Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The Rap GTPases coordinate actin and microtubule cytoskeleton reorganization, and promote antigen extraction… Wang, Jia Chao 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_february_wang_jia.pdf [ 12.27MB ]
Metadata
JSON: 24-1.0363009.json
JSON-LD: 24-1.0363009-ld.json
RDF/XML (Pretty): 24-1.0363009-rdf.xml
RDF/JSON: 24-1.0363009-rdf.json
Turtle: 24-1.0363009-turtle.txt
N-Triples: 24-1.0363009-rdf-ntriples.txt
Original Record: 24-1.0363009-source.json
Full Text
24-1.0363009-fulltext.txt
Citation
24-1.0363009.ris

Full Text

THE RAP GTPASES COORDINATE ACTIN AND MICROTUBULE CYTOSKELETON REORGANIZATION, AND PROMOTE ANTIGEN EXTRACTION AT THE B CELL IMMUNE SYNAPSE by  Jia Chao Wang  B.Sc., The University of British Columbia, 2010  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Microbiology and Immunology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   December 2017  © Jia Chao Wang, 2017 ii  Abstract  B cells that bind antigens displayed on antigen-presenting cells (APCs) form an immune synapse (IS), a polarized cellular structure that optimizes the dual functions of the B cell receptor (BCR), signal transduction and antigen internalization. Immune synapse formation involves the polarization of the microtubule-organizing center (MTOC) towards the APC. I showed that BCR-induced MTOC polarization requires the Rap1 GTPases (which has two isoforms, Rap1a and Rap1b), an evolutionarily conserved regulator of cell polarity, as well as cofilin-1, an actin-severing protein that is regulated by Rap1. MTOC reorientation towards the antigen contact site correlated strongly with cofilin-1-dependent actin reorganization and cell spreading. I also showed that BCR-induced MTOC polarization requires the dynein motor protein as well as IQGAP1, a scaffolding protein that can link the actin and microtubule cytoskeletons. At the periphery of the immune synapse, IQGAP1 associates closely with F-actin structures and with the microtubule plus-end-binding protein CLIP-170. Moreover, the accumulation of IQGAP1 at the antigen contact site depends on F-actin reorganization that is controlled by Rap1 and cofilin-1. I also demonstrate that the hematopoietic-cell specific cortactin-homologue, HS1, is essential for regulating actin cytoskeletal remodeling during immune synapse formation and acts downstream of Rap to promote BCR-induced antigen gathering. Additionally, inhibiting the Rap1-cofilin-1 pathway, CLIP-170 expression, or cytoskeletal dynamics impairs the ability of B cells to acquire antigens from APCs. Thus, Rap1 coordinates actin and microtubule organization at the IS, facilitating antigen acquisition from APCs.     iii  Lay Summary   B cells are a type of cell of the immune system that make protective antibodies when they are activated by foreign agents. Antibodies are important for defending the body from infectious agents. B cell activation is tightly controlled because the inappropriate activation of B cells can be harmful to the body and cause autoimmune disease. This study investigates the role of the cytoskeleton and its regulators in controlling B cell activation. When B cells recognize a foreign agent, the cytoskeleton can rapidly change its structure and location within the cell. The cytoskeleton reorganizes cellular components into positions that promote B cell activation. Using loss of function approaches and microscopy imaging, I identified proteins that control cytoskeletal reorganization in B cells. This study provides new insights into how cytoskeletal reorganization supports B cell activation. iv  Preface Status of data chapters and my contributions to them The material contained in this thesis is derived from the following publications or manuscripts:  A version of Chapters 3 and 4 has been published.  Jia C. Wang, Jeff Y.-J. Lee, Sonja Christian, May Dang-Lawson, Caitlin Pritchard, Spencer A. Freeman, Michael R. Gold, The Rap1-cofilin pathway coordinates actin reorganization and MTOC polarization at the B cell immune synapse. J Cell Sci., 2017, 130(6):1094-1109.  I performed and analyzed all the experiments contained in these chapters except for those in Figure 3.5E-G, which were performed by Jeff Y.-J. Lee. For Figure 3.16, the cell stretching was performed by Sonja Christian but I carried out the imaging and data analysis. May Dang-Lawson sequenced some of the plasmid constructs that I used in the experiments contained in these chapters. Caitlin Pritchard and I performed and analyzed the supplementary experiments for this publication. Spencer Freeman was involved in the early stages of concept formation. Michael Gold and I conceived the experiments and wrote the published paper.  A portion of Chapter 3 has been submitted as an invited research article. This manuscript has been peer-reviewed and is currently under revision.  Jia C. Wang, Jeff Y.-J. Lee, Caitlin Pritchard, and Michael R. Gold, The roles of the Rap1 and Rap2c GTPases in B cell receptor-induced reorientation of the microtubule-organizing center.  I performed and analyzed all the experiments contained in this chapter except for Figure 3.10A, where I performed the experiment but Caitlin Pritchard performed the western transfer, and v  Figure 3.10B, which was performed and analyzed by Jeff Y.-J. Lee. Michael Gold and I conceived the experiments and wrote the manuscript.   Chapter 5 is being prepared for submission in January 2018.  Jia C. Wang and Michael R. Gold, HS1 regulates BCR-induced actin reorganization and IS formation. I conceived, performed and analyzed all the experiments in this chapter except for Figure 5.1A, which was performed by Kate Choi. I wrote the manuscript and Michael Gold contributed to the manuscript edits.  A version of Chapter 6 is being prepared for submission in January 2018. Jia C. Wang, May Dang-Lawson, and Michael R. Gold, The Rap1-cofilin pathway and the B cell cytoskeleton are important for antigen acquisition from APCs.  I performed and analyzed all the experiments in this chapter. The plasmid construct shown in Figure 6.9 was made by May Dang-Lawson. Pauline Johnson provided the OT-II mice and contributed to the early stages of experimental design for the T cell activation experiments. Michael Gold and I conceived the experiments for this manuscript. I wrote the manuscript and Michael Gold contributed to the manuscript edits.  I conceived, analyzed and performed all the experiments contained in the Appendices except for Figure A.1, which was performed by Jeff Y.-J. Lee.   vi  Animal studies were conducted in the Modified Barrier Facility at the University of British Columbia. All animal work was performed under strict accordance with the recommendations of the Canadian Council for Animal Care, certification number 6615 – 15. Protocols were approved by the Animal Care Committee (ACC) (A15-0162) of the University of British Columbia.  vii  Table of Contents  Abstract .......................................................................................................................................... ii Lay Summary ............................................................................................................................... iii Preface ........................................................................................................................................... iv Table of Contents ........................................................................................................................ vii List of Tables .............................................................................................................................. xvi List of Figures ............................................................................................................................ xvii List of Abbreviations ................................................................................................................. xxi Acknowledgements ................................................................................................................. xxvii Dedication ................................................................................................................................. xxix Chapter 1: Introduction ................................................................................................................1 1.1 The role of B cell activation in immune responses ......................................................... 1 1.1.1 BCR signaling pathways ............................................................................................. 2 1.2 Antigen presentation to B cells ....................................................................................... 6 1.2.1 B cells migrate to secondary lymphoid organs to encounter antigen ......................... 6 1.2.2 Antigen capture and presentation in the secondary lymphoid organs ........................ 9 1.2.2.1 Follicular dendritic cells ................................................................................... 11 1.3 The B cell immune synapse .......................................................................................... 13 1.3.1 BCR organization...................................................................................................... 13 1.3.2 B cells that bind to cognate antigen form signaling BCR microclusters .................. 16 1.3.3 B cells form an immune synapse with antigen-presenting cells ............................... 17 1.4 The antigen acquisition function of the B cell immune synapse .................................. 22 viii  1.4.1 Immune synapse formation regulates BCR signaling by promoting BCR internalization ....................................................................................................................... 22 1.4.2 B cells acquire antigens from APCs and present them to T cells ............................. 24 1.4.3 B cells present antigens to gain T cell help............................................................... 26 1.4.4 B cells extract antigens from APC membranes ........................................................ 29 1.4.5 BCR-mediated antigen endocytosis and trafficking ................................................. 31 1.5 The actin cytoskeleton .................................................................................................. 32 1.5.1 Actin structures ......................................................................................................... 33 1.5.2 Regulators of actin polymerization ........................................................................... 37 1.5.2.1 Profilin .............................................................................................................. 37 1.5.2.2 Nucleation factors promote actin filament assembly ........................................ 38 1.5.3 Arp2/3-mediated polymerization of dendritic actin networks .................................. 40 1.5.3.1 HS1 and Arp2/3 regulate the formation of dendritic actin networks ................ 41 1.5.3.2 The role of HS1 in immune cells ...................................................................... 44 1.5.4 Small GTPases regulate actin nucleation .................................................................. 45 1.5.5 Actin filament disassembly ....................................................................................... 46 1.5.5.1 The actin-severing protein cofilin regulates actin dynamics ............................ 47 1.5.5.2 Regulation of cofilin ......................................................................................... 49 1.5.6 The actin cytoskeleton and immune synapse formation ........................................... 52 1.6 The MTOC and cell polarity ......................................................................................... 56 1.6.1 The MTOC and its roles in cell polarity ................................................................... 59 1.6.1.1 The structure of the centrosomal MTOC .......................................................... 60 1.6.1.2 Establishing MTOC polarity ............................................................................. 62 ix  1.6.2 MTOC polarization in lymphocytes ......................................................................... 64 1.6.2.1 Mechanisms underlying MTOC reorientation in lymphocytes ........................ 67 1.7 The roles of IQGAP1 and the Rap GTPases in cytoskeletal regulation ....................... 70 1.7.1 The IQGAP1 scaffolding protein .............................................................................. 70 1.7.1.1 The expression and structure of IQGAP1 ......................................................... 71 1.7.1.2 IQGAP1 regulates cytoskeletal organization .................................................... 72 1.7.2 The Rap GTPases regulate cell polarity and cytoskeletal remodeling ..................... 73 1.7.2.1 The Rap GTPases .............................................................................................. 74 1.7.3 Rap GEFs and GAPs ................................................................................................. 75 1.7.4 Rap GTPases localize to cellular membranes ........................................................... 77 1.7.5 Rap1 regulates cell adhesion and actin cytoskeleton remodeling ............................. 78 1.7.6 Rap activation and cell polarity ................................................................................ 79 1.7.7 Rap1 in B cells .......................................................................................................... 81 1.8 Hypothesis and Specific Aims ...................................................................................... 85 Chapter 2: Methods .....................................................................................................................86 2.1 Cell isolation and culture .............................................................................................. 86 2.1.1 Primary B cell isolation and culture.......................................................................... 86 2.1.2 B cell lines and culture .............................................................................................. 86 2.1.3 T cell isolation and culture ........................................................................................ 87 2.2 Transfections and transductions .................................................................................... 87 2.2.1 Transient transfection of B cell lines ........................................................................ 87 2.2.2 Retrovirus-mediated transduction ............................................................................. 88 2.3 siRNA- and shRNA-mediated gene silencing .............................................................. 89 x  2.3.1 siRNA knockdown .................................................................................................... 89 2.3.2 Lentivirus-mediated expression of shRNAs ............................................................. 90 2.4 Immunoblotting............................................................................................................. 91 2.5 Assessing BCR-mediated endocytosis using flow cytometry ...................................... 92 2.6 Cytoskeletal inhibitors .................................................................................................. 93 2.6.1 Pharmacological inhibitors ....................................................................................... 93 2.6.2 Cofilin inhibitory peptides ........................................................................................ 94 2.7 B cell interactions with anti-Ig-coated beads ................................................................ 94 2.8 B cell spreading on planar surfaces .............................................................................. 95 2.8.1 B cell spreading on coated coverslips ....................................................................... 95 2.8.2 Mechanical stretching of B cells ............................................................................... 95 2.9 B cell:APC interactions................................................................................................. 96 2.9.1 Antigen-presenting cells ........................................................................................... 96 2.9.2 B cell interactions with APCs in suspension ............................................................ 96 2.9.3 B cell interactions with adherent APCs .................................................................... 97 2.9.4 T cell activation by APC-activated B cells ............................................................... 97 2.10 Cell staining for fluorescence microscopy .................................................................... 98 2.10.1 Immunostaining for fluorescence microscopy ...................................................... 98 2.10.2 Fluorescent stains for visualizing cellular structures .......................................... 100 2.10.3 Immunostaining for STED super resolution microscopy ................................... 102 2.10.4 Immunostaining for ground state depletion super-resolution microscopy imaging of microtubules ................................................................................................................... 102 2.10.5 Proximity ligation assays .................................................................................... 103 xi  2.11 Microscopy and image analysis .................................................................................. 103 2.11.1 Confocal microscopy .......................................................................................... 103 2.11.2 TIRF microscopy ................................................................................................ 104 2.11.3 STED microscopy ............................................................................................... 104 2.11.4 Ground state depletion microscopy .................................................................... 104 2.11.5 Scanning electron microscopy ............................................................................ 105 2.11.6 Fluorescence recovery after photobleaching ...................................................... 105 2.12 Quantification of MTOC polarization ........................................................................ 106 2.12.1 Quantification of F-actin and IQGAP1 at bead:cell contact sites ....................... 109 2.12.2 Quantification of internalized antigen from APCs ............................................. 110 2.13 Statistical analyses ...................................................................................................... 111 Chapter 3: The Rap1-Cofilin-1 pathway regulates MTOC reorientation towards the B cell immune synapse .........................................................................................................................112 3.1 Introduction ................................................................................................................. 112 3.1.1 Cytoskeletal reorganization drives B cell immune synapse formation ................... 112 3.1.2 MTOC polarization supports immune synapse formation ...................................... 113 3.1.3 Polarity proteins control the polarization of the MTOC ......................................... 113 3.1.4 The functions of Rap1 and Rap2 proteins ............................................................... 115 3.1.5 Cofilin-mediated actin remodeling during B cell activation ................................... 115 3.1.6 Rationale and hypothesis ........................................................................................ 116 3.2 Results ......................................................................................................................... 117 3.2.1 Microtubules and Rap1 are required for cSMAC formation .................................. 117 3.2.2 BCR clustering induces MTOC polarization towards anti-Ig-coated beads ........... 120 xii  3.2.3 Microtubule dynamics and PKCζ activity are important for BCR-induced MTOC reorientation ........................................................................................................................ 124 3.2.4 BCR-induced MTOC polarization towards anti-Ig-coated beads depends on Rap1 activation ............................................................................................................................. 127 3.2.5 MTOC reorientation towards APCs depends on Rap1 ........................................... 130 3.2.6 Rap1 and Rap2 play non-redundant roles in BCR-induced MTOC reorientation .. 138 3.2.7 MTOC polarization requires cofilin-mediated actin severing ................................ 141 3.3 Discussion ................................................................................................................... 155 3.3.1 Summary of findings............................................................................................... 155 3.3.2 The role of Rap1 and Rap2c isoforms in MTOC reorientation .............................. 156 3.3.3 The Rap1-cofilin pathway is important for MTOC reorientation ........................... 