Miller Z, Nakatani E, Schulte CF, Tolmie DE, Kent Wenger R, Yao H, Markley JL (2008) BioMagResBank. Nucleic Acids Res 36(Database issue):

Biochemical Assessment of the Adhesome D402–D408. https://doi.org/10.1093/nar/ gkm957 11. Bouchet BP, Gough RE, Ammon YC, van de Willige D, Post H, Jacquemet G, Altelaar AM, Heck AJ, Goult BT, Akhmanova A (2016) Talin-KANK1 interaction controls the recruitment of cortical microtubule stabilizing complexes to focal adhesions. Elife 5. https://doi. org/10.7554/eLife.18124 12. Cowell AR, Jacquemet G, Singh AK, Paatero I, Brown DG, Ammon Y-C, Akhmanova A, Ivaska I, Goult B (2020) Talin Rod Domain Containing Protein 1 (TLNRD1), a novel actin bundling protein which promotes filopodia formation. BioRxiv. https://doi.org/10. 1101/2020.05.19.103606 13. Azizi L, Cowell AR, Mykuliak VV, Goult BT, Turkki P, Hytonen VP (2020) Cancer associated talin point mutations disorganise cell adhesion and migration. bioRxiv: 2020.2003.2025.008193. https://doi.org/ 10.1101/2020.03.25.008193 14. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9(7):671–675. https://doi.org/10.1038/nmeth.2089

147

15. Spudich JA, Watt S (1971) The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of the tropomyosin-troponin complex with actin and the proteolytic fragments of myosin. J Biol Chem 246(15):4866–4871 16. Anthis NJ, Wegener KL, Ye F, Kim C, Goult BT, Lowe ED, Vakonakis I, Bate N, Critchley DR, Ginsberg MH, Campbell ID (2009) The structure of an integrin/talin complex reveals the basis of inside-out signal transduction. EMBO J 28(22):3623–3632. https://doi. org/10.1038/emboj.2009.287 17. Niggli V, Kaufmann S, Goldmann WH, Weber T, Isenberg G (1994) Identification of functional domains in the cytoskeletal protein talin. Eur J Biochem 224(3):951–957. https:// doi.org/10.1111/j.1432-1033.1994. 00951.x 18. Camp D, Haage A, Solianova V, Castle WM, Xu QA, Lostchuck E, Goult BT, Tanentzapf G (2018) Direct binding of Talin to Rap1 is required for cell-ECM adhesion in Drosophila. J Cell Sci 131(24):jcs225144. https://doi. org/10.1242/jcs.225144

Chapter 10 Network Analysis of Integrin Adhesion Complexes Frederic Li Mow Chee and Adam Byron Abstract Cell-surface adhesion receptors mediate interactions with the extracellular matrix (ECM) to control many fundamental aspects of cell behavior, including cell migration, survival, and proliferation. Integrin adhesion receptors recruit structural and signaling proteins to form multimolecular adhesion complexes that link the plasma membrane to the actomyosin cytoskeleton. The assembly and turnover of adhesion complexes are tightly regulated, governed in part by the networks of physical protein interactions and functional signaling associations between components of the adhesome. Proteomic profiling of adhesion complexes has begun to reveal their molecular complexity and diversity. To interrogate the composition of cell–ECM adhesions, we detail herein an approach for the network analysis of adhesion complex proteomes. Integration of these proteomic data with adhesome databases in the context of predicted protein interactions enables the mapping of experimentally defined adhesion complex networks. Computational analysis of resultant network models can identify subnetworks of putative functionally linked adhesion protein communities. This approach provides a framework to predict functional adhesion protein relationships and generate new mechanistic hypotheses for further experimental testing. Key words Bioinformatics, Cell adhesion, Cell signaling, Data analysis, Integrins, Interaction networks, Network analysis, Proteomics

1

Introduction Cells interact with their microenvironment, including the extracellular matrix (ECM), soluble factors, and other cells, using transmembrane adhesion receptors, primarily integrins. The engagement of integrin adhesion receptors with their extracellular ligands triggers the clustering of integrins on the cell surface and the formation of intracellular, integrin-associated complexes of adhesion proteins. These adhesion complexes scaffold integrins to the contractile cytoskeleton and transmit biochemical and biomechanical signals bidirectionally across the plasma membrane [1–4]. The direct and indirect associations of different signaling and adaptor proteins with adhesion complexes orchestrate the processing of adhesion signals to control fundamental aspects of

Miguel Vicente-Manzanares (ed.), The Integrin Interactome: Methods and Protocols, Methods in Molecular Biology, vol. 2217, https://doi.org/10.1007/978-1-0716-0962-0_10, © Springer Science+Business Media, LLC, part of Springer Nature 2021

149

150

Frederic Li Mow Chee and Adam Byron

cell behavior, including differentiation, migration, and proliferation [5–8]. Dysregulation of these structural and signaling adhesion networks is implicated in various disease processes, such as cancer progression and metastasis [9–12]. 1.1 Integrin Adhesion Complexes

Numerous intracellular signaling and adaptor proteins associate with integrins in adhesion complexes, collectively termed the adhesome. A literature-curated adhesome database contains 232 proteins that have been associated with integrin adhesion complexes in various experimental settings [9, 13]. The development of biochemical methods for the isolation of ligand-induced adhesion complexes, coupled with quantitative mass spectrometry (MS), has enabled the proteomic characterization of integrin adhesion complexes [14–20]. These methodological advances revealed an unanticipated molecular complexity and diversity of adhesion complex proteomes [21–29], highlighting the assorted and dynamic nature of their roles in cell adhesion signaling and other cellular processes. The integration in silico of multiple adhesion complex proteomes from different cell types enabled the construction of an experimentally defined meta-adhesome database [30]. Using this database, a core set of 60 frequently identified integrin-associated proteins—a consensus adhesome—was established, from which emergent modules of interconnected adhesion proteins were identified [30]. Adhesion complex assembly is tightly and dynamically regulated, and the regulation of mechanosensitive protein interactions by biomechanical force plays a central role in the recruitment of adhesion proteins to integrins [31–35]. Modulation of adhesion protein conformation, including the relief of adhesion protein autoinhibition, is an important mechanism for the appropriate sequential assembly of adhesion complexes [36–39]. Linked to this, the formation of adhesion protein precomplexes, which may enable hierarchical adhesion protein interactions, appears to contribute to the spatiotemporal control of adhesion complex formation and maturation [40–42]. Thus, analysis and modelling of the subnetworks of protein interactions in adhesion complexes will advance our understanding of the molecular mechanisms regulating integrin-mediated cell adhesion. In addition, while many consensus adhesome components have well-defined roles in cell adhesion, several have been reported to localize and function at sites away from the cell–ECM interface, such as the nucleus [43– 46]. Interrogation of the interactions of adhesion proteins with binding partners that do not have primary roles in cell adhesion may yield new insights into the noncanonical functions of adhesion protein networks.

Adhesome Network Analysis

1.2 Network Biology Approaches

151

The interactions of proteins drive biological processes; indeed, many proteins have no function as unbound monomers. It is the precise and spatiotemporally defined association of proteins in complex networks that controls the transmission and processing of signaling information in the form of biochemical reactions. To enable the assessment of underlying structure in protein networks, the system can be represented as graphs consisting of nodes (proteins) connected by edges (associations). Various graph theoretic approaches have been developed to analyze biological network architecture and properties, such as subnetworks of proteins that participate in a common cellular process (termed modules) or wellconnected subnetworks that are perturbed under a given condition (termed active modules) [47–49]. Products of coexpressed genes have a propensity to interact with each other to form modules that are associated with a given cellular response. For example, integration of an early-embryogenesis interactome from Caenorhabditis elegans with transcriptional profiles at different developmental stages revealed that proximal proteins within the network tended to be coexpressed, and these proteins also phenoclustered, indicating that proximal proteins in the interactome may drive similar biological processes [50]. Similarly, proteins within disease-specific functional modules have been postulated to act collaboratively to drive disease phenotypes [51, 52]. Thus, the analysis of network topology can reveal important biological properties of subsets of proteins in an interactome. The networks of physical protein interactions and functional signaling associations between adhesion complex proteins play a central role in regulating adhesion complex assembly and function. In this chapter, a computational approach for the network analysis of adhesion complex proteomes is detailed (Fig. 1). First, the processing of adhesion complex proteomic data is described, which is necessary for the identification and quantification of isolated adhesion proteins and the appropriate normalization of the quantitative MS data that will be the input for the network analysis. Second, the integration of adhesion complex proteomic data with interactomic data is described, which enables the mapping of experimentally defined adhesion complex networks. Computational analysis of resultant network models can identify key network hubs and subnetworks of putative functionally linked adhesion protein communities. Results obtained using this approach enable the interrogation of the composition of cell– ECM adhesions, the prediction of functional adhesion protein relationships, and the generation of new mechanistic hypotheses for further experimental testing.