159 3.3.4 Perspectives............................................................................................................. 163 Chapter 4: IQGAP1 and CLIP-170 couple actin reorganization to MTOC polarization ..164 4.1 Introduction ................................................................................................................. 164 4.2 Results ......................................................................................................................... 167 4.2.1 BCR-induced MTOC polarization requires IQGAP1 and CLIP-170 ..................... 167 4.2.2 IQGAP1 and CLIP-170 are located at microtubule-actin interfaces ...................... 170 4.2.3 Rap-dependent actin reorganization promotes IQGAP1 accumulation at the periphery of the immune synapse ....................................................................................... 180 4.2.4 BCR-induced MTOC polarization requires dynein activity ................................... 185 4.3 Discussion ................................................................................................................... 187 4.3.1 Summary of findings............................................................................................... 187 4.3.2 IQGAP1 localizes to the antigen contact site ......................................................... 187 xiii  4.3.3 IQGAP1 may facilitate MTOC movement by capturing microtubule plus ends at the cell cortex ............................................................................................................................ 189 4.3.3.1 The cortical capture of microtubule plus ends ................................................ 190 4.3.3.2 Force generation promotes MTOC reorientation and polarization ................. 194 4.3.4 Perspectives............................................................................................................. 196 Chapter 5: HS1 regulates BCR-induced actin reorganization and IS formation ................198 5.1 Introduction ................................................................................................................. 198 5.2 Results ......................................................................................................................... 200 5.2.1 BCR-induced tyrosine phosphorylation of HS1 depends on the Rap GTPases...... 200 5.2.2 Phosphorylated-HS1 is enriched at the antigen contact site ................................... 202 5.2.3 The localization of HS1 at the cell periphery depends on F-actin .......................... 206 5.2.4 HS1 is important for BCR signaling at the immune synapse ................................. 207 5.2.5 HS1 is required for APC-induced cSMAC formation ............................................ 210 5.2.6 HS1 is important for the assembly of the peripheral actin cytoskeleton during B cell spreading ............................................................................................................................. 213 5.2.7 HS1 promotes the formation of sustained membrane protrusions at the B cell periphery ............................................................................................................................. 217 5.2.8 HS1 controls the actin organization around BCR microclusters ............................ 221 5.2.9 HS1 is important for MTOC reorientation towards the B cell IS ........................... 225 5.3 Discussion ................................................................................................................... 228 5.3.1 Summary of findings............................................................................................... 228 5.3.2 HS1 phosphorylation .............................................................................................. 228 5.3.3 HS1 controls actin dynamics in B cells .................................................................. 230 xiv  5.3.4 HS1 regulates IS formation in B cells..................................................................... 234 5.3.5 HS1 regulates BCR signaling ................................................................................. 236 5.3.6 Perspectives............................................................................................................. 237 Chapter 6: The Rap1-cofilin pathway and the B cell cytoskeleton are important for antigen acquisition from APCs ...............................................................................................................238 6.1 Introduction ................................................................................................................. 238 6.2 Results ......................................................................................................................... 240 6.2.1 Antigen acquisition from APCs requires MTs and actin ........................................ 240 6.2.2 Lysosome accumulation near the antigen contact site depends on Rap1 ............... 247 6.2.3 B cells acquire antigens from APCs in a Rap1-dependent manner ........................ 248 6.2.4 Rap1 is dispensable for the internalization of soluble antigen ............................... 251 6.2.5 Actin remodeling proteins are required for antigen internalization from APCs ..... 253 6.2.6 LPS-activated B cells extract antigen more efficiently than resting B cells ........... 255 6.2.7 B cells can present antigens that are acquired from APCs to T cells ..................... 257 6.3 Discussion ................................................................................................................... 261 6.3.1 Summary of findings............................................................................................... 261 6.3.2 Antigen acquisition at the IS ................................................................................... 262 6.3.3 BCR-mediated antigen extraction ........................................................................... 265 6.3.4 BCR-mediated internalization of extracted antigen ................................................ 268 6.3.5 Antigen trafficking to degradative compartments .................................................. 269 6.3.6 Antigen presentation to T cells ............................................................................... 270 6.3.7 Perspectives............................................................................................................. 272 Chapter 7: Overall discussion and perspectives .....................................................................274 xv  7.1 Summary ..................................................................................................................... 274 7.2 Rap regulates cofilin activation .................................................................................. 276 7.2.1 Rap isoforms and cofilin activation ........................................................................ 276 7.2.2 The mechanisms by which Rap1 may regulate cofilin dephosphorylation ............ 277 7.2.3 Negative regulation of cofilin ................................................................................. 278 7.3 Rap1 localization during IS formation ........................................................................ 279 7.3.1 Rap GEFs and Rap1 activation ............................................................................... 280 7.3.2 Scaffolding proteins and Rap1 localization ............................................................ 281 7.4 The role of B cell immune synapse formation in disease ........................................... 282 7.5 Conclusions ................................................................................................................. 288 Bibliography ...............................................................................................................................289 Appendices ..................................................................................................................................322 Appendix A ............................................................................................................................. 322 A.1 siRNA-mediated depletion of Rap1a and Rap1b proteins in B cells ...................... 322 Appendix B ............................................................................................................................. 323 B.1 Super-resolution microscopy images of microtubules ............................................ 323 Appendix C ............................................................................................................................. 325 C.1 The Arp2/3 complex is important for B cells to acquire antigen from APCs......... 325  xvi  List of Tables Table 2.1 Plasmid constructs used to express proteins in B cells. ................................................ 88 Table 2.2 siRNAs .......................................................................................................................... 89 Table 2.3 shRNAs ......................................................................................................................... 90 Table 2.4  Antibodies for immunoblotting ................................................................................... 92 Table 2.5 Inhibitors targeting cytoskeletal regulators ................................................................... 93 Table 2.6 Cofilin inhibitory peptide sequences ............................................................................ 94 Table 2.7 Primary antibodies used for immunostaining ............................................................... 99 Table 2.8 Secondary antibodies used for immunostaining ......................................................... 100 Table 2.9  Cellular stains used for detecting cellular components ............................................. 101 Table 2.10 Percent of cells with polarity index that would randomly be ≤ 1.0, 0.75 or 0.5 when 4.5-μm beads were used .............................................................................................................. 109 Table 2.11 Percent of cells with polarity index that would randomly be ≤ 1.0 or 0.75 when 3-μm beads were used. ......................................................................................................................... 109  xvii  List of Figures Figure 1.1. BCR signaling pathway ................................................................................................ 5 Figure 1.2. Lymph node structure ................................................................................................... 8 Figure 1.3. APCs in the LNs capture and present antigen to B cells ............................................ 11 Figure 1.4. Actin-dependent BCR mobility and IS formation ...................................................... 21 Figure 1.5. Naïve B cells acquire antigen from APCs after forming an IS .................................. 24 Figure 1.6. B cells acquire antigens to gain T cell help ................................................................ 26 Figure 1.7. T cell-dependent B cell activation .............................................................................. 29 Figure 1.8. Actin structures in a migrating cell ............................................................................ 36 Figure 1.9. Mechanisms of actin assembly ................................................................................... 39 Figure 1.10. The structures of cortactin and HS1 ......................................................................... 43 Figure 1.11. Model for the association of HS1 with the Arp2/3 complex and F-actin ................. 44 Figure 1.12. Cofilin-mediated actin severing ............................................................................... 49 Figure 1.13. Cofilin regulation...................................................................................................... 52 Figure 1.14. B cells form a polarized IS structure ........................................................................ 58 Figure 1.15. The structure of the centrosome ............................................................................... 61 Figure 1.16. Proteins involved in the cortical capture of microtubules ........................................ 64 Figure 1.17. The MTOC is reoriented towards the antigen contact site during immune synapse formation ....................................................................................................................................... 66 Figure 1.18. Mechanisms that can reorient the MTOC in lymphocytes ....................................... 70 Figure 1.19. IQGAP1 structure and protein interaction motifs .................................................... 72 Figure 1.20. Rap activation ........................................................................................................... 75 Figure 1.21. BCR-mediated activation of Rap GTPases .............................................................. 84 xviii  Figure 2.1. Quantification of MTOC polarization ...................................................................... 107 Figure 2.2. Quantification of antigen internalization .................................................................. 110 Figure 3.1. B cells extend lamellipodia across antigen-bearing surfaces ................................... 121 Figure 3.2. Primary B cells reorient the MTOC towards anti-Ig-coated beads .......................... 124 Figure 3.3. Nocodazole and paclitaxel block BCR-induced MTOC polarization ...................... 125 Figure 3.4. PKCζ activity is required for BCR-induced MTOC polarization towards anti-Ig-coated beads ................................................................................................................................ 126 Figure 3.5. Rap1 is important for MTOC polarization towards anti-Ig-coated beads ................ 128 Figure 3.6. Rap activation is required for MTOC polarization towards anti-Ig-coated beads ... 130 Figure 3.7. Rap1 is important for MTOC polarization towards APCs ....................................... 132 Figure 3.8. Rap activation is required for MTOC polarization towards APCs........................... 133 Figure 3.9. Relationship of BCR signaling to MTOC polarization ............................................ 137 Figure 3.10. siRNA-mediated depletion of either Rap1a/b or Rap2c impairs BCR-induced MTOC reorientation.................................................................................................................... 141 Figure 3.11. BCR-induced MTOC polarization requires an intact actin cytoskeleton ............... 142 Figure 3.12. Rap1 depletion inhibits BCR-induced cofilin dephosphorylation .......................... 143 Figure 3.13. Cofilin controls BCR-induced MTOC polarization ............................................... 146 Figure 3.14. Cofilin-mediated actin reorganization is required for the MTOC to approach the plasma membrane ....................................................................................................................... 149 Figure 3.15. Cofilin knockdown blocks actin reorganization and MTOC polarization ............. 150 Figure 3.16. Cell spreading alone is not sufficient to elicit MTOC polarization........................ 154 Figure 4.1. IQGAP1 and CLIP-170 are required for BCR-induced MTOC reorientation ......... 169 xix  Figure 4.2. Expressing the C-terminal fragment of IQGAP1 (IQGAP1-CT) inhibits BCR-induced MTOC reorientation.................................................................................................................... 169 Figure 4.3. Super-resolution images of B cell microtubules and F-actin ................................... 171 Figure 4.4. STED super-resolution images of actin, microtubules, and CLIP-170 at the antigen contact site .................................................................................................................................. 175 Figure 4.5. Co-localization of IQGAP1, CLIP-170, and F-actin ................................................ 176 Figure 4.6. IQGAP1 and CLIP-170 localize at the cell periphery .............................................. 179 Figure 4.7. Rap1 promotes IQGAP1 accumulation at the IS by controlling actin organization 183 Figure 4.8. IQGAP1 is dispensable for B cell spreading and actin reorganization .................... 184 Figure 4.9. Dynein is required for MTOC polarization .............................................................. 186 Figure 5.1. HS1 is phosphorylated upon BCR stimulation with soluble anti-Ig antibodies ....... 201 Figure 5.2. Phosphorylated HS1 is localized at the antigen-contact site with anti-Ig-coated beads..................................................................................................................................................... 203 Figure 5.3. Phosphorylated HS1 is localized at the cell periphery in cells spreading on immobilized anti-Ig antibodies ................................................................................................... 205 Figure 5.4. Clones of A20 and A20/D1.3 cells transduced with HS1 shRNA ........................... 208 Figure 5.5. CD79 phosphorylation at the antigen contact site is dependent on HS1 .................. 209 Figure 5.6. HS1 depletion blocks the formation and coalescence of BCR microclusters .......... 212 Figure 5.7. HS1 is important for B cell spreading and for the formation of peripheral F-actin structures ..................................................................................................................................... 215 Figure 5.8. HS1 promotes the assembly of the peripheral F-actin structures ............................. 216 Figure 5.9. HS1 knockdown does not affect actin dynamics ...................................................... 218 Figure 5.10. Peripheral actin protrusions are less stable in HS1-depleted B cells ...................... 220 xx  Figure 5.11. Actin-dense clusters form around BCR-antigen clusters at the interface between a B cell and an APC .......................................................................................................................... 224 Figure 5.12. HS1 is required for BCR-induced MTOC reorientation towards the B cell IS ...... 226 Figure 5.13. The Arp2/3 inhibitor CK-666 blocks BCR-induced MTOC polarization .............. 227 Figure 5.14. Model for the role of HS1 in IS formation. ............................................................ 233 Figure 6.1. Antigen internalized from APCs is trafficked to early endosomes .......................... 242 Figure 6.2. Internalization of antigens extracted from APCs requires the microtubule and actin networks as well as PKCζ activity .............................................................................................. 245 Figure 6.3. CLIP-170 is required for antigen acquisition from APCs ........................................ 246 Figure 6.4. Lysosome accumulation near the antigen contact site depends on Rap1 ................. 247 Figure 6.5. Rap1 is important for primary B cells to internalize antigens that are acquired from APCs ........................................................................................................................................... 250 Figure 6.6. Rap activation is not required for the internalization of soluble antigens ................ 253 Figure 6.7. Cofilin is required for antigen internalization from APCs ....................................... 254 Figure 6.8. LPS-activated primary B cells acquire more antigen than BAFF-treated cells........ 256 Figure 6.9. Experimental system for B cell-mediated T cell activation ..................................... 258 Figure 6.10. OT-II T cells exhibit increased intracellular IL-2 levels in response to antigen presented by B cells .................................................................................................................... 260 Figure 6.11. Increased IL-2 expression in OT-II cells cultured with MD4 B cells and APCs presenting HEL-OVA ................................................................................................................. 261 Figure 6.12. Model for Rap- and cofilin-dependent antigen extraction ..................................... 267 Figure 7.1. Model for Rap and cofilin-dependent MTOC reorientation during IS formation .... 275 Figure 7.2. Aberrant B cell signaling during infection and disease ............................................ 286 xxi  List of Abbreviations  +TIPs: plus-end tracking proteins A: alanine Abl: Abelson kinase Abp1: actin binding protein 1 ACC: animal care committee ADF: actin depolymerizing factor Ag: antigen APC: antigen-presenting cell APC: adenomatous polyposis coli aPKC: atypical protein kinase C BLNK: B cell linker protein BSA: bovine serum albumin CBL: Casitas B-lineage lymphoma CC: coiled coil CCP: clathrin-coated pit Cdc42: cell division cycle 42 CFP: cyan fluorescent protein CFSE: 5(6)-Carboxyfluorescein N-hydroxysuccinimidyl ester CHD: calponin homology domain CK2: casein kinase 2 CME: clathrin-mediated endocytosis xxii  CR3: complement receptor 3 cSMAC: central supramolecular activation cluster D: Aspartic Acids DAG: diacylglycerol DC: dendritic cell DMEM: Dulbecco’s Modified Eagle Medium dSMAC: distal supramolecular activation cluster dSTORM: direct stochastic optical reconstruction microscopy EAE: experimental autoimmune encephalomyelitis EB: end-binding protein EEA-1: early endosome antigen-1 F-actin: filamentous actin FAK: focal adhesion kinase FCS: fetal calf serum FDC: follicular dendritic cell FERM: four-point-one, ezrin, radixin, moesin FN: fibronectin FRAP: fluorescence recovery after photobleaching FRET: fluorescent resonance energy transfer G-actin: globular actin GAP: GTPase-activating protein GEF: guanine nucleotide exchange factor Grb2: growth factor receptor-bound protein 2 xxiii  GRD: GAP-related domain GSD: ground state depletion GSDIM: ground state depletion followed by individual molecule return GST: glutathione-S-transferase HDAC6: histone deacetylase 6 HEL: hen egg lysozyme HIV: human immunodeficiency virus HS1: hematopoietic lineage cell-specific protein-1 HTH: helix-turn-helix IC: immune complex Ig: immunoglobulin IP3: inositol trisphosphate ITAM: immunoreceptor tyrosine activation motifs JMY: junction-mediating and -regulatory protein JNK: jun N-terminal kinase KD: knockdown Lat A: latrunculin A LIMK: LIM Kinase LIS1: lissencephaly 1 MFI: mean fluorescence intensity mHBS: modified HEPES-buffered saline mHEL: membrane hen egg lysozyme MIIC: MHCII-containing multivesicular bodies xxiv  MTOC: microtubule organizing center NHE1: sodium/hydrogen exchanger 1 NK cell: natural killer cell NPF: nucleation-promoting factor NTA: N-terminal arginine N-WASP: neural Wiskott–Aldrich Syndrome protein OVA: ovalbumin PAK: p21-activated kinase PCAF: p300/CBP-associated factor lysine acetyltransferase PCM: pericentriolar material PFA: paraformaldehyde PI(s): polarity index or polarity indices (plural form) PI3K: phosphoinositide 3-kinase PIP2: phosphatidylinositol 4,5-bisphosphate PIP3: phosphatidylinositol (3,4,5)-trisphosphate PIPKγ: phosphoinositol-4-monophosphatase 5 kinase type Iγ PKD: protein kinase D PLCγ2: phospholipase C gamma 2 PLL: poly-L-lysine Pro: proline pSMAC: peripheral supramolecular activation cluster P-Tyr: phosphotyrosine R: amino acid repeat xxv  RapGAPII: Rap GTPase-activating protein II Rac1: Ras-related C3 botulinum toxin substrate 1 RGCT: Ras GAP C-terminus ROCK: Rho-associated protein kinase ROI: region of interest S: serine S1P: sphingosine-1-phosphate SCS: subcapsular sinus SEM: scanning electron microscopy SFK: Src family kinase SH: Src homology Sirt1: sirtuin1 SLE: systemic lupus erythematosus SLO: secondary lymphoid organ SSH: Slingshot STED: stimulated emission depletion TCR: T cell receptor TIRFM: total internal reflection fluorescence microscopy TLR: toll-like receptor WASP: Wiscott-Aldrich Syndrome protein WT: wild type WIP: WASP-interacting protein Y: tyrosine xxvi  YFP: yellow fluorescent protein Zap-70: ζ-chain-associated protein kinase of 70 kDa    xxvii  Acknowledgements  I would like to offer my heartfelt gratitude to the faculty, staff and students at UBC who have inspired and supported my development as a scientist and as an individual during my graduate studies. My studies and my research could not have been possible without funding support from the CIHR, NSERC and UBC.  I am especially indebted to my supervisor, Dr. Michael Gold, who always believed in my vision and my potential to be a scientist. I couldn’t have accomplished all that I have and maintained the courage to pursue my passion for discovery if it weren’t for his patient guidance, active support, and generous encouragement. He is an amazing scientist and has enthusiastically taught me in more ways than I could ever give him credit for. He will forever be a role model to me as I continue to strive for excellence in my career goals.   I am grateful to all the wonderful students and faculty members in the Life Sciences Institute and in the department of Microbiology and Immunology that I have had the immense pleasure of meeting and working with. My sincere appreciation goes to my committee members, Drs. Pauline Johnson, Ninan Abraham and Calvin Roskelley for always believing in me. I am so grateful for the ways they’ve challenged me to strive for excellence and integrity in my research. I thank the teaching faculty, particularly Drs. Tracy Kion, David Oliver and Marcia Graves, for giving me the opportunity to learn and teach in their classrooms.   xxviii  To my lab members May Dang-Lawson, Kate Choi, Caitlin Pritchard, Madison Bolger-Munro, Dr. Sonja Christian and Dr. Spencer Freeman, I thank them for inspiring me to be a better scientist and for nurturing my growth as a person. I will always remember the laughter we generated, the tears we shared and the western blots we puzzled over. I thank all the members of the Matsuuchi, Johnson, Abraham, Harder and Perona-Wright labs for their friendship, food adventures, stimulating scientific discussions, celebrations, and commiserations over failed experiments. Their kindness and positive attitude in every circumstance encouraged me to believe in myself and achieve more than I thought I could.   No one has been more important to me in the pursuit of my studies than my loved ones including my parents, family, partner and friends for their unending support through their acts of care, their zealous belief in my success and earnest prayers. My parents, who sacrificed so much and gave me all the opportunities to pursue my dreams, I am so grateful for their love and steadfast faith in me. I thank Duy for his encouragement, acts of love, and strength that sustained me to push through the toughest hurdles. And most importantly, for making all of this possible, I thank my father God who created me, and to whom I owe my very existence and everything I possess.     xxix  Dedication  I dedicate this thesis to my parents, Yan Sun and Samuel Xianen Wang, for loving me more than I’ll ever know, and unselfishly supporting me in all my crazy endeavors.   1 Chapter 1: Introduction 1.1 The role of B cell activation in immune responses B lymphocytes are an essential arm of the adaptive immune system that help defend the body against infection by producing antibodies and cytokines, and by acting as antigen-presenting cells (APCs). When naïve mature B cells bind their cognate antigen using their B cell receptor (BCR), the resulting intracellular signals initiate B cell activation, and the differentiation of B cells into effector cells and memory cells.  Antigen-activated B cells can differentiate into plasma cells that secrete antibodies that target pathogens and toxins by neutralizing the target (i.e. prevents attachment and entry into cells), marking it for opsonization by phagocytes, causing antibody-induced cellular toxicity, or lysing pathogens by activating the complement pathway [1]. Activated B cells also secrete cytokines that modulate the immune response [2] such as the effector cytokines IL-4, GM-CSF, TNFα, lymphotoxin alpha, and IL-6 [2], and the regulatory cytokines IL-10 and TGFβ [3]. B cells acquire BCR-bound antigens from APCs and present them to T cells in order to elicit T cell helper responses (e.g. expression of co-stimulatory molecules such as CD40 ligand, and secretion of IL-4, IL-5, and other B cell-activating cytokines) that drive B cell differentiation and Ig class switching [4]. This antigen-presenting function of B cells to T helper cells is especially important for affinity maturation in germinal centers (GC) where B cells undergo immunoglobulin (Ig) gene hypermutation and selection to produce high affinity antibodies [5]. Collectively, these B cell responses play a key role in protecting the body against pathogens. However, the aberrant activation of B cells and the dysregulation of B cell effector functions can initiate autoimmunity, which can lead to serious human diseases such as type 1 diabetes, multiple 2 sclerosis, and systemic lupus erythematosus (SLE) [6, 7]. Therefore, the precise control of B cell activation is essential for directing appropriate B cell responses.   1.1.1 BCR signaling pathways The first step in B cell activation is the recognition of cognate antigen via the BCR. The BCR consists of an antigen-binding subunit, the membrane Ig, and a signaling subunit. The membrane Ig portion of the BCR consists of two identical heavy chains as well as two identical light chains that are joined by disulfide bonds to form a Y-shaped structure (Figure 1.1). The variable regions of the Ig heavy and light chains that result from random gene recombination combine to form antigen-binding sites, one on each arm of the Y-structure. The signaling subunit consists of the Igα (CD79a) and Igβ (CD79b) polypeptides, which non-covalently associate with the membrane Ig. Upon antigen binding, BCRs aggregate into micron scale structures termed microclusters (discussed in section 1.3.2) that recruit signaling effectors to the BCR and activate BCR signaling pathways [8]. The clustering of BCRs by multivalent antigens or anti-Ig antibodies (also referred to as BCR crosslinking) triggers rapid phosphorylation of the immunoreceptor tyrosine activation motifs (ITAMs) in the cytoplasmic domains of Igα and Igβ. This is mediated by Src-family kinases (SFKs) such as Lyn, as well as the Syk tyrosine kinase [9]. The phosphorylation of ITAMs is the first step in initiating the BCR signaling cascade, which has been reviewed in [10, 11], but will be described briefly here (Figure 1.1). Antigen binding promotes the association of BCRs with the co-receptor CD19, which binds Lyn and brings it in close proximity to the BCR ITAMs [12]. The phosphatases CD45 and CD148 act as critical regulators of the initial Lyn-dependent ITAM phosphorylation. Both CD45 and CD148 activate Src-family kinases by 3 dephosphorylating the inhibitory C-terminal tyrosine residue in Lyn and other SFKs [13, 14] (Figure 1.1). Once this autoinhibition is relieved, Lyn phosphorylates the ITAMs in Igα and Igβ. In B cells, ITAM phosphorylation creates binding sites for Src homology-2 domain (SH2)-containing proteins including SFKs and the Syk tyrosine kinase, which then amplify BCR signaling.  The recruitment of Syk to the ITAMs allows Syk to be phosphorylated by SFKs. This increases Syk activity, allowing Syk to phosphorylate many downstream signaling proteins that are involved in BCR signaling [10, 11]. BCR-induced phosphorylation of the cytoplasmic domain of CD19 allows CD19 to recruit phosphoinositide 3-kinase (PI3K) to the plasma membrane [15]. PI3K phosphorylates the membrane lipid phosphatidylinositol 4,5-bisphosphate (PIP2) to generate the second messenger phosphatidylinositol 3,4,5-trisphosphase (PIP3), which recruits PH domain-containing proteins such as Bruton’s tyrosine kinase (Btk) and the Akt pro-survival kinase to the plasma membrane, where they are activated [15]. Akt plays a key role in regulating cellular metabolism and inhibits the FOXO transcription factors that promote apoptosis [16]. Syk and the SFKs phosphorylate Btk, which in turn leads to the activation of phospholipase C gamma 2 (PLCγ2). PLCγ2 cleaves PIP2 to generate the second messengers diacylglycerol (DAG) and 1,4,5-inositol trisphosphate (IP3). IP3 causes the release of calcium from intracellular stores, which activates the transcription factor Nuclear factor of activated T-cells (NFAT). DAG is important for the activation of protein kinase C (PKC) as well as the Ras, Rap, and Rac GTPases. Multiple PKC isoforms are activated by the BCR and these play key roles in regulating multiple substrates. For example, PKCβ is essential for BCR-induced activation of the pro-survival and pro-proliferation transcription factor NFκB [17]. DAG-dependent recruitment of RasGRP proteins to the plasma membrane lead to activation of Ras 4 [15]. Activated Ras controls a signaling cascade that activates extracellular-signal-regulated kinase (ERK), which, along with the Rac-Jun N-terminal kinase (JNK) pathway, phosphorylate key transcription factors including AP-1 (c-Jun/c-Fos heterodimer) [15]. As described later in section 1.5.4, the Rap and Rac GTPases regulate cytoskeletal organization. The Rap GTPases also link BCR signaling to integrin activation. These BCR signaling pathways cause changes in cytoskeletal organization, metabolic pathways, protein synthesis, and gene expression that lead to cell cycle progression, proliferation, and differentiation into antibody-producing cells (Figure 1.1) [15, 18, 19].  BCR signaling can also be negatively regulated by the B cell-specific Siglec family member CD22. CD22 contains three tandem cytosolic immunoreceptor inhibitory motifs (ITIMs) that are phosphorylated by activated Lyn upon BCR crosslinking [20]. Phosphorylated ITIMs recruit the SHP-1 and SHIP1 phosphatases via their SH2 domains. SHP-1 dephosphorylates the ITAM motifs in the BCR, a process facilitated by the interaction of CD22 with the BCR [21]. SHIP1 dephosphorylates PIP3 and opposes PI3K signaling [22]. This negative regulation of BCR signaling is particularly important for controlling B cell activation. Mice with loss-of-function mutations in the CD22/Lyn/SHP-1 signaling pathway develop lupus-like disease and are 8-fold more likely to develop autoimmunity than wild type mice [23]. Likewise, the phosphorylation of FcγRIIB ITIM motifs negatively regulates BCR signaling when immune complexes (ICs), which are antigens that are bound by antibodies, bring the FcγRIIB in close proximity to the BCR [22]. This is important for terminating BCR signaling when a sufficient antibody response has been mounted.   5   Figure 1.1. BCR signaling pathway  B cell receptors (BCRs) that bind cognate antigen initiate BCR signaling via the Igα/β signaling subunits of the BCR. CD45/CD148 phosphatases dephosphorylate the inhibitory tyrosines of Src family kinases (SFKs) to alleviate their autoinhibition. The immunoreceptor tyrosine-based activation motifs (ITAMs; depicted as red bars in the cytoplasmic domains of Igα/Igβ) are phosphorylated by SFKs such as Lyn, and can then recruit the Syk kinase. Syk initiates signal transduction pathways that lead to the activation of phospholipase C gamma 2 (PLC2), phosphoinositide 3-kinase (PI3K; comprised of the subunits p85 and p110), and the Ras, Rap, and Rac GTPases. Collectively, these signaling pathways result in cytoskeletal reorganization, integrin activation, and the activation of transcription factors including c-Jun, c-Fos, nuclear factor of activated T-cells (NFAT), and nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB), which regulate the expression of genes that are associated with B cell survival and activation. The activation of Akt also inhibits the pro-apoptosis transcription factor Forkhead box O (FOXO). Abbreviations: B cell linker protein (BLNK), Bruton’s tyrosine kinase (Btk), diacylglycerol (DAG), extracellular signal–regulated kinases (ERK), inositol trisphosphate (IP3), jun N-terminal kinases (JNK), protein kinase C (PKC). See text for additional details. Solid lines indicate direct interactions and dashed lines indicate indirect interactions. Adapted from [15, 24].  6 1.2 Antigen presentation to B cells Circulating naïve follicular B cells home to secondary lymphoid organs (SLOs) where they sample antigens that are concentrated in these organs and presented by APCs. APCs are important not only for initial B cell activation but also for the subsequent affinity maturation of B cells within GCs, a process that generates B cells that possess higher affinities for their cognate antigen. Here I discuss how antigens are presented to B cells and how this impacts B cell responses.   1.2.1 B cells migrate to secondary lymphoid organs to encounter antigen SLOs are specialized immune tissues that facilitate B cell activation as they are strategically positioned to sample antigens from many different sources. The spleen can sample blood-borne antigens, which are brought to this organ by DCs and macrophages [5]. The draining lymph nodes (LNs) filter antigens from the lymph fluid that drains from the skin (e.g. antigens that enter after a barrier breach) or mucosal surfaces [5, 25]. The Peyer’s patches are located strategically in the gut to sample antigens transported from the intestinal lumen [25].  B cells traffic via the blood and lymphatic systems and circulate through SLOs where they become activated if they encounter their cognate antigen. B cells can enter passively into the spleen but their entry into LNs occurs via specialized structures called high endothelial venules (HEV) (Figure 1.2) [26]. B cell entry into LNs via the HEVs involves binding of L-selectin on their surface to 6-sulpho sialyl Lewis X motifs on the surface of the HEV. This results in rolling adhesion along the endothelial wall [27]. HEVs also secrete the chemokine CCL21, which binds to the chemokine receptor CCR7 on B cells and induces the activation of the LFA-1 integrins [27]. Activated LFA-1 on B cells binds to ICAM-1 and ICAM-2 on the endothelial wall and 7 mediate firm adhesion so that B cells can switch to a crawling motion before transmigrating across the endothelium into the LN [27]. Once in the lymphoid tissue, a gradient of the chemokine CXCL13, which is produced by follicular dendritic cells (FDCs) and follicular reticular cells (FRCs), attracts B cells to migrate towards the area of high chemokine concentration. The chemokine receptors CXCR5, CCR7, and EBI2 (also known as GPR183) sense these chemokine gradients and direct B cells to migrate towards the B cell follicles [28, 29]. Intravital microscopy imaging has demonstrated that lymphocytes can also migrate towards the follicles by moving along a network of reticular fibers and along FDC dendritic processes [26]. Within the SLOs, B cells migrate in a CXCL13- and sphingosine-1-phosphate (S1P)-dependent manner while searching for antigens that have been concentrated in these tissues. Upon binding cognate antigen, B cells upregulate CCR7 so that they can then migrate to the boundary of the B cell and T cell zones to search for T cell help [28].    8  Figure 1.2. Lymph node structure  The lymph node is organized into compartments in which B cells enter via high endothelial venules (HEVs) that are located near the subcapsular sinus (SCS) in the cortical region. This cortical region is juxtaposed to the T cell zone in the paracortical region of the lymph node (LN) where T cells accumulate. Dendritic cells (DCs) enter the LN via the afferent lymphatics, SCS, or HEVs. Reticular fibers formed by fibroblastic reticular cells (FRCs) act as guidance networks for lymphocyte and DC migration to different compartments. FRCs and FDCs also produce the chemokines CCL21, CCL19 and CXCL13 that guide the intranodal migration and positioning of T and B cells. B cells are densely packed into discrete follicles with follicular dendritic cells (FDCs) that cluster in the center of these follicles. Circulating follicular B cells spend up to 24 hr exploring a LN and scanning for cognate antigen presented on FDCs in the follicles. B cells that have encountered cognate antigen within the follicle move towards the boundary of the B cell follicle and the T cell zone to interact with T helper cells [28]. Adapted from [27] with permission.  9 1.2.2 Antigen capture and presentation in the secondary lymphoid organs Collectively, the vast diversity of BCR specificities on the surfaces of the millions of B cells in the body can confer recognition of a large number of different antigens. However, the clonal nature of the adaptive immune system, which is established via VDJ recombination, ensures that all the BCRs on a single B cell have an identical antigen-binding site that binds only one or a few structurally-related antigens with sufficient binding affinity (Kd < 10-6 M) to induce BCR signaling [30]. To facilitate the encounter between a rare B cell and its cognate antigen, antigens from throughout the body are collected and concentrated within the SLOs, where they are captured and presented by several different types of APCs. Although B cells can readily recognize fluid phase antigens in vitro (e.g. stimulating B cells with soluble, multivalent antigens), B cell activation by antigens presented on the surface of APCs may be physiologically relevant in vivo [12, 31]. The concentration of antigens in the SLOs, and on the surfaces of APCs, may be especially important in naïve animals where only a few B cells may recognize that antigen. The entry of antigens into the B cell follicles proceeds by different mechanisms, depending on the nature of the antigen (e.g. its molecular size), the pre-existence of antibodies bound to the antigens, the original location of the antigen, and the type of SLO [28]. The blood and lymphatic systems deliver interstitial fluid (lymph), bacteria and viruses, ICs, and antigen-bearing DCs to SLOs (Figure 1.3) [32]. Antigens in the blood, which often become coated with complement proteins, are filtered and concentrated in the spleen where they are scavenged by DCs and macrophages, and then transferred to FDCs [5]. Antigens that do not enter the bloodstream are transported via the lymphatic system and lymph fluid to draining LNs, where their entry into the follicles occurs in a manner that depends on their size and whether they have 10 been opsonized. Soluble antigens of low molecular mass (<70 kDa) diffuse freely across the SCS and can be observed entering LNs within minutes of subcutaneous injection [26, 28, 33]. The SCS acts as a sheath that surrounds the LN and creates a sieve-like barrier around the B cell follicles. Hence, antigens larger than 70 kDa, ICs that include antibody- and complement-coated antigens, and particulate antigens (e.g. bacteria and viruses) must enter the B cell follicles by cell-mediated pathways [8]. Complement- and IgG-opsonized ICs are readily bound by SCS macrophages via complement receptor 3 (CR3; also called macrophage receptor 1 (MAC1)), FcγRIIB and CD169 (also called Siglec-1) [5, 28]. SCS macrophages can also bind non-opsonized antigens via Siglec-1 or the mannose receptor CD206 [5]. Antigens that SCS macrophages capture in the SCS can subsequently be displayed on their follicular surface and directly presented to B cells. Alternatively, the antigen can be transferred to non-cognate B cells in the follicle, which then shuttle the antigen from the SCS to deep within the follicles and transfer them to FDCs [5]. Antigens that are not opsonized or bound by SCS macrophages can be captured via pattern recognition receptors on DCs or macrophages that are located in the medullary interfollicular region [5]. These antigens can then be phagocytosed or transferred to FDCs for presentation to B cells. Antigens can also be transported to LNs [26] by DCs that home to LNs from the peripheral tissues [28]. Conventional DCs continuously recycle engulfed antigen to the cell surface without degrading the antigen. These DCs accumulate around the HEVs and can present intact antigen to B cells as they enter the LNs from the circulation [5]. Although DCs and SCS macrophages may be important for the early activation of B cells during primary immune responses, FDCs are thought to be the most important APC for the affinity maturation of B cells [5].   