152

Frederic Li Mow Chee and Adam Byron

Fig.1 A pipeline for network analysis of adhesion complex proteomes. (a) To generate the proteomic data required for network analysis, adhesion complexes must be isolated and analyzed by quantitative MS. Several methodologies for the proteomic characterization of integrin adhesion complexes have been described elsewhere [14, 16, 17, 19, 20]. (b) Key stages of the network analysis workflow detailed herein. A proteomic data processing phase is followed by an adhesion network analysis phase. (c) Further analysis of network models, including integration with additional datasets, can predict functional associations, reveal new mechanistic insights, and prime further experiments

2

Materials For the computational pipeline detailed herein, the only equipment required is suitable computer infrastructure. A modern, mid-level personal computer is sufficient for the analysis of most adhesion networks. However, for the analysis of modularity of very large networks (e.g., the human interactome), computation time will benefit substantially from use of a high-performance compute cluster. There are several software packages required for running the analysis pipeline, as well as environments for executing R and

Adhesome Network Analysis

153

Python scripts. In addition, an MS dataset of adhesion complex proteomes, enabling the quantification of protein abundance in at least two experimental conditions, is required as input for processing and analysis. 2.1 Proteomic Data Processing

1. Personal computer (e.g., Linux-enabled machine) with at least mid-range, up-to-date system specifications (e.g., at least 8 GB memory, modern quad-core processor). 2. Web browser (e.g., Google Chrome). 3. Raw MS data files derived from proteomic analysis of isolated integrin adhesion complexes. 4. Data analysis software platform for identification and quantification of proteins from raw MS data (e.g., MaxQuant [53]). 5. Peptide search engine (e.g., Andromeda [54]). 6. Software framework, if required (e.g., Microsoft .NET Framework). 7. Vendor-specific raw MS data file reader library, if required (e.g., Thermo Fisher Scientific MSFileReader). 8. Data analysis software platform for statistical analysis of processed MS data (e.g., Perseus [55]). 9. Spreadsheet program (e.g., Microsoft Excel) (optional). 10. R or R integrated development environment (e.g., RStudio Desktop). 11. Normalyzer package (implemented online, accessed via http:// quantitativeproteomics.org/normalyzer [56]). 12. mice package (download at https://CRAN.R-project.org/ package¼mice [57]).

2.2 Adhesion Network Analysis

1. Personal computer (e.g., Linux-enabled machine) with at least mid-range, up-to-date system specifications (e.g., at least 8 GB memory, modern quad-core processor). 2. High-performance compute cluster access (optional). 3. Web browser (e.g., Google Chrome). 4. R or R integrated development environment (e.g., RStudio Desktop). 5. Protein interaction network edge list (e.g., BioGRID dataset, download at https://downloads.thebiogrid.org/BioGRID [58]). 6. BioNet package (download at https://bioconductor.org/ packages/BioNet [59, 60]). 7. heinz package (download at https://github.com/ls-cwi/ heinz).

154

Frederic Li Mow Chee and Adam Byron

8. LEMON library (download at https://lemon.cs.elte.hu/trac/ lemon/wiki/Downloads), OGDF library (download at https://ogdf.uos.de/releases), and CPLEX library (academic licence available, IBM). 9. louvain package (download at https://louvain-igraph. readthedocs.io/en/latest/index.html). 10. igraph package (download at https://igraph.org/r [61]). 11. Python or Python integrated development environment (e.g., Spyder) or notebook (e.g., Jupyter Notebook). 12. pandas library (download at https://pandas.pydata.org). 13. Java platform (minimum Oracle Java Standard Edition 8). 14. Cytoscape software (download at https://cytoscape.org/down load.html [62]). 15. CyRest package (download at https://github.com/idekerlab/ cy-rest-R/tree/develop/utility [63]). 16. brainGraph library (download at https://cran.r-project.org/ web/packages/brainGraph).

3

Methods The starting point for the protocol detailed herein is raw quantitative proteomic data derived from MS analysis of biochemically isolated adhesion complexes. Various methods exist for the proteomic characterization of integrin adhesion complexes, and these have been described elsewhere [14, 16, 17, 19, 20]. To enable adhesion complex network analysis, quantitative MS data are preprocessed and normalized and missing values are imputed (Fig. 1). Processed proteomic data are then used to build an interaction network model that can be analyzed using various computational approaches (Fig. 1). The entire computational pipeline for the network analysis of adhesion complex proteomes can be completed within 4 days: proteomic data processing requires 1–2 days (Subheading 3.1) and adhesion network analysis requires 1–2 days (Subheading 3.2), depending on dataset size and network complexity.