11  Figure 1.3. APCs in the LNs capture and present antigen to B cells  Antigens arriving through the afferent lymph vessels can enter the B cell follicles and be presented to B cells by several mechanisms. Small soluble antigens can directly traverse the subcapsular sinus (SCS) and enter into the B cell follicles. Larger antigens, immune complexes (ICs), and particulate antigens can be transported by SCS macrophages that bind antigens using a variety of different receptors (e.g. DC-SIGN, MAC1, complement receptors (CRs), Fc receptors (FcR)). Antigen is then transported across the SCS by the SCS macrophages where they are directly presented to B cells. Alternatively, antigens can be transferred to non-cognate B cells, which shuttle the antigen to follicular dendritic cells (FDCs). FDCs deep within the follicles can then present the antigen to cognate B cells. See text for details. The figure is from [8] and is used with permission.  1.2.2.1 Follicular dendritic cells FDCs are specialized APCs that efficiently capture antigens, secrete chemokines that attract B cells, and present antigens to B cells in order to stimulate B cell activation. In addition, FDCs can store antigens for long periods of time and, therefore, represent a record of previous antigen exposure. Because FDCs are located throughout the B cell follicle, and are not in physical contact with SCS macrophages, transport mechanisms are required to transfer antigens to the FDCs. Non-cognate B cells can acquire antigens from SCS macrophages in a CR1/2 dependent manner and then shuttle this antigen to FDCs within the follicle [28]. The higher 12 levels of CR1/2 on FDCs compared to follicular B cells may be important for FDCs to acquire this antigen from the non-cognate B cell. The FDC can then present this antigen to B cells and activate cognate B cells [5]. Medullary DCs and macrophages can also transfer antigen to FDCs [5]. FDCs may also take up small soluble antigens that cross the SCS and enter into the follicles.  In addition to complement receptors, FDCs express Fc receptors, DC-SIGN, Siglec-1/CD169, and other lectins that can capture and present antigens to B cells [5, 26, 28, 34-37]. The most important of these receptors for B cell activation are likely the CR1 and CR2 (CD21) receptors that present C3-opsonized antigens to cognate B cells [5]. Many of these receptors retain bound antigens in an intact form at the cell surface whereas others cycle them through non-degradative intracellular compartments and then traffic the antigen back to the cell surface [26, 38]. ICs are often seen bound to the reticular processes of FDCs, which are called IC-coated bodies or iccosomes due to their beaded structure, which are 0.25–0.38 μm in diameter [26]. Opsonized antigens can be displayed on FDCs within hours of immunization and can remain on the surface of APCs for days, weeks, or sometimes months [5, 26]. The antigen capture function of FDCs, as well as the prolonged presentation of captured antigens on their surfaces, may be particularly important for activating the limited number of B cells that can recognize a specific antigen. Intravital two-photon microscopy has shown that B cells rapidly scan the surface of APCs and interact only briefly if they do not encounter their cognate antigen. However, when B cells interact with APCs bearing cognate antigen, they typically interact with that APC for 20-30 min, although these interactions can last much longer [39]. The length of the interaction between a B cell and an APC may depend on the affinity and avidity of the BCR for that antigen as well as the number of antigen molecules on the APC surface. Although FDCs can activate naive B 13 cells, they are essential for the affinity maturation of B cells in GCs, which results in high affinity antibodies (discussed in section 1.4.3).  1.3 The B cell immune synapse B cells that contact APCs bearing cognate antigen initiate BCR signaling that lead to dynamic morphological changes that form an immune synapse (IS) at the B cell:APC interface [30]. BCRs that bind cognate antigen rapidly aggregate to form microclusters that, along with other adhesion proteins and the actin cytoskeleton, are progressively organized into an IS. The formation of the IS promotes B cell activation by enhancing BCR signaling and antigen acquisition; the latter function allows B cells to present antigen to T cells and elicit T cell help [28, 31]. Importantly the morphological changes that underlie IS formation require remodeling of the actin and microtubule cytoskeletons. In this section I will discuss BCR organization as well as the cytoskeletal regulators that establish B cell polarity and drive IS formation at the B cell:APC contact site.  1.3.1 BCR organization Up to hundreds of thousands of BCRs can be expressed on the surfaces of B cells [40]. Understanding how this multitude of BCRs are organized within the plasma membrane and how this organization impacts BCR signaling is critical for determining how B cell activation is controlled. Initial models of B cell activation proposed that BCRs exist as monomers on the cell surface and that BCR signaling is initiated by antigen-induced BCR clustering (also called crosslinking). This was consistent with observations that BCR signaling and B cell activation could be stimulated by multivalent antigens and intact anti-Ig antibodies but not by monovalent 14 antigens or monovalent F(ab) fragments of anti-Ig antibodies [41]. However, native gel electrophoresis suggested that BCRs form oligomers even in resting cells [42]. More recently, direct stochastic optical reconstruction microscopy (dSTORM) super-resolution imaging by Mattila et al. showed that BCRs in naïve mature B cells exist as separate IgM and IgD nanoclusters of 120-160 nm in diameter with ~20-120 BCRs in each nanocluster [40]. These nanoclusters in resting B cells are thought to reside in protein islands, regions of high protein density within the plasma membrane where the lipid composition limits the diffusion of proteins [43]. This pre-clustering of BCRs is thought to promote tonic BCR signaling [44], which is essential for B cell survival [45]. Upon antigen engagement, BCR nanoclusters aggregate even further into larger proteins complexes (i.e. BCR microclusters).  Early models of membrane protein organization suggested that the fluid-like movement of lipids and proteins within the plasmid membrane reflected planar Brownian diffusion [46]. However, Kusumi et al. showed that even abundant phospholipids in cellular membranes moved slower than the diffusion coefficient predicted for Brownian motion [47]. The observation that these membrane components moved within confined compartments, and occasionally “hopped” to another compartment, led to the model of a compartmentalized plasma membrane that limits the free diffusion of lipids and proteins embedded within it [47-49]. It is now generally accepted that most, if not all, plasma membrane-associated proteins are organized into submicron diameter compartments or “protein islands”.  Likewise, the BCRs on the surface of B cells are not randomly distributed, but are rather compartmentalized within the plasma membrane along with other membrane proteins [40, 50]. Membrane compartments that organize BCRs within the plasma membrane can form via several mechanisms. Lipids such as sphingolipids and sterols can self-organize into lipid “rafts” or microdomains. Proteins that partition into these lipid 15 microdomains exhibit slower diffusion than in non-raft membrane compartments [51]. The underlying membrane-associated actin cytoskeleton may also form membrane compartments by organizing networks of membrane-associated proteins that form pickets within the plasma membrane [47, 49, 52]. Within these compartments, BCRs and other membrane proteins on the B cell surface may exist in dense “hot spots” termed protein islands. Protein-protein interactions may promote the formation of BCR nanoclusters. Reth and colleagues observed that IgD BCR complexes form in a manner that is dependent on amino acids in the transmembrane domain of this class of BCRs [42]. Taken together, it is likely that a combination of different mechanisms contributes to the non-random distribution and organization of BCRs at the B cell surface.  The BCR nanoclusters in resting B cells are kept relatively immobile by the membrane-associated actin cytoskeleton [44, 53]. Ezrin plays a key role in creating plasma membrane compartments by anchoring the MSK to transmembrane proteins (e.g. CD44, ICAMs, selectins, CD43, CD95) that act as pickets [54]. The extracellular domains of these pickets may also impede the movement of the BCR within the membrane. Antigen-induced BCR signaling results in localized cofilin-mediated severing of the MSK [55, 56] as well as the inactivation of ezrin, which uncouples the actin network from the plasma membrane [53]. This increases lateral BCR mobility, which promotes BCR:BCR collisions as well as BCR collisions with other membrane proteins such as the activatory coreceptor CD19 [50]. This results in the formation of microclusters of BCRs with CD19, which leads to the formation of BCR microsignalosomes that mediate BCR signaling [12, 40]. This is consistent with the “collision-coupling” model [57]. Indeed, treating B cells with actin-disrupting drugs is sufficient to induce the coalescence of BCR nanoclusters [44], the interaction of BCRs with CD19 [40], and robust antigen-independent BCR signaling [50]. This suggests that BCR nanoclusters are confined within actin-based 16 compartments in resting B cells and that this limits BCR signaling in the absence of antigen by preventing the interaction of BCRs with the activating coreceptor CD19 [50, 58]. Therefore, not only do BCRs exhibit non-random distribution on the B cell surface but their organization may serve functional purposes in promoting appropriate BCR signaling in the presence of antigen.   1.3.2 B cells that bind to cognate antigen form signaling BCR microclusters The formation of BCR microclusters is one of the earliest observable events in the activation of B cells by membrane-associated antigens. Within 1-2 min of contacting antigen on planar lipid bilayers, BCR nanoclusters aggregate into micron-scale structures that consist of 100-500 BCRs [12, 50]. How antigen binding promotes the transition of BCRs from nanoclusters to microclusters is not fully understood. Antigen-induced BCR signaling initiates remodeling of the MSK, which releases BCR nanoclusters from actin-based restraints and allows for the increased diffusion of nanoclusters within the plasma membrane. Upon antigen binding by BCRs, ezrin is transiently dephosphorylated, which abrogates its ability to link the MSK to the plasma membrane [53]. At the same time, the localized activation of the actin-severing protein cofilin (see Section 1.5.5) is thought to remove actin-based diffusion barriers [55, 56]. This results in increased mobility of BCRs within the plasma membrane, which allows IgM and IgD BCR nanoclusters from adjacent compartments to associate and progressively concatenate into microclusters that contain both IgM and IgD BCRs [41, 50, 53]. Ezrin proteins are then re-phosphorylated and recruited to microclusters where they mediate the formation of actin corrals to preserve the stability of antigen-bound BCR microclusters [53].  BCR microclusters recruit signaling proteins to form a microsignalosome, the basic unit of BCR signaling [9] (see 1.1.1 above). CD19, an activating co-receptor for BCR signaling, 17 rapidly associates with BCR microclusters upon actin cytoskeleton remodeling [12]. CD19 acts as a co-receptor that enhances BCR signaling by recruiting signaling effectors such as PI3K and Vav to BCR microsignalosomes [12]. It is thought that during BCR microcluster formation, remodeling of the actin cytoskeleton promotes the association between BCR nanoclusters with pre-existing CD19 nanoclusters [50, 59]. The membrane organization of CD19 is governed by the tetraspanin CD81, which forms microdomains and amplifies BCR signaling by regulating the association of CD19 with BCR nanoclusters [40]. The phosphatase CD45 also associates initially with BCR microclusters to shield them from CD22 nanoclusters, which negatively regulate BCR signaling by recruiting the tyrosine phosphatases SHP-1 and SHIP1 [60]. After a few minutes of engaging antigen, antigen-bound BCR microclusters coalesce into a large central supramolecular activation cluster (cSMAC) that is characteristic of a lymphocyte IS.   1.3.3 B cells form an immune synapse with antigen-presenting cells The formation of the IS was first described for T cells [61-63], but also occurs during the activation of natural killer (NK) cells, phagocytes, and B cells [15]. In B cells, IS formation upon binding cognate antigen occurs in three steps: (1) BCR microcluster formation, (2) B cell spreading across the surface of the APC to increase the chance of engaging additional cognate antigens, and (3) B cell contraction to gather antigen-bound BCR microclusters into a large central cluster along with the higher ordered organization of other receptors and cytoskeletal structures (Figure 1.4) [8, 9, 15, 32, 64, 65].  To exceed the signaling threshold required for B cell activation, B cells must maximize the extent of antigen encounter and BCR microcluster formation. To accomplish this, B cells extend their membrane across the surface of the APC to increase the area of contact with the 18 APC [64]. This step may be particularly important for B cell activation by rare antigens that are presented at low densities as even membranes presenting low amounts of cognate antigen can support IS formation and B cell activation [66]. B cell spreading is initiated by BCR signaling where the affinity and avidity of antigen-binding determines the extent of B cell spreading, the number of BCR microclusters formed, and the magnitude of BCR signaling [64]. The Lyn and Syk tyrosine kinases are critical for initiating B cell spreading, as are PLCγ, BLNK, Vav, and Btk [9, 67]. The spreading of B cells across the APC surface is driven by dynamic reorganization of the actin cytoskeleton and is a prerequisite for IS formation [32]. The regulation and function of actin cytoskeleton remodeling during IS formation will be discussed in section 1.5. Formation of the IS organizes BCR microclusters, integrins and the actin cytoskeleton into discrete ring structures, which collectively form a bulls-eye pattern. The center of the IS consists of a cSMAC that contains a large central cluster of antigen-bound BCR microclusters [8, 31, 32, 66]. This is surrounded by a peripheral SMAC (pSMAC) in which ligand-bound LFA-1 and VLA-4 integrins are enriched [32]. Surrounding and underlying the pSMAC is the distal SMAC (dSMAC), which is comprised of a ring of polymerized filamentous actin (F-actin). After ~2-5 min of spreading across the APC surface, the B cell undergoes a contraction phase in which antigen-bound BCR microclusters are gathered into larger clusters that are brought together to form the cSMAC [32, 66]. During IS formation, membrane proteins with large extracellular domains, such as the phosphatases CD45 and CD148, are moved to the periphery of the IS, beyond the pSMAC. This size-based protein sorting at the IS is referred to as kinetic segregation [32, 63, 66]. This mechanism of segregating bulky phosphatases may also exclude negative regulators of BCR (e.g. CD22) signaling from the IS [59]. 19 The LFA-1 and VLA-4 integrins in the pSMAC are adhesion molecules that bind ICAM-1 and VCAM-1, respectively, which are expressed on the surface of APCs [68-71]. Integrins mediate B cell adhesion to the APC and lower the threshold for B cell activation [69, 72]. In experiments using lipid bilayers containing defined amounts of antigen and the LFA-1 ligand ICAM-1, Batista and colleagues showed that integrin-mediated adhesion reduces the amount of antigen required for B cells to adhere to these lipid bilayers and become activated [72]. Even under limiting antigen conditions, integrins are recruited to the antigen contact site where they facilitate IS formation as well as B cell activation, as assessed by CD86 upregulation [72]. In T cells, single particle tracking of TCRs revealed that integrin-mediated adhesion slows down the centripetal movement of TCR microclusters towards the cSMAC and, in this way, prolongs TCR microcluster signaling [73]. The ability of integrins to bind ligands is not constitutive and is instead induced by intracellular signaling pathways in a process known as “inside-out signaling” [31, 69]. BCR-induced integrin activation requires SFKs, Syk, Vav, PI3K, Btk, PLCγ2, Ca2+ release, PKC, the Rac2 GTPase, and the Rap1 GTPases [31, 70, 74-77]. The resulting recruitment of RIAM and talin to the cytoplasmic domains of the integrin  and β chains cause a conformational change in the integrin extracellular domains that increases its affinity for adhesion molecules [70, 74, 75, 77, 78].  Cytoskeletal reorganization is critical for IS formation. The initial actin structures that form at the lymphocyte:APC interface mimic those found at the leading edge of migrating cells. Dustin and colleagues have termed the IS a radially-symmetric version of the leading edge [79]. The dSMAC surrounding and underlying the pSMAC occurs at the periphery of the B cell:APC interface in the outer lamellipodium and is enriched in dendritic actin [80-83]. Inside the dSMAC ring lies the lamellum which is comprised of linear actin fibers that lie parallel to the edge of the 20 cell and form arcs that generate myosin II-dependent contraction forces [81]. During the contraction phase of IS formation, both actin- and microtubule-dependent mechanisms mediate BCR microcluster coalescence. Upon binding antigen presented on an APC, the B cell polarizes its microtubule-organizing center (MTOC) and the microtubule network towards the site of antigen contact. These cytoskeletal remodeling events and polarity-establishing mechanisms that drive BCR microcluster coalescence and IS formation are discussed below in sections 1.5 and 1.6.   21  Figure 1.4. Actin-dependent BCR mobility and IS formation  BCRs that bind antigen presented by APCs initiate BCR signaling, which induces remodeling of the membrane associated cytoskeleton (MSK). The disruption of ezrin-mediated connections between the plasma membrane and the actin cytoskeleton, as well as localized actin severing mediated by cofilin, increases BCR mobility within the plasma membrane. BCRs that bind antigen can then aggregate into microclusters, and recruit signaling proteins to form microsignalosomes. Integrins bind to Intercellular Adhesion Molecule 1 (ICAM-1) on the surface of the APC, which facilitate cell-cell adhesion and lower the threshold for B cell activation. The B cell then spreads over the surface of the APC to increase the chance of encountering cognate antigen. The B cell subsequently contracts and organizes receptors into concentric rings called supramolecular activation clusters (SMACs) surrounding a central cluster of antigen-bound BCR microclusters termed the central supramolecular activation cluster (cSMAC). A ring of integrins, which is termed the peripheral SMAC (pSMAC), surrounds the cSMAC. The outermost ring, or distal SMAC (dSMAC), is comprised of F-actin. This bulls-eye pattern of protein organization is characteristic of immune synapses formed by B cells, T cells, and natural killer cells. See text for explanation. Adapted from [15, 84] with permission.  22 1.4 The antigen acquisition function of the B cell immune synapse  In vivo, B cells form contacts with APCs that last 20-30 min in order to sustain BCR signaling and promote B cell activation [39]. The formation of an IS with antigen-bearing APCs is important for facilitating initial BCR signaling as well as for supporting later B cell interactions with T cells. At the IS, antigens are concentrated in a central site and this optimizes the acquisition of BCR-bound antigens and their subsequent delivery to antigen processing and MHC II loading compartments [9, 29, 32, 65, 85-87]. B cells that acquire antigens can then migrate to the B cell-T cell border to present antigens to T helper cells. The amount of antigen acquired determines the capacity of a B cell to present antigens to T cells and elicit T cell help [30, 64, 66, 88]. T cells provide B cells with co-stimulatory signals that include membrane-bound ligands such as CD40L and ICOS and cytokines such as IL-4, IFNγ and IL-21 [89]. T cell help is required for the GC response in which B cells undergo affinity maturation and selection for high affinity antibodies, as well as differentiation into plasma cells and long-lived memory B cells. Therefore, the ability of B cells to effectively acquire and present antigens is essential for generating a robust antibody response.  1.4.1 Immune synapse formation regulates BCR signaling by promoting BCR internalization The IS in T cells is thought to function as a “molecular machine” for sustaining signaling for several hours to direct T cell differentiation [62, 90]. In contrast, B cells form transient contacts with APCs lasting minutes to half an hour [32, 64]. Initially it was thought that the IS served as a central signaling platform. However, the observation that calcium signaling begins before the IS forms has led to the idea that the cSMAC acts instead as a site for the termination of signaling via the internalization of antigen-bound TCRs or BCRs [37, 91, 92]. In naïve mature 23 B cells, cSMAC formation is followed by the acquisition of antigen from APC membranes [87] (Figure 1.5). Antigen-bound BCRs are ubiquitinated and rapidly internalized for degradation whereas BCRs that are not ubiquitinated remain at the cell surface and continue signaling until the IS is formed [37]. Many diffuse large B cell lymphomas have elevated BCR expression at the cell surface due to defects in BCR internalization arising from a mutation in Igβ. This indicates that impaired BCR internalization can lead to chronic BCR signaling that supports the survival of malignant cells [93]. Similarly, beige mice, which harbor a mutation similar to that in Chediak-Higashi syndrome in humans, have a defect in lysosome function that causes internalized BCRs to accumulate in late endosomes, instead of being trafficked to lysosomes where they are degraded. These B cells exhibit increased long-term BCR signaling via ERK and p38, which correlates with excessive antibody production in vivo [94]. Indeed, defects in several endocytic machinery components or regulators of vesicular trafficking result in aberrantly enhanced BCR signaling and autoimmunity [37]. Therefore, antigen internalization at the IS is an important mechanism for preventing excessive BCR signaling and B cell effector functions in vivo. In contrast to the role of BCR internalization in terminating BCR signaling, internalized BCRs have been shown to recruit signaling effectors to intracellular vesicles and initiate additional BCR signaling that promotes antigen trafficking and B cell activation [86]. Pierce and colleagues showed that the active, phosphorylated forms of ERK and Akt co-localize with early endosomes containing BCR-antigen complexes prior to the fusion of these vesicles with degradative lysosomal compartments [86]. However, these findings remain to be confirmed by other groups. It is not known whether vesicles containing BCRs bound to antigens that are acquired from APC membranes also support BCR signaling [39]. Altogether, these reports 24 highlight the importance of BCR-mediated antigen internalization at the IS in the regulation of BCR signaling.   Figure 1.5. Naïve B cells acquire antigen from APCs after forming an IS   IS formation is important for the acquisition of antigen from APCs. The reorientation of the MTOC (orange circle) and the microtubule network (orange), along with peripheral F-actin dynamics (not shown), facilitate the gathering of antigen from the periphery of the B cell:APC contact site. The formation of BCR microclusters (blue circles) triggers BCR signaling, which is important for recruiting the machinery required for extracting the antigen from the APC membrane and then internalizing it. Importantly, naïve mature B cells must gather antigen into a central supramolecular activation cluster (cSMAC) in order to extract the antigen from the APC membrane whereas antigen extraction by GC B cells (not shown) is mediated by BCR microclusters that form at the periphery of the cell:cell interface [87]. Extracted antigen is internalized into intracellular vesicles and trafficked to antigen processing and major histocompatibility complex II (MHC II)-loading compartments. See text for details.   1.4.2 B cells acquire antigens from APCs and present them to T cells The ability of B cells to take up antigens and then present them to T cells was first described by Lanzavecchia in 1985 [95]. Although B cells can use fluid phase pinocytosis to sample the external environment and take up antigens, receptor-mediated internalization of antigens via the BCR has been estimated to be 1000 times more efficient [96, 97]. It is now appreciated that B cells can acquire antigens from APC membranes using its BCRs, internalize 25 those antigens, and then present them to T cells (Figure 1.6). Batista et al. showed that when transgenic B cells expressing hen egg lysozyme (HEL)-specific BCRs were allowed to interact with APCs expressing GFP-tagged HEL on their surface, GFP fluorescence co-localized with the BCR inside the B cell [66]. Moreover, incubating these B cells with antigen-specific T cell hybridomas resulted in T cell activation, as judged by the production of IL-2 [66]. These results provided the first evidence that B cells can acquire antigens from APCs, present them in the form of peptide-MHC II complexes to T cells, and stimulate T cell activation. More recent work by Tolar and colleagues using high throughput imaging approaches showed that naïve follicular mature B cells gathered antigen-bound BCRs into a cSMAC before internalizing the antigen [87]. In contrast, activated B cells undergoing affinity maturation in the GC do not always gather antigen into a cSMAC and BCR-bound antigen is internalized primarily at peripheral BCR microclusters [87]. Nevertheless, at least for naïve follicular B cells, IS formation is essential for one of the primary functions of the BCR, the internalization of antigen and its subsequent delivery to antigen processing and MHC II loading compartments. [15].    26  Figure 1.6. B cells acquire antigens to gain T cell help   The binding of BCRs to cognate antigens on the surfaces of APCs allows the B cell to extract antigen from the APC surface and then internalize BCR-antigen complexes. The MTOC and the microtubule network is reoriented towards the site of IS formation. Lysosomes and major histocompatibility complex (MHC II)-containing vesicles, which are transported along microtubules, are moved towards the reoriented MTOC and close to the IS. Antigen that is internalized is processed into peptides that are loaded onto MHC II molecules, which are then trafficked to the B cell surface. Activated B cells migrate to the border of the B cell:T cell zones to present antigen to CD4+ T helper cells, which provide co-stimulatory signals that promote B cell differentiation, proliferation, and Ig class switching. B cells that have undergone somatic hypermutation also present internalized antigens to GC T cells in order to receive survival signals. This is an important mechanism for selecting for mutated BCRs that are highly effective at binding and internalizing antigen. Adapted from [15, 29] with permission.   1.4.3 B cells present antigens to gain T cell help One of the earliest discoveries with regard to T cell function is providing help to B cells, hence giving rise to the term “T helper” cells. Early reports by Miller and Mitchell discovered that only the adoptive transfer of both T and B cells into an irradiated mouse could elicit an 27 antibody response when the mice were subsequently immunized with sheep erythrocytes [98-100]. These experiments showed that T cells, which do not produce antibodies, are required for the B cell-dependent production of antibodies in response to certain types of antigens. Importantly, it was suggested that B cells and T cells recognized the same antigen. We now describe the antigens that are bound by BCRs and lead to B cell activation as T-dependent (TD) or T-independent (TI). TD antigens are protein antigens are recognized by both B cells and T cells (in the context of MHC II) and require cooperation between the two lymphocyte types [8]. On the other hand, TI antigens can induce B cell activation in the absence of T cells and include TLR ligands and mitogens that activate B cells, as well as polysaccharides with multiple repeating epitopes that engage and crosslink multiple BCRs [101]. TI antigens induce the rapid differentiation of B cells into short-lived plasmablasts that provide immediate protection against infection; however, these cells usually produce antibodies with low affinity. Some TI antigens can bind to both the BCR and to pattern recognition receptors, which can synergize with BCR signaling to enhance antibody production and even induce class switch recombination [102, 103]. However, naïve B cells often express BCRs that bind to antigen with low affinity and, therefore, B cell activation by TI antigens result in the production of low affinity antibodies [8]. In most cases, B cells require multiple rounds of affinity maturation in the GCs to produce high affinity antibodies, a process that requires interactions with T helper cells and, therefore, TD antigens.  Upon initial antigen-dependent interactions with APCs in the B cell follicles, B cells activated by TD antigens migrate to the B cell:T cell zone border where they interact with activated T cells (Figure 1.7). T cells providing help to B cells induce the formation of a GC, which is characterized by cyclic events that include BCR somatic hypermutation, proliferation 28 and clonal selection to ensure affinity maturation [89]. Activated B cells can in turn trigger activated CD4+ T cells to differentiate into T follicular helper (TFH), in part via ICOSL-ICOS interactions [23]. Within the GCs, B cells undergo multiple rounds of cell division and somatic hypermutation to introduce mutations in the Ig heavy chain variable regions in order to create a clone that can potentially bind antigen with higher affinity [5, 89]. These B cell clones then present antigen and compete for TFH cell-derived help [89]. This process requires FDCs to continuously present antigen to B cells. B cell clones compete for small amounts of cognate antigen on the surface of FDC and only those with the highest affinity BCRs for the antigen receive antigen-induced survival signals [5].  The amount of antigen that a B cell acquires determines its capacity to compete for TFH help and receive additional signals for survival, proliferation, differentiation, class-switch recombination and somatic hypermutation [104]. GC B cells that are unable to compete for TFH help are deleted in the GC as they are programmed to undergo apoptosis if they do not receive T cell-derived survival signals [104]. Affinity matured B cells can become either plasma cells that secrete high affinity antibodies or memory cells that enter the circulation and have the potential to survive for long periods of time within lymphoid organs such as the bone marrow. Alternatively, they can re-enter the GC to undergo additional rounds of affinity maturation [5]. Therefore, at multiple steps during the humoral response, the ability of B cells to extract and internalize antigens from APCs, and then present those antigens to T cells, is a critical determinant of B cell fate.    29   Figure 1.7. T cell-dependent B cell activation   B cells that encounter cognate antigen within the B cell follicle migrate to the T cell:B cell boundary [26, 105] where they form mobile long-lived contacts with T cells. This interaction allows B cells to produce short-lived extrafollicular plasmablasts [89]. These contact-dependent responses require several receptor-ligand interactions including major histocompatibility complex II (MHC II):T cell receptor (TCR) and CD40:CD40L. Some activated T and B cells move back into the follicle and continue short-lived interactions that require ICOS:ICOSL interactions to support B cell expansion and differentiation until germinal centres (GCs) develop from the clustering of B cells around follicular dendritic cells. It is during the GC response that B cell affinity maturation occurs. Adapted from [89] with permission.   1.4.4 B cells extract antigens from APC membranes B cells extract antigens from the membranes of APCs and may employ multiple mechanisms to do so. When BCRs bind to antigens presented by capture receptors on APCs, pulling forces exerted by the BCRs can extract the antigen from the APC. This type of antigen extraction is mediated by actin-associated myosin IIA, which generates forces on the BCR [106, 107]. Myosin-mediated pulling forces can either extract the antigen from the APC or, if the BCR-binding affinity is too low, the pulling forces would detach the BCR from the antigen [107]. This allows B cells to distinguish weak versus strong affinity antigen such that only B 30 cells that bind with high affinity for the antigen would be able to acquire that antigen, present it to T cells, and receive T cell-derived signals [107]. In addition to dissociating the antigen from the capture receptor on the APC, B cells can exert sufficient force to tear off portions of the APC membrane bearing the cognate antigen. This process is termed trogocytosis and the membrane patch that is ripped off is then internalized along with the BCR-bound antigen [108]. Finally, B cells can also secrete lysosomal enzymes into the synaptic space between the B cell and the APC to proteolytically cleave antigens from their capture receptors [15]. This mechanism of antigen extraction is thought to be important for the extraction of high affinity-binding antigens that are particularly difficult to extract using mechanical forces alone [109].  The antigen extraction mechanisms used by BCRs is also determined by the stiffness of the antigen-presenting membrane or substrate. Using plasma membrane sheets derived from cells as well as APCs with various rigidities, Tolar and colleagues demonstrated that the tensile strength of the antigen-presenting membrane determines the type of antigen extraction mechanism that the B cell employs [109]. When BCRs bind antigen presented on the surface of softer membranes such as those found in DCs, mechanical forces generated by actomyosin contraction are the main mechanism for antigen extraction [109]. However, when presented with antigens on more rigid surfaces, including stiffer biological membranes such as those of FDCs, mechanical forces may not be sufficient to extract antigens. In this case, B cells resort to lysosome exocytosis and the proteolytic cleavage of antigens from the APC surface [15, 85, 109].  The accumulation of lysosomes close to the IS, as well as their directed exocytosis towards the APC, requires the reorientation of the MTOC towards the IS [15]. The B cell:APC interaction acts as a polarity cue and the repositioning of the MTOC facilitates the localization of 31 endosomes, lysosomes, and MHC II-loading compartments close to the contact site as these organelles are moved along microtubules towards the MTOC [110]. BCR-bound antigens that are acquired from APCs are trafficked to these membrane compartments that collectively form the multivesicular bodies, which are the sites of antigen processing and loading onto MHC II molecules [97]. Internalized antigen is subsequently presented at the cell surface as peptide-MHC II complexes that can activate CD4+ T cells, which deliver the co-stimulatory signals required for B cell activation [29, 97].   1.4.5 BCR-mediated antigen endocytosis and trafficking BCRs continually undergo endocytosis and recycling in the absence of antigen. However, BCR crosslinking by multivalent antigens initiates BCR signaling that increases the rate of BCR endocytosis thereby increasing antigen internalization and processing [111, 112]. The uptake of both soluble and membrane-bound antigens typically requires clathrin-mediated endocytosis (CME) [37, 107, 112-115]. Clathrin molecules aggregate and coat the plasma membrane in structures called clathrin-coated pits (CCP), which facilitate the invagination of the plasma membrane around the BCR-bound antigen and the eventual pinching off of these invaginations to form vesicles [37]. CCPs associate with the BCR Igα/β subunit in the BCR via the adaptor protein AP-2. AP-2 recruits the SFK Lyn, which phosphorylates the clathrin heavy chain [37]. Phosphorylated clathrin heavy chains associate with clathrin light chain to couple CCP formation to actin polymerization, which is a requirement for membrane invagination [37]. Depleting Lyn abolishes all BCR-mediated antigen internalization in the DT40 chicken B cell line [116]. BCR-mediated endocytosis also requires signaling via Syk, Btk, Vav, and Bam32, which likely regulate actin polymerization at CCPs. Wiskott-Aldrich syndrome protein (WASP) family 32 proteins and Arp2/3 are recruited via these signaling proteins to nucleate actin polymerization [37]. Additionally, actin-binding protein 1 (Abp1) links actin polymerization to vesicle neck constriction [112]. Although actin-mediated antigen internalization can occur independently of clathrin-mediated uptake, these two molecular mechanisms often work in concert during BCR-mediated antigen endocytosis [114].  Multiple signaling components regulate the trafficking of internalized antigen into processing compartments. The CCP is stripped of its clathrin coat and converted into an early endosome. Conversion into an early endosome can result in either recycling back to the plasma membrane, which is common for BCRs under steady-state conditions, or maturing into a late endosome and fusing with a lysosome for degradation of the vesicle contents [37]. The trafficking of BCRs from early endosomes to late endosomes requires ubiquitination of BCRs at the plasma membrane. The Igβs of crosslinked BCRs are ubiquitinated by Cbl ubiquitin ligases, which promotes vesicle sorting to late endosomes [117, 118]. Late endosomes can then fuse with MHC II-containing multivesicular bodies and lysosomes, resulting in the proteolytic processing of antigens into peptides [37, 119]. Antigen peptides are loaded onto MHC class II molecules and then trafficked to the plasma membrane for presentation to T cells [120].   1.5 The actin cytoskeleton  B cell activation and IS formation is a highly dynamic process that requires substantial reorganization of many cellular components. Actin cytoskeleton remodeling is a key driving force that promotes IS formation as well as the functions of the IS that lead to B cell activation. Not only does the submembrane actin cytoskeleton mediate receptor organization and tonic BCR signaling at steady state, but the remodeling of the MSK upon antigen binding is also an 33 important mechanism that controls BCR mobility and signaling. Additionally, actin-based forces promote B cell spreading as well as the subsequent membrane contraction to form the cSMAC where antigens are internalized. Actin polymerization also drives antigen acquisition and facilitates the trafficking of internalized antigen to processing compartments. Therefore, the actin cytoskeleton is a critical regulator of B cell signaling and activation. A large number of proteins regulate the formation, dynamics, and remodeling of distinct actin structures within cells. In this section, I focus on those that play key roles in lymphocyte activation.   1.5.1 Actin structures  The actin cytoskeleton is comprised of a meshwork of filamentous actin fibers that support essential cellular functions. These include determining cell shape, supporting locomotion, establishing cell polarity, controlling intracellular trafficking, providing structural scaffolds, supporting cell division, and mediating other dynamic cellular processes [121-123]. It has been estimated that 1-5% of the total cellular protein is comprised of actin, a 42 kDa protein [124]. The cytosolic concentration of actin is ~0.5 mM, making it one of the most abundant proteins in the cell [122]. The high concentration of actin in the cell reflects the presence of actin structures that occupy large spaces in the cell and which form the actin cortex, a dense meshwork of F-actin arrays that underlies the cell membrane to form the “skeleton” of the cell [50]. The cell cortex defines the cell shape and opposes mechanical stress [50].   