3.1 Proteomic Data Processing

MS data derived from unlabeled samples are used in this protocol because label-free methods of quantification are straightforward to implement, low cost, and broadly applicable to most cell systems. We use MS ion current–based label-free quantification (LFQ) data derived from a high-resolution mass spectrometer to enable accurate quantification with high sensitivity [64]. Owing to the lack of

Adhesome Network Analysis

155

an internal standard against which to correct experimental and technical variability in label-free methods, reproducible, wellcontrolled sample preparation and MS analysis are vital for accurate protein identification and quantification. It is important to determine the most appropriate normalization technique to reduce variation between and within experimental conditions and to eliminate systematic error [65]. Furthermore, despite technological advances in instrumentation and data analysis that improve sensitivity, missing data remains a considerable issue in MS datasets, which can result in loss of statistical power and introduce bias [66]. Therefore, it is also important to apply an appropriate data imputation technique to eliminate missing values in the dataset [65, 67, 68]. 3.1.1 Protein Identification and Quantification

1. Using a personal computer, launch the latest version of desired data analysis software for identification and quantification of proteins from raw MS data (e.g., MaxQuant and its integrated peptide search engine, Andromeda). Ensure the correct versions of any dependent software frameworks (e.g., .NET Framework) and vendor-specific libraries (e.g., MSFileReader) are installed (see Note 1). 2. In the Raw files tab in MaxQuant, load raw MS data files. 3. In the Group-specific parameters tab in MaxQuant, select a multiplicity of 1 for label-free quantification. Select the enzyme parameters as appropriate (e.g., Trypsin/P with a maximum of two missed cleavages). Select protein N-terminal acetylation and methionine oxidation as variable modifications (see Note 2). Select LFQ (label-free quantification). 4. In the Global parameters tab in MaxQuant, load the FASTA file corresponding to the database to be searched (see Note 3). Select carbamidomethylation of cysteine as a fixed modification. Enable matching between runs. 5. Concatenate the database to be searched with a decoy database containing reversed sequences from the original database. Accept peptide and protein false-discovery rates of 1%. 6. Click the Start button to begin the analysis. 7. In the Viewer tab in MaxQuant, visually inspect and quality control the processed MS data. 8. Launch the latest version of desired data analysis software for statistical analysis of processed MS data (e.g., Perseus). 9. Using the Generic matrix upload function in Perseus, load the MaxQuant proteinGroups.txt results file (saved in the MaxQuant . . .\combined\txt folder). Select the columns that contain LFQ intensities as main columns.

156

Frederic Li Mow Chee and Adam Byron

10. Using the Filter rows based on categorical column function in Perseus, remove proteins only identified by site, potential contaminants, and reverse database hits. 11. Using the Rename columns function in Perseus, edit column names to improve readability or relevance if necessary. 12. Using the Categorical annotation rows function in Perseus, group samples by like experimental conditions. 13. Using the Filter rows based on valid values function in Perseus, filter the dataset to exclude proteins that do not meet a given threshold of sample identification (e.g., for an experiment with four biological replicates, require protein identification in at least three replicates for all experimental conditions) (see Note 4). 14. Export the matrix of protein LFQ data to a tab-delimited text file using the Generic matrix export function in Perseus. 15. In a spreadsheet program or R, save the matrix exported from Perseus as a new file, and reformat it to give a single top row of headers. Discard all columns except for the protein identifiers (or associate gene names) and the protein LFQ data for all relevant samples. 3.1.2 Data Normalization and Missing-Value Imputation

1. In a spreadsheet program or R, save the reformatted matrix exported from Perseus as a new file, and insert a new top row of headers defining sample group, such that samples from the same experimental condition are labeled with the same number and different experimental conditions are labeled with different numbers. Label the protein identifiers header with “0”. 2. Using a web browser, access the online implementation of the Normalyzer package (see Notes 5 and 6). 3. Under the Normalize tab in Normalyzer, import the reformatted dataset, assign the project a name, and click the Submit button. 4. Using the normalization report (saved in the compressed Normalyzer output folder), select the most appropriately normalized dataset, basing the decision on the different diagnostic plots (e.g., MA plot, relative log expression plot). 5. Launch RStudio. Set path and load libraries. 6. In the Environment/History panel in RStudio, under the Environment tab, import the normalized dataset as a data frame, with protein names (“Gene.names”) set as row names and sample names (incorporating experimental condition and replicate) set as column names. Let this data frame be called “data.”

Adhesome Network Analysis

157

7. Missing-not-at-random single-value imputation is performed using the minimum LFQ intensity value in the respective replicates if, and only if, LFQ values for all replicates of the same condition are missing (see Notes 7 and 8). As in the code that follows, use a for loop to iterate through each row in the dataset, and use a conditional statement to test whether all replicates within a condition are missing; if they are, impute with the lowest LFQ value within the replicate. for (i in 1:nrow(data)) { if (is.na(data[i, A:B]) == TRUE) { data[i, A:B]