F-actin is assembled via the polymerization of globular actin monomers (G-actin) and these two states exist in a dynamic equilibrium [125]. An actin monomer is comprised of a single polypeptide chain with four subdomains (SD1-4), which is folded into a U-shaped structure [124]. The formation of new actin filaments requires the initial self-assembly of actin monomers 34 into a trimer. This is a rate-limiting nucleation step that is required to form a seed for further actin polymerization [126, 127]. Polymerization of G-actin into a polymer occurs with the assembly of each monomer in the same orientation and is associated with a conformational change that results from hydrolyzing bound ATP to ADP [124, 125]. Actin filament growth and shrinkage result from the addition and loss of G-actin monomers from the filament ends [124]. The two ends of the polarized actin filament are termed the barbed end and the pointed end based on the orientation of myosin heads that bind to the filament, and the rates of actin polymerization at the two ends [128]. At steady state, ATP-bound actin monomers are more readily added to the more dynamic barbed end of an actin filament and this represents the major mechanism for filament growth and elongation in vivo. Much slower monomer addition occurs at the pointed end [129]. The filament itself acts as an ATPase, converting ATP-bound actin to ADP+Pi and ADP-actin, which results in the accumulation of ADP-actin towards the pointed end [130]. Actin dynamics, i.e. the relative rates of actin polymerization, depolymerization, filament severing, and remodeling into different structures, is influenced by the concentration of G-actin in the cell and is regulated by multiple regulatory proteins (see section 1.5.2).   Actin filaments can be assembled into higher ordered structures or networks that have different physical properties and support different functions. Actin filaments can be organized into branched arrays, such as those found in the lamellipodia of migrating cells, which exert outward forces on membranes to generate forces for cell movement or changes in cell shape [131, 132]. Alternatively, actin filaments can be organized into linear bundles in which the filaments are arranged in a parallel or anti-parallel manner [50, 125]. Parallel actin bundles have all of their barbed ends facing in one direction [131] and are often found in finger-like membrane protrusions such as filopodia or microvilli [133]. Anti-parallel actin bundles can associate with 35 motor proteins, such as myosins, to establish contractile actin networks such as stress fibers [131, 134]. Actin bundling is mediated by proteins such as L-plastin, fascin, fimbrin, filamin, formins, and α-actinin, which crosslink actin filaments and build different types of bundled actin structures (Figure 1.8) [124, 131]. These actin-bundling proteins control the dimensions and stiffness of actin bundles, which may be important for supporting distinct cellular processes. For example, the cell cortex is comprised of crosslinked actin filaments with interspersed contractile bundles that control cell shape and maintain the mechanical integrity of the cell [131]. Several types of actin structures can exist simultaneously in a single cell, especially in migrating cells (Figure 1.8). These actin structures can be rapidly assembled and disassembled as the cell moves. The ability of the cells to remodel its actin cytoskeleton into different types of structures in different parts of the cell is important for rapid responses to external stimuli.  36  Figure 1.8. Actin structures in a migrating cell  Actin-binding proteins that crosslink or bundle actin filaments can establish distinct actin structures with different locations within the cell.  i) The cell cortex is comprised of branched networks of crosslinked filaments as well as contractile bundles.  ii) Stress fibers are anti-parallel bundles of actin that are stabilized by fimbrin and α-actinin. Contractile stress fibers are located at the basal surface, below the nucleus, and contain myosin II. Highly compact, parallel bundles of contractile actin are found at the apical surface, on top of the nucleus. Purple circles represent the sites of focal adhesion formation.  iii) Dendritic or branched actin filaments that are nucleated by the Arp2/3 complex are found in the lamellipodium at the leading edge. Crosslinked actin networks that are established by actin crosslinking proteins such as α-actinin and filamin are found close to the leading edge but behind the lamellipodium.  iv) Parallel actin bundles that are found within filopodia at the leading edge of the cell are formed by fascin, α-actinin, or fimbrin.  Reproduced from [131] with permission. 37 1.5.2 Regulators of actin polymerization All actin structures are formed by two general steps: (1) nucleation, where actin nucleators initiate filament formation, and (2) polymerization, where G-actin monomers are added to the end of a filament [50]. Although actin monomers have been observed to self-assemble into filaments, this self-assembly is slow and the resulting filaments are unstable [135]. Therefore, actin nucleating proteins and other actin-binding proteins that promote actin polymerization, bundling, capping and branching are essential for cells to rapidly remodel their actin networks in response to external stimuli.   1.5.2.1 Profilin When actin monomers are present at concentrations above the “critical concentration” of 0.1 μM, actin monomers can self-assemble into filaments. Although this critical concentration is several orders of magnitude lower than the concentration of actin monomers in the cell, actin filament self-assembly rarely occurs in cells [132]. The high concentration of monomers in the cell is maintained in part by proteins that bind G-actin, such as profilin, which prevent these monomers from spontaneously forming dimers or trimers that can nucleate the formation of new actin filaments [129, 132]. Although profilin prevents spontaneous actin polymerization, the binding of profilin to actin monomers increases the rate of formin-mediated actin polymerization at the barbed ends of actin filaments while inhibiting branched actin polymerization by the Arp2/3 complex [132]. Thus, profilin regulates actin monomer availability and influences the type of F-actin networks that are assembled.  38 1.5.2.2 Nucleation factors promote actin filament assembly The spontaneous association of actin monomers creates highly unstable polymerization intermediates of actin dimers and trimers, which rapidly dissociate. This makes spontaneous nucleation inefficient and energetically costly [126, 127, 129]. However, actin nucleators, such as formin proteins, can stimulate the formation of a filament by efficiently seeding polymerization from a pool of profilin-bound actin monomers (Figure 1.9) [129]. Formins that are recruited to and activated at the plasma membrane initiate the de novo formation of linear unbranched actin filaments [126, 136]. Other proteins that mediate linear actin assembly include spire, cordon-blue, leimodin, and junction-mediating and -regulatory protein (JMY) [129]. In contrast, the Arp2/3 complex nucleates the polymerization of branched actin filaments by binding to the side of a pre-existing primer, or “mother” filament [131]. With the assistance of a nucleation promoting factor (NPF) [127], Arp2/3-mediated nucleation initiates the polymerization of a daughter filament at a characteristic 70o angle with the mother filament, resulting in a Y-shaped branch structure [123]. This type of actin polymerization is called dendritic nucleation [123]. Although the Arp2/3 complex remains associated with the pointed end of the growing actin filament, formins move processively with the growing filament by remaining associated with the barbed end [127].   39  Figure 1.9. Mechanisms of actin assembly  The Arp2/3 complex is a multi-subunit complex that includes the Arp2 and Arp3 proteins, which structurally resemble actin monomers [127]. Arp2/3-mediated actin nucleation requires an NPF. In this example, the NPF N-WASP recruits actin monomers via its WH2 domain, and the Arp2/3 complex via its acidic domain, allowing Arp2/3 to bind to the actin monomer. This complex is stabilized by N-WASP and mimics an actin trimer, where the addition of another actin monomer will create a stable seed. Upon nucleation, the Arp2/3 complex remains associated with the pointed end of the actin filament and does not affect the rate of assembly at the barbed end. Formins are thought to act by stabilizing actin dimers and trimers that form spontaneously, thereby allowing actin nucleation. Formins bind to the barbed end of an actin filament via its donut-shaped FH2 domain and they move processively with the filament as it grows. Formins promote the elongation of an actin filament in two ways, by binding to the barbed end via their FH2 domain and thereby preventing capping proteins from binding and by using their FH1 arm-like domains to sequentially add profilin-bound actin monomers to the barbed end. Adapted from [129] with permission.    40 1.5.3 Arp2/3-mediated polymerization of dendritic actin networks An important function of branched actin networks is the generation of intracellular force [131]. During IS formation, branched actin networks form at the periphery of the B cell:APC interface and drive B cell spreading, which maximizes antigen binding. This dendritic actin network resembles the branched actin network that drives the formation of membrane protrusions at the leading edge of a migrating cell. The formation of dendritic actin networks requires the binding of the Arp2/3 complex to a pre-existing actin filament as well as an NPF that activates Arp2/3.  Several NPFs can stimulate Arp2/3-mediated actin nucleation (see [137] for a review of NPFs). The major Arp2/3-activating NPFs are WASP, the suppressor of cyclic AMP repressor/WASP-family verprolin-homologous protein (SCAR/WAVE), WASP and Scar homologous protein (WASH), and the WASP homologue associated with actin, membranes, and microtubules (WHAMM) [123, 127]. These NPFs share a common VCA domain, which consists of a verprolin/WASP-homology-2 domain (V) that binds G-actin, a central region (C), and an acidic (A) region that binds to Arp2/3 [123]. The VCA domain is sufficient to activate Arp2/3-dependent actin polymerization in vitro [123]. The binding of this domain to the Arp2/3 complex causes a conformational change that allows Arp2/3 to bind to the side of an existing actin filament and then bind an actin monomer [123]. The trimeric complex of Arp2, Arp3, and the actin monomer act as a nucleus onto which actin monomers can be added in order to generate a new actin filament [123].   The structural properties of the dendritic network make it well suited for generating pushing forces against the cell membrane. The activation and location of different Arp2/3 NPFs can determine when, as well as where in the cell, these dendritic networks are assembled. 41 Membrane-proximal actin filaments push against the membrane when new monomers are incorporated at their barbed ends [127]. Capping proteins are then stochastically incorporated at the free barbed ends to maintain shorter actin filaments and prevent the formation of filopodial protrusions [138]. These capped ends are replaced by newly nucleated branched filaments so that the lamellipodium is dominated by branched actin networks [138]. The cooperation between capping proteins and the Arp2/3 complex allows force to be distributed along the entire leading edge of the cell, which gives the dendritic network properties that resemble an expanding gel [127]. Other proteins such as cortactin and the related protein hematopoietic-lineage-cell specific homologue 1 (HS1) can collaborate with the Arp2/3 complex to promote the formation of branched actin networks.    1.5.3.1 HS1 and Arp2/3 regulate the formation of dendritic actin networks Cortactin is a ubiquitously-expressed actin-binding protein that promotes Arp2/3-dependent actin polymerization and plays an important role in cell adhesion, spreading, migration and endocytosis. Hematopoietic lineage cell-specific protein 1 (HS1) [139, 140], or Hematopoietic Cell-Specific Lyn Substrate 1 (HCLS1) [139, 141], a 75-kDa cytoplasmic protein, is a cortactin homologue that is expressed exclusively in hematopoietic cells [140]. HS1 is important for T cell IS formation [142], the antigen-uptake function of APCs [143], B cell development and activation [144], and lymphocyte migration [145].   HS1 and cortactin share several structural similarities that mediate their interactions with other actin-binding proteins and signaling effectors (Figure 1.10). HS1 binds F-actin via three 37-amino acid helix-turn-helix (HTH) repeats and a coiled-coil (CC) region [146, 147]. In vitro, both HS1 and cortactin can interact with Arp2/3 via their N-terminal domain and act as NPFs to 42 increase the rate of Arp2/3-dependent branched actin polymerization [139, 148]. However, HS1 is less efficient at stimulating Arp2/3-dependent actin polymerization than either cortactin or N-WASP, due to structural differences [139, 149]. Instead, the main role of HS1 in lymphocytes is thought to be as an actin filament-stabilizing protein because the loss of HS1 in T cells results in transient lamellipodial protrusions as well as defects in cell spreading [142]. This may reflect a failure to sustain branched actin structures that support membrane protrusions. HS1 can also interact simultaneously with Arp2/3 and F-actin (Figure 1.11). In macrophages, HS1-GFP fusion proteins co-localize with Arp2/3 complexes [139]. However, HS1 has only a modest binding affinity for F-actin compared to cortactin, at least in vitro [139]. This raises the possibility that the F-actin stabilizing function of HS1 depends on other proteins that link HS1 to F-actin, or that it occurs only at sites where there is a high local density of F-actin, e.g. at the lamellipodium of a migrating cell.   The localization and function of HS1 is controlled by the phosphorylation of key regulatory tyrosine residues by SFKs (e.g. Lck, Lyn, Fgr) and other intracellular protein tyrosine kinases such as Syk, Zap-70, and ITK [139]. SFKs phosphorylate HS1 at Y222 whereas Y388 and Y405 in murine HS1 (Y378 and Y397 in human HS1) can be phosphorylated by SFKs as well as Syk [150, 151]. Ligand-induced clustering of the BCR, TCR, and mast cell FcεRI results in rapid phosphorylation of HS1 at these tyrosine residues [152]. The phosphorylation of HS1 creates binding sites for multiple SH2-domain-containing proteins including SFKs, PLCγ, and Vav1 [142], which are important for regulating actin-remodeling pathways.      43   Figure 1.10. The structures of cortactin and HS1  Abbreviations in alphabetical order: Abelson kinase (Abl), coiled-coil region (CC), casein kinase 2 (CK2); extracellular response kinase (Erk; elsewhere in thesis, ERK), focal adhesion kinase (FAK), histone deacetylase 6 (HDAC6), N-terminal arginine (NTA), neural Wiskott–Aldrich Syndrome protein (N-WASP), p21-activated kinase (PAK), p300/CBP-associated factor lysine acetyltransferase (PCAF), proline-rich region (Pro), amino acid repeat (R)/helix turn helix (HTH) motifs, Src family kinase (SFK), sirtuin1 (Sirt1). Amino acid numbers are for the human isoforms. Adapted from [153] with permission.  44  Figure 1.11. Model for the association of HS1 with the Arp2/3 complex and F-actin   HS1 binds to the Arp2/3 complex via its N-terminal DDW residues, and to the newly-formed branched filament via its HTH motifs (shown as R1-R3) and CC domain. Adapted from [147] with permission.   1.5.3.2 The role of HS1 in immune cells HS1 is a hematopoietic cell-specific actin-binding protein that has been implicated in lymphocyte activation and development. Although HS1 knockout mice exhibit normal B cell development, antigen-induced proliferation and apoptosis are impaired in mature B cells [154]. Antibody responses to TD antigens are impaired in these mice, suggesting that HS1 is also important for T cell activation [154]. In T cells from HS1-deficient mice, the initial TCR-induced tyrosine phosphorylation response is normal but calcium flux is impaired [154] as is the NFAT- and NFκB-dependent expression of IL-2 in response to TCR or TCR/CD28 engagement [142]. Moreover, T cells from HS1-deficient mice exhibit aberrant actin dynamics when allowed to spread on anti-CD3-coated coverslips or phospholipid bilayers containing peptide-MHC 45 complexes [142]. This impaired cytoskeletal response is characterized by disorganized actin structures and short-lived membrane protrusions [142]. In terms of the role of HS1 in B cell IS formation, when B cells are presented with polarized arrays of antigen, HS1 localizes at the antigen contact site and recruits the Arp2/3 complex, which is important for cell spreading, the initial phase of IS formation [136]. This function of HS1 may modulate B cell activation as loss-of-function mutations in HS1 are linked to autoimmune diseases such as SLE that are characterized by autoreactive B cells and increased IgM production [155, 156]. Further investigation is required to delineate the mechanisms by which the actin-regulating activities of HS1 control BCR signaling and B cell activation.  1.5.4 Small GTPases regulate actin nucleation The formation and regulation of actin structures are orchestrated by actin-nucleating and actin-regulatory proteins whose activity is controlled both spatially and temporally by small GTPases that act as molecular switches. These small GTPases translate external cues (e.g. integrin engagement, cytokines, growth factors, hormones, chemokines) into intracellular cytoskeletal responses [157]. RhoA, Cdc42, and Rac, which are members of the Rho family of GTPases, each play distinct and critical roles in actin network formation and remodeling [121]. RhoA is required for the activation of formins [134]. Formins nucleate actin filaments via a conserved formin homology-2 (FH2) domain, and promote filament elongation via their formin homology-1 (FH1) domain, which binds to profilin-primed actin monomers (Figure 1.9) [134]. In contrast, activation of the Arp2/3 complex is controlled by Rac and Cdc42 GTPases, which regulate NPFs that activate the Arp2/3 complex [121, 123]. The NPFs WASP, N-WASP, and WAVE bind to the active GTP-bound form of Cdc42 and Rac. This induces a conformational 46 change in the NPFs that relieves the autoinhibition and reveals their VCA domain, allowing the NPFs to bind to and activate the Arp2/3 complex [132]. Actin nucleation occurs primarily at the cell cortex as the active forms of these GTPases are usually associated with the inner face of the plasma membrane [134].   1.5.5 Actin filament disassembly In addition to allowing cells to remodel their actin cytoskeleton in response to external stimuli, the disassembly of existing actin networks allows cells to replenish the pool of free actin monomers that are available for the assembly of new actin structures [158]. After nucleation, filament assembly continues until actin monomer pools are depleted or an actin capping protein is added to the barbed end [132]. In contrast, actin filament disassembly occurs continuously via the release of ADP-actin from the pointed ends of filaments [159]. This plays an important role in a process known as actin treadmilling, which occurs at the lamellipodium [159]. During actin treadmilling, barbed end growth is oriented towards the leading edge of the cell. As growing filaments push against the membrane at the leading edge, this results in the retrograde movement of the filament, which is coupled with depolymerization at the pointed ends that are oriented towards the cell body [159]. The observation that actin filaments disassemble much faster in cells than in solution, and that disassembly can occur along the entire length of the filament, led to the discovery of proteins that disassemble actin filaments via severing or fragmentation [159]. These actin-disassembly proteins play a critical role in regulating cellular actin dynamics and are required for many cellular functions including cell migration, nuclear architecture, apoptosis, lipid metabolism and transcription and are reviewed in [124, 158-164]. Although several actin-binding proteins contribute to actin remodeling, I discuss here cofilin, which was identified as a 47 major actin severing protein in B cells and is important for both actin depolymerization and remodeling during B cell activation [55, 56, 82].   1.5.5.1 The actin-severing protein cofilin regulates actin dynamics Cofilin-1 (also called n-cofilin) is the major form of cofilin in non-muscle cells whereas cofilin-2 (also called m-cofilin) is expressed mainly in differentiated muscle cells [161, 162]. Cofilins and other actin depolymerizing factor (ADF) family proteins are actin-severing proteins and major regulators of actin dynamics [161, 162, 165, 166]. Cofilin/ADF proteins overlap in their actin-severing activity and are required for critical processes such as tissue morphogenesis, cell migration, growth, proliferation, differentiation, maintenance of cell shape, and cell polarization [161, 162]. Although cofilin and ADF proteins have structural similarities, they may have distinct functions in different cell types or during different cellular processes. This is illustrated by the observation that disrupting the cofilin-1 gene in mice results in embryonic lethality whereas ADF deficiencies result in post-natal blindness but not death [162].  The 20-kDa cofilin protein has a conserved ADF-homology domain. Cofilin preferentially binds to older ADP-actin-containing filaments and changes their mechanical properties so as to promote filament disassembly (Figure 1.12) [161]. Cofilin-mediated actin severing occurs in a concentration-dependent manner. At high concentrations, cofilin can bind cooperatively to actin filaments with a 1:1 ratio of cofilin to actin subunits [167]. However, actin filaments that are saturated with cofilin proteins are more resistant to severing compared to filaments that are partially decorated [124, 168]. In contrast, lower cofilin concentrations result in cofilin-decorated segments and bare segments that occur randomly along the filament. Cofilin destabilizes the non-covalent bonds between actin subunits so that cofilin-decorated segments 48 acquire structural flexibility with reduced torsional rigidity [124, 168]. Severing is induced at the boundaries between decorated and non-decorated segments [167-169]. Small patches of cofilin-decorated actin introduce twists and kinks in the actin filament such that filament buckling is concentrated at these sites when other forces are applied to the filament [168]. These forces often result from compression from the cell’s leading edge due to actin assembly or contraction forces that are generated by myosin motor proteins [167, 168]. Hence cofilin binding itself does not induce severing but destabilizes regions of actin filaments such that applied mechanical forces result in filament severing.  Cofilin plays an essential role in actin turnover dynamics by severing existing actin filaments, liberating actin monomers, debranching existing actin filaments, and may create new sites for Arp2/3-mediated branched actin polymerization. Cofilin is often enriched in cellular locations where extensive actin turnover take place such as at the leading edge of motile cells or ruffling membranes [160]. When cofilin is inactivated (see section 1.5.5.2), actin filament turnover decreases and F-actin accumulates due to an increased rate of polymerization versus depolymerization [161]. At physiological ratios of cofilin to actin monomers (1:25-1:4) within a filament, cofilin induces F-actin severing and depolymerization. This replenishes the pool of G-actin monomers for actin assembly [158]. In addition, cofilin can promote the debranching of older ADP-bound filaments in dendritic networks such as those found at the leading edge of migrating cells [160]. This occurs when cofilin binding reduces the affinity of actin filaments for Arp2/3 by propagating structural changes along the filament, or when cofilin competes with Arp2/3 for binding sites, thereby releasing Arp2/3 complexes from actin filaments [170]. This allows the released Arp2/3 complexes to nucleate new branched actin filaments. Importantly, cofilin-mediated F-actin severing creates barbed ends at the sites where F-actin was severed, 49 which may be preferred sites for Arp2/3-dependent actin nucleation and elongation [161, 171, 172].   Figure 1.12. Cofilin-mediated actin severing  Cofilin that is dephosphorylated by phosphatases (e.g. Slingshot) can bind in a cooperative fashion to actin subunits within a filament. Cofilin binding can induce twists that cause the actin filament to become more fragile and susceptible to fragmentation or severing by mechanical forces [173]. Cofilin-mediated actin severing also replenishes the G-actin monomer pool that is available for filament assembly and increases the number of filament ends that can act as sites for dendritic nucleation by the Arp2/3 complex.   1.5.5.2 Regulation of cofilin  The activity of cofilin, i.e. its ability to bind to F-actin filaments, is spatially and temporally regulated by post-translational modifications, lipid binding, and cofilin-binding proteins (Figure 1.13) [161]. Cofilin phosphorylation/dephosphorylation is the primary post-translational mechanism that regulates cofilin-mediated actin dynamics. The activity and the localization of cofilin kinases and phosphatases are both regulated. Phosphorylation of cofilin on 50 S3 inhibits its ability to bind to F-actin. Cofilin can be phosphorylated on S3 by LIM kinase (LIMK) 1/2, testicular protein kinases (TESKs) and Nck-interacting kinase (NIK)-related kinase (NRK) [161, 164]. In B cells, only LIMK1 and LIMK2 are highly expressed [161]. The LIMKs are activated primarily by Rho GTPases via the Rho-associated protein kinase (ROCK) [160]. Rac or Cdc42 can also activate LIMK via PAK4, which phosphorylates LIMK1/2 at T508/T505, or via myotonic dystrophy kinase-related Cdc42-binding kinase α (MRCKα) [174].  In response to extracellular stimuli, cofilin S3 is dephosphorylated, and this allows cofilin to bind to F-actin. This dephosphorylation is mediated primarily by the Slingshot (SSH) 1-3 phosphatases, although this reaction can also be carried out by chronophin, protein phosphatase 1 (PP1), and protein phosphatase 2A (PP2A) [163]. SSH1 is subject to both positive and negative regulation. SSH1 binds to F-actin and localizes to F-actin structures such as those found in the lamellipodia, stress fibers, and cortical actin networks. The cofilin-dephosphorylating activity of SSH1 is greatly enhanced by its binding to F-actin [175]. The increased cofilin activity at lamellipodia, a site of active actin remodeling, is dependent on coronin 1B, which appears to target SSH1 to the lamellipodia [176]. The ability of SSH1 to activate cofilin is negatively regulated by 14-3-3 proteins, which bind to SSH1 when it is phosphorylated at S937 and S978. This prevents SSH1 from interacting with F-actin [177]. The sequestering of SSH1 by 14-3-3 proteins decreases the amount of active cofilin and results in the accumulation of F-actin [161, 176]. Protein kinase D (PKD, also known as PKCµ), which is strongly activated by BCR signaling [178] can phosphorylate SSH1 at S978 [174]. PKD can also inactivate SSH1 by phosphorylating it on S402 within the phosphatase domain [179]. In T cells, calcineurin promotes Ca2+-dependent activation of SSH1 by dephosphorylating SSH1 and preventing its interaction with 14-3-3 proteins [180]. In B cells, BCR signaling results in cofilin 51 dephosphorylation and increased cofilin-mediated actin severing [55]. The Rap GTPases mediate this response but the mechanism is not known [55].  Membrane sequestration is another mechanism for controlling cofilin function. Binding to PIP2 inhibits the function of cofilin [132]. External cellular stimuli such as chemoattractants and antigens induce the hydrolysis of PIP2 by PLC, which can result in localized release and activation of cofilin. This promotes the formation of membrane protrusions at these sites [160]. PI3K, which is also activated by chemoattractants, can phosphorylate PIP2 to yield PI(3,4,5)P3, which also releases cofilin from the plasma membrane [181]. PIP2-mediated sequestration of cofilin acts independently of LIMK activity to inhibit cofilin function [182]. Cofilin function can also be regulated by intracellular pH which is controlled by the Na+-H+ exchanger NHE1, which pumps protons out of the cell [160]. Receptor signaling often results in cytoplasmic alkalinization and the interaction between PIP2 and cofilin is reduced at higher pH [160]. Cortactin can also bind cofilin and sequester it such that it cannot sever actin filaments and this interaction is also disrupted at higher pH [183].    52  Figure 1.13. Cofilin regulation  Cofilin is regulated by post-translational modifications, protein-protein interactions, and changes in cytoplasmic pH. Phosphorylation of cofilin on S3 by LIM kinases (LIMKs) or testicular protein kinases (TESKs) renders it inactive. Cofilin can also bind to PIP2, which sequesters cofilin at the plasma membrane and prevents it from binding to F-actin. Cortactin binding can also inhibit the function of cofilin. Cofilin is activated when it is dephosphorylated on S3 by phosphatases such as slingshot (SSH), protein phosphatase 1 (PP1) and protein phosphatase 2A (PP2A). Increases in cytoplasmic pH that are mediated by the proton pump sodium/hydrogen exchanger 1 (NHE1) also promote cofilin activation [160]. Abbreviations: Phospholipase (PLC), inositol trisphosphate (IP3), diacylglycerol (DAG), phosphatidylinositol 4,5-bisphosphate (PIP2), phosphoinositide-3-kinase (PI3K), phosphatidylinositol (3,4,5)-trisphosphate (PIP3). See text for details.   1.5.6 The actin cytoskeleton and immune synapse formation During B cell activation, remodeling of the actin cytoskeleton is critical for the formation of the IS and for B cell activation. Actin regulators are activated by BCR signaling and control multiple aspects of IS formation, including the release of BCRs from actin-based confinement zones and the formation of the cSMAC. In this section I describe the role of the actin 53 cytoskeleton and its regulators in mediating BCR microcluster formation, B cell spreading and membrane contraction, all of which are important for IS formation.  Upon BCR signaling, the submembrane actin cytoskeleton is disassembled and uncoupled from the plasma membrane in order to allow increased BCR mobility, clustering, and signaling [44]. The removal of diffusion barriers allows BCR nanoclusters to aggregate with one another and with CD19 clusters to form signaling microclusters that recruit Lyn and Syk, the adaptor proteins BLNK and Grb2, and the signaling enzymes Vav and PLCγ [12, 40, 44, 50, 53, 67, 112, 184] (see sections1.1.1, 1.3.1, and 1.3.2). The dephosphorylation of ezrin uncouples the actin cytoskeleton from the plasma membrane and increases BCR mobility [9]. The disassembly of the submembrane actin cytoskeleton is mediated by the actin-severing protein cofilin, which is rapidly dephosphorylated and activated upon BCR signaling [55]. Cofilin-mediated actin severing is an important step in remodeling the actin cytoskeleton. Actin severing replenishes the G-actin pool of monomers and generates additional free barbed ends that can be used by the actin-assembly proteins to assemble new actin networks, which are important for BCR signaling and B cell activation (see section 1.5.5).  B cell spreading is an important pre-requisite for B cell IS formation and can shape the outcome of B cell activation. Spreading allows the B cell to encounter additional antigen molecules, which increases BCR microsignalosome formation such that the level of BCR signaling exceeds the threshold for B cell activation [9]. Additionally, spreading increases the ability of B cells to elicit T cell help, which relies on the amount of antigen that is bound by BCRs and internalized for presentation to T cells [64]. Therefore, the molecular regulators of cytoskeletal remodeling are critical for determining the outcome of B cell activation. Upon engaging cognate antigen, a dendritic network of actin that exhibits actin treadmilling activity 54 forms at the periphery of the antigen contact site while F-actin is progressively cleared from the center of the contact site [82]. The peripheral ring of dendritic actin generates forces that initiate symmetrical membrane spreading at the planar B cell:APC interface, which resembles a radial lamellipodium [9, 32, 64, 65]. This spreading response requires BCR signaling and actin dynamics as both inhibiting SFKs and disrupting the actin cytoskeleton abrogates B cell spreading [64]. A screen using mutants of the DT40 B cell line showed that multiple BCR signaling effectors including Lyn, Syk, Btk, Vav, BLNK and PLCγ2 are required for B cell spreading but that the three IP3 receptors are dispensable [9, 112, 185].  The actin severing function of cofilin is also important for the actin cytoskeleton remodeling that drives B cell spreading [55, 82]. BCR signaling triggers Rap-dependent cofilin dephosphorylation and activation [55, 82]. Song and colleagues showed using total internal reflection fluorescence microscopy (TIRFM) that dephosphorylated cofilin is recruited to the site of antigen contact near BCR microclusters, where it promotes actin severing at the cytoplasmic face of the cortical actin network [186].  Although cofilin is important for actin severing, other cytoskeletal regulators promote the formation of the dendritic actin network that drives B cell spreading. BCR signaling recruits Vav to the plasma membrane, where it acts as a GEF that activates the Rac and Cdc42 GTPases (see section 1.5.5) [187]. Rac and Cdc42 are important for activating the NPFs WASP and N-WASP in B cells [112]. Global deficiencies in WASP result in complex immune disorders including autoimmunity, immunodeficiency, and lymphatic cancers [112]. WASP deletion in mice is sufficient to cause B cell-mediated autoimmunity [188]. In B cells, both WASP and N-WASP are responsible for B cell spreading and appear to have overlapping functions [189]. Although both WASP and N-WASP are important for activating Arp2/3-dependent actin nucleation, and 55 this is required for T cell spreading [80], a requirement for Arp2/3 during B cell spreading has not been demonstrated.  B cell spreading persists for only a few minutes before the B cell membrane contracts, BCR microclusters coalesce into a cSMAC, and integrins and the actin cytoskeleton are organized into a bullseye pattern as described in section 1.3. During this contraction phase, the cell membrane retracts while the actin cytoskeleton assembles curved bundles of actin filaments (called actin arcs) that form in the inner lamellum of the peripheral F-actin network and condense into a contractile ring [80-82]. Although actin dynamics are clearly important for the centripetal movement of antigen receptor microclusters, the mechanism is not completely understood. Yi et al. showed that the rate of TCR microcluster movement is nearly identical to the rate of actin retrograde flow in in the lamellipodium of T cells that have spread on lipid bilayers containing peptide-MHC complexes [81]. However once the TCRs reach the inner lamellum, their centripetal movement is 3-fold slower [81]. This suggests that actin retrograde flow drives the initial centripetal movement of TCR microclusters but that another mechanism is responsible for the final coalescence of TCR microclusters into a cSMAC. Indeed, Murugesan et al. showed that formin-dependent actin arcs in the lamellum promote TCR microcluster coalescence by sweeping the microclusters towards the center of the antigen contact site during the contraction phase [80]. Thus distinct types of actin-based forces drive receptor centralization during T cell contraction [81].  In B cells, the Rap1-cofilin pathway is important for BCR microclusters to coalesce into a cSMAC [55, 82]. Although cofilin is required for B cell spreading, which enhances BCR microcluster formation, some microclusters still form when cofilin-mediated actin severing is blocked [55]. However, these microclusters remain dispersed across the B cell:APC contact site, 56 which suggests that cofilin-dependent actin processes are important for microcluster centralization. How cofilin-mediated actin dynamics promotes cSMAC formation remains to be elucidated. Another mechanism by which the B cell switches from cell spreading to contraction is through the SHIP-1 phosphatase, which dephosphorylates PIP3. This decreases the activation of Btk, an upstream regulator of the Vav-WASP pathway [189]. Although actin-dependent forces are important for moving peripheral BCR microclusters towards the center of the antigen contact site, the cSMAC is largely devoid of dense actin structures. Therefore, other mechanisms must also be involved in the centralization of microclusters towards the center of the B cell:APC contact site. Indeed, in B cells the microtubule-based motor protein dynein has been implicated in cSMAC formation, which requires the reorientation of the MTOC towards the antigen-contact site [190]. I describe in the next section the mechanisms by which cell polarity cues induce MTOC reorientation, as well as the functions of MTOC polarization during IS formation.   1.6 The MTOC and cell polarity  B cells often encounter antigen that is presented as a polarized array on the surface of an APC. In these situations, an IS forms at the site of antigen contact where the concentration and organization of receptors, signaling proteins, and cytoskeletal components at one pole of the cell initiates cell polarity (Figure 1.14). Cell polarity is established when cellular components and structures are organized in an asymmetric manner within the cell [134]. Two fundamental properties establish cell polarity: (1) the asymmetric distribution of mobile cellular components between opposite ends of a cell and (2) the organization of inherently polarized structures along the axis of polarity [134]. In general, external polarity stimuli cause the actin cytoskeleton to undergo rapid remodeling, which breaks symmetry by establishing the polarized organization of 57 regulatory proteins and by dynamically assembling distinct cytoskeletal structures [134]. The microtubule cytoskeleton, in most cases, is thought to stabilize this asymmetric distribution in order to maintain polarity [134]. The maintenance of polarity is particularly important in cells that remain in polarized states such as neurons and epithelial cells. In contrast, during lymphocyte migration and IS formation, short-term polarization, combined with the ability to respond rapidly to external stimuli, may be more important for making an appropriate response.     58  Figure 1.14. B cells form a polarized IS structure  The spatially-restricted antigen binding and BCR signaling that occur at the B cell:APC interface establish a cell polarity in which signaling proteins and cytoskeletal components are reorganized at the site of antigen contact. The image shows a cross-section of a B cell interacting with an APC. An IS forms in which antigen-bound BCRs are gathered at the center of the contact site and surrounded by a ring of intercellular adhesion molecule-1 (ICAM-1)-bound lymphocyte function-associated antigen-1 (LFA-1) integrins. Signaling proteins are recruited to the antigen-bound BCRs at the contact site. Actin is reorganized such that it is cleared from the central region of the B cell:APC contact site where BCRs are gathered, and accumulates at the periphery of the contact site, underlying the integrins. Additionally, cortical microtubule capture proteins (e.g. IQGAP1, dynein) at the antigen contact site bind to microtubule plus ends to move the MTOC towards the site of antigen contact.  59 1.6.1 The MTOC and its roles in cell polarity Microtubules are built from polymers of α and β tubulin dimers that are assembled into hollow cylindrical structures. The microtubule cytoskeleton consists of an aster of microtubules that are attached via their minus-ends to a central MTOC while their plus ends extend outwards towards the cell periphery. The MTOC is a cellular structure that assembles and organizes microtubule arrays [191]. In interphase cells, the main MTOC is the centrosome. Other acentrosomal MTOC structures that can be found in cells include the mitotic spindle, basal bodies, and secondary MTOCs that are associated with membranes and organelles [191].  The MTOC creates cellular asymmetry and controls cell polarity by organizing intracellular components, controlling the position of the nucleus and the Golgi apparatus, and supporting the directional trafficking of vesicles within the cell. This is essential for polarized cellular processes such as cell migration and asymmetric cell division (reviewed in [134, 192] and [193]). During interphase, the MTOC is located in a perinuclear position close to or attached to the nuclear envelope via the transmembrane adaptor protein Velcro [136, 194]. However, when the cell receives a polarity cue, the MTOC can be repositioned to mediate cell polarization. When macrophages, fibroblasts, endothelial cells, astrocytes, and other adherent cells undergo persistent directed migration towards chemoattractants, the MTOC is reoriented such that it is between the nucleus and the leading edge of the cell to maintain cell polarity [134, 193, 195, 196]. In these polarized cells, the MTOC organizes microtubules so that they are asymmetrically distributed along the polarity axis [193]. Stable microtubules are extended towards the leading edge whereas dynamic microtubules extend towards the rear of the cell [134]. MTOC polarization also reorients the Golgi apparatus and the endocytic recycling compartments, which are associated with the MTOC, towards the leading edge. The polarized delivery of vesicles from 60 the Golgi apparatus to the leading edge occurs along the stable microtubules and plays a key role in supporting lamellipodial cell spreading and forward cell movement [134]. MTOC polarization is also important for cell type-specific processes such as ciliogenesis in epithelial cells, spindle orientation during the asymmetric cell division of stem cells and embryonic cells, axon specification in neuronal cells, and T and B cell IS formation (discussed below) [134, 193]. The structure of the MTOC is well suited for its role in anchoring microtubules and positioning the microtubule network during cell polarization.   1.6.1.1 The structure of the centrosomal MTOC In most cells, including lymphocytes, the centrosome functions as the major MTOC [194]. The centrosome is an organelle that is comprised of a pair of cylindrical centrioles that display 9-fold radial symmetry [197]. The centrosome nucleates and anchors microtubules [197]. The architecture of the centriole is a cartwheel structure that symmetrically organizes 9 sets of microtubule triplets (Figure 1.15). Although both centrioles can nucleate microtubules, only the mother centriole anchors the minus-ends of microtubules [197]. The appendages on the mother centriole can also promote membrane docking of the centrosome [194]. This is particularly important for ciliogenesis as the mother centriole can attach to the plasma membrane and become a basal body that gives rise to the cilia and flagella [198].  The centrioles are surrounded by a dense matrix of proteins called the pericentriolar material (PCM) which is comprised of a complex of over 100 different proteins and assumes a multilayered donut-shaped cloud around the centriole [194]. The PCM is important for supporting microtubule nucleation as well as the anchoring of microtubules [199]. Several key proteins involved in PCM formation include scaffolding proteins (e.g. Cep152, pericentrin), 61 kinases (e.g. polo-like kinase), phosphatases and structural proteins (e.g. γ-tubulin), which are reviewed in [199]. This multilayered organization may act as a scaffold for recruiting microtubule nucleation factors [197, 200]. Microtubule polymerization at the MTOC is thought to involve 10-13 γ-tubulin proteins, along with other nucleation proteins, which form a ring like complex called the γ-tubulin ring complex (γ-TuRC) [201]. This ring complex acts as a seed onto which tubulin dimers assemble into tubules [201]. The resulting microtubules can be highly dynamic and can rapidly switch between phases of growth and shrinkage, a phenomenon termed dynamic instability [202]. Microtubule dynamics, as well as the interaction of microtubules with the cell cortex, can exert forces on the centrosome, which are important for controlling its position in the cell. The structure of the centrosome is suited to withstand such forces and prevent fragmentation so that when the MTOC moves, the attached microtubules are repositioned along with the MTOC [203].    Figure 1.15. The structure of the centrosome  The left panel shows a cartoon of a lymphocyte with a centrosome that deforms the nucleus. The centrosome is comprised of a mother centriole (MC, red) and a daughter centriole (DC, green) that are structurally and functionally distinct (center panel). The mother centriole anchors microtubules and may be important for docking the centrosome during polarity processes such as ciliogenesis (right panel). Acquired from [197] with permission.  62 1.6.1.2 Establishing MTOC polarity Because the MTOC plays an essential function in intracellular organization, how the positioning of the MTOC is regulated is a key question. In contrast to the actin severing and repolymerization that occurs at one pole of the cell when cell polarity is established, polarization of the microtubule network involves proteins that capture the plus ends of microtubules and move the MTOC to one pole of the cell [134]. Importantly, the signaling pathways and proteins involved in repositioning the MTOC can be cell-type-specific and depend on the nature of the polarity cue that the cell receives.  Interphase cells often maintain their centrosomes in a central position within the cell via pushing and pulling forces that result from microtubule dynamics and interactions with the cell cortex [192]. Cells growing on micropatterned surfaces can sense cell shape and position the centrosome to the center of the cell [197]. In lymphocytes, the MTOC is often seen within an invagination of the nucleus, suggesting that the forces that maintain MTOC position are strong enough to deform the nucleus [204]. When the cell receives an external stimulus, repositioning of the MTOC involves the cortical capture of microtubule plus ends, which facilitates movement of the MTOC by allowing forces to be exerted on the captured microtubules.  The forces exerted on the MTOC can be divided into pushing forces and pulling forces. Pushing forces can result from the polymerization of microtubules or from the movement of other organelles to which microtubules are attached. Pushing forces alone can be sufficient to reorient the MTOC, especially in smaller cells such as yeast where shorter microtubules that resist buckling can push against the cell membrane as they polymerize, and thereby reorient the MTOC [205]. However, in very large cells, long microtubules will buckle rather than transmit pushing forces [206]. Contractile actomyosin networks can also exert pushing forces on the 63 nucleus that move the MTOC. These pushing forces can be transmitted directly to the MTOC if it is closely associated with the nucleus, or transmitted via microtubules that connect the MTOC to the nucleus [192]. Pulling forces are often generated by mechanisms originating from the cell cortex. This requires both a force-generating mechanism and a microtubule cortical capture mechanism, which together reposition the MTOC towards the source of the pulling force. Cortical capture requires: (1) proteins that are associated with microtubule plus ends (collectively referred to as plus-end-tracking proteins (+TIPs)), and (2) cortical proteins that capture the microtubule plus ends. Some of the cortical proteins involved in capturing microtubules include IQGAP1, mDia1, LL5β, PAR6, aPKC, β-catenin, dynein APC and CLASPs (Figure 1.16) [134, 196]. Capture events mediated by +TIPs can also regulate dynamic instability, which can be harnessed as a force-generating mechanism to move the MTOC [134, 196, 207, 208]. However, the major mechanism by which pulling forces are generated to move the MTOC are mediated by the minus-end directed movement of cortically-anchored dynein along microtubules [206]. Dynein-dependent pulling forces are important for positioning the mitotic spindle during cell division as well for reorienting the MTOC during cellular processes such as cell migration [196] and T cell IS formation [209-211].    64  Figure 1.16. Proteins involved in the cortical capture of microtubules   Some cortical capture proteins are associated with the actin cytoskeleton while others can be associated with the cell membrane. Typically, cortical capture proteins bind to the +TIPs that form complexes at the plus ends of microtubules. Cortical capture can influence the dynamic instability of microtubules. Microtubule-stabilizing factors including CLASPs, APCs, CLIPs, spectraplakins, and dynein-dynactin complexes can capture microtubules at the cell cortex. See text for details.    1.6.2 MTOC polarization in lymphocytes Multiphoton microscopy has shown that lymphocytes assume polarized cell morphologies and exhibit rapid motility within resting SLOs, even in the absence of antigen stimulation [8, 212]. The MTOC in these highly motile leukocytes is polarized towards the rear of the cell [213] and this is thought to concentrate the Golgi apparatus, as well as the microtubule and intermediate filament networks, behind the nucleus in the uropod during rapid cell migration [214]. However, upon antigen encounter, the lymphocyte depolarizes and then re-establishes intracellular polarity towards the APC. The MTOC is rapidly reoriented towards the site of antigen contact, docks in close proximity to the IS and contributes to IS formation and the directed secretion of intracellular vesicles [85, 190, 213, 215].  65 MTOC polarization in lymphocytes has been well-studied in cytotoxic T cells (CTLs) forming contacts with target cells [216], where MTOC docking at the IS polarizes cytotoxic granules to the target cell [217]. The proximity of the MTOC to the contact surface is critical for directional granule secretion as inhibition of MTOC polarization in CTLs results in a disorganized IS architecture, pre-mature termination of TCR signaling, reduced IL-2 production [218] and impaired cytotoxic function [219].  In B cells, MTOC reorientation plays key a role in IS formation, antigen acquisition and processing, and B cell differentiation. During IS formation, the MTOC polarizes close to the plasma membrane (within a 50 nm distance from the membrane [215]) and the microtubules align along the inner face of the plasma membrane at the contact site between the B cell and the antigen-bearing membrane [190]. Minus-end directed dynein motor proteins move along microtubules to facilitate the coalescence of BCR microclusters into a cSMAC (Figure 1.17) [190]. The polarity proteins Par3 and PKCζ are involved in recruiting dynein to the IS [220, 221]. Dynein can bind to BCR microclusters via an adaptor protein complex that consists of Cbl, Dok-3, and Grb2 [190]. Dynein activity is also required for TCR microcluster centralization [215]. Analogous to the role of MTOC polarization in T cells, which directs the polarized secretion of cytotoxic granules, MTOC reorientation in B cells guides the polarization of lysosomes to the IS [85]. Lysosomes exocytosis at the IS is important for the antigen extraction function of the IS [85, 109]. Reorientation of the MTOC also positions organelles such as the Golgi apparatus, degradative compartments and MHC II loading compartments close the IS to facilitate the processing of internalized antigens [222, 223]. During the formation of B cell:T helper cell conjugates, the MTOCs of both cells reorient towards the IS, enhancing the mutual 66 activation of both lymphocytes [224]. MTOC polarization is also important for determining B cell fate. PKCζ, Bcl 6 and the IL-21R polarize with the MTOC at the IS in pre-mitotic B cells and remain polarized through mitosis and asymmetric division. The APC-proximal daughter cells that preferentially inherit these proteins are thought to become plasma cells whereas the APC-distal cells become memory B cells [225]. The signaling pathways used by the BCR to promote MTOC polarization to the IS are not fully understood but may be similar to the pathways used in T cells.       Figure 1.17. The MTOC is reoriented towards the antigen contact site during immune synapse formation  When B cells engage antigen presented on the surface of an APC, the actin and microtubule cytoskeletons are coordinately reorganized to facilitate the formation of the IS. Actin polymerization drives B cell spreading over the surface of the APC. At the same time, the MTOC is reoriented and moves towards the site of antigen contact such that microtubules align along the inner face of the plasma membrane at the contact site between the B cell and the antigen-bearing membrane. The retrograde flow of the peripheral actin moves BCR microclusters towards the center of the antigen contact site. Dynein motor complexes then associate with BCR microclusters and propel the antigen-bound microclusters centripetally to form the cSMAC of the IS. Adapted from [15, 82] with permission.  67 1.6.2.1 Mechanisms underlying MTOC reorientation in lymphocytes  Three models that describe mechanisms for repositioning the MTOC in T lymphocytes have been proposed: cortical sliding, and end-on capture-shrinkage, and cortical capture/actin reorganization (Figure 1.18) [211, 217, 218, 226]. Dynein is required for reorienting the MTOC in T cells. Dynein motor proteins that are anchored at the cell cortex via adaptor proteins can generate pulling forces that reposition the MTOC as they move processively along microtubules [134]. Dynein-mediated pulling during MTOC reorientation results in microtubules sliding across the cell cortex (cortical-sliding model) or depolymerizing at the plus ends (end-on-capture shrinkage model) [211, 226]. Alternatively, reorganization of the actin cytoskeleton is linked to reorienting the MTOC to the IS [217]. In T cells, actin polymerization at the periphery of the antigen contact site, coupled with actin clearance from the center of the contact site, is correlated with MTOC reorientation and docking at the IS [217]. This mechanism of reorienting the MTOC may require capture proteins that link the microtubule plus ends to the actin cytoskeleton such as the actin-microtubule crosslinking protein IQGAP1. Importantly, multiple mechanisms may act simultaneously or in a coordinated manner to reposition the MTOC towards APCs.  In T cells MTOC polarization appears to occur in two steps [226]. This biphasic mode of MTOC polarization is characterized by an initial rapid MTOC movement towards the plasma membrane, followed by slower movement as the MTOC docks at the IS [226]. Using transmission electron microscopy to visualize the docking of the centrosome at the IS, Stinchecombe et al. observed a close association between the centrosome and the plasma membrane adjacent to the site where a CTL forms an IS with a target cell [217, 219]. The initial fast MTOC movement across the volume of the cell has been termed a “fast direction sensing and reorientation phase” [227]. This phase is controlled, at least in part, by DAG and novel PKCs (nPKCs) [227]. A change in speed 68 occurs once the MTOC moves within 2 μm of the IS [226]. This second slower or stabilizing phase appears to be important for maintaining the polarized MTOC at the IS and requires the Par3-Par6-aPKC and Scribble-Dlg polarity complexes [85, 227]. The slower speed may reflect a switch from a rapid initial cortical sliding mechanism to the slower end-on capture shrinkage mechanism. Alternatively, resistive forces that slow down MTOC movement may come into play as the MTOC nears the plasma membrane. These opposing forces could be due to microtubules that extend rearward from the MTOC and connect the MTOC to the nucleus or to the distal cell cortex. Alternatively, the presence of organelles or other cellular structures may impede the close approximation of the MTOC to the plasma membrane at the IS. In order for the centrosome to contact the plasma membrane, the dense amorphous mass of PCM may need to be repositioned away from the IS or the centrioles may need to push past the dense PCM to reach the plasma membrane (personal communication, Nasser Rusan, NIH). This may account for the slower rate of MTOC movement during the membrane-docking phase compared to the initial rapid reorientation phase. Although the dynamics of MTOC reorientation in T cells has been described and possible mechanisms have been proposed, much less is known about the mechanisms that reorient the MTOC to the B cell IS.  69    70 Figure 1.18. Mechanisms that can reorient the MTOC in lymphocytes  BCR signaling induces the remodeling of the actin cytoskeleton at the antigen contact site. Concomitantly, the MTOC is reoriented to the site of antigen contact. +TIPs (e.g. EBs, CLIP-170, CLASPs) interact with capture proteins at the cell cortex (e.g. IQGAP1, spectraplackins) that are often associated with the actin cytoskeleton (A). Actin-driven cell spreading over the APC at the antigen contact site can exert pulling forces on the microtubules to reposition the MTOC towards the APC. Alternatively, the +TIP dynein that is associated with the cell cortex (via adaptor or cortical capture proteins) at the antigen contact site can generate pulling forces on microtubules via its minus end-directed movement (B, C). The microtubule plus ends can then either depolymerize as it approaches the cell cortex (end-on capture-shrinkage model, B) or slide along the cell cortex (cortical sliding model, C). These mechanisms may act at the same time on different microtubules to reorient the MTOC. See text for details.   1.7 The roles of IQGAP1 and the Rap GTPases in cytoskeletal regulation  1.7.1 The IQGAP1 scaffolding protein  During many different cellular processes, including cell migration and IS formation, the reorganization of the actin cytoskeleton and the microtubule network often occur in a coordinated manner. Proteins that can physically link actin filaments to microtubules may be essential for coupling the reorganization of the two cytoskeletons. IQGAP1 is a ubiquitously-expressed scaffolding protein that can bind to both F-actin and microtubule +TIPs, suggesting that it can couple localized actin remodeling to the reorientation of the microtubule network. IQGAP1 has been shown to play a role in establishing MTOC polarity during cell migration [228] and T cell interactions with APCs [82, 217, 229, 230]. Because of its ability to interact with many proteins, IQGAP1 also regulates other cellular processes including cell adhesion, polarity, and protein trafficking [231, 232]. In addition to IQGAP1, other actin-microtubule crosslinking proteins such as spectraplakins and septins may promote MTOC reorientation. Whether these actin-microtubule crosslinking proteins each have distinct functions in different 71 cell types or in different cellular processes is not clear. The roles of these proteins in B cells have not been investigated.  1.7.1.1 The expression and structure of IQGAP1  There are three isoforms of IQGAPs (1, 2, and 3) that are expressed in different tissues [230]. The most ubiquitously expressed and well-studied isoform is IQGAP1, which is the major IQGAP isoform in lymphocytes [230, 233]. The 190-kDa IQGAP1 scaffolding protein contains multiple protein interaction domains and more than 50 binding partners for IQGAP1 have been identified [230]. The domain structure and interaction partners of IQGAP1 are shown in Figure 1.19. The N-terminal calponin homology domain (CHD) allows IQGAP1 to bind directly to the actin cytoskeleton and the N-WASP NPF [230]. C-terminal to the CHD is a coiled coil domain (CC), which consists of a conserved repeat of hydrophobic and charged amino acids in an HXXHCXC pattern [234]. These repeats can bind the Four-point-one, ezrin, radixin, moesin (FERM) domain of ezrin [235]. The WW motif in the center region of IQGAP1 is important for binding both proline-rich proteins [234] and the ERK1/2 kinases [230]. IQGAP1 also contains four tandem repeats of isoleucine-glutamine (IQ) motifs that can bind a variety of signaling proteins including calmodulin, Rap1 [234, 236], Mek1/2 [237] and phosphoinositol-4-monophosphatase 5 kinase type Iγ (PIPKγ) [238]. The C-terminal Ras GAP-related domain (GRD) of IQGAP1 can interact with Cdc42 and Rac1, as well as Rap1 [230]. However, the IQGAP1 GRD does not have GTPase-activating activity because it lacks the catalytic arginine finger that promotes GTP hydrolysis [230, 234]. The Ras GAP C-terminus (RGCT) domain can interact with several microtubule plus-end binding proteins including CLIP-170. This RGCT 72 domain of IQGAP1 also binds the phospholipid PIP2, which can anchor IQGAP1 to the plasma membrane and allow it to recruit its binding partners to membrane sites [238].    Figure 1.19. IQGAP1 structure and protein interaction motifs   IQGAP1 contains 6 protein interaction domains: a calponin homology domain (CHD), a coiled coil domain (CC), two WW domains, 4 tandem isoleucine/glutamine-containing (IQ) domains, a Ras GAP-related domain (GRD), and a Ras-GAP C terminus domain (RGCT). These domains mediate interactions with several binding partners. See text for details. Adapted from [230] with permission.   1.7.1.2 IQGAP1 regulates cytoskeletal organization IQGAP1 promotes remodeling of the actin and microtubule cytoskeletons by regulating the function and localization of its binding partners. The CHD of IQGAP1 binds actin filaments and promotes actin filament cross-linking and bundling [239, 240]. This domain can also 73 promote branched actin assembly via Arp2/3 by recruiting the N-WASP NPF [241]. Additionally, the C-terminal domain of IQGAP1 promotes the stabilization of actin filaments by capping barbed ends [242].  In addition to regulating actin dynamics, the ability of IQGAP1 to bind to both actin and microtubules is important for coupling actin cytoskeleton to positioning of the MTOC. IQGAP1 can capture microtubules by binding the +TIP CLIP-170, which promotes the stability of captured microtubules and supports repositioning of the MTOC [228, 239]. IQGAP1 is also an effector of the Rac1 and Cdc42 GTPases, which are key mediators of receptor-induced cytoskeletal reorganization. In live cell imaging experiments, a tripartite complex of activated Rac1/Cdc42, IQGAP1, and CLIP-170 were observed at cortical sites where microtubules were captured at the leading edge of migrating fibroblasts [228]. Depleting IQGAP1 in these cells showed that it is required for stabilizing microtubule interactions with the cell cortex [228]. Hence IQGAP1 may play a critical role in coordinating actin and microtubule organization during cell polarization.  1.7.2 The Rap GTPases regulate cell polarity and cytoskeletal remodeling  The Rap GTPases, which are members of the Ras-like family of GTPases, are evolutionarily conserved regulators of cell polarity [243, 244], cell adhesion [245, 246] and actin cytoskeleton remodeling [55, 246, 247]. Our lab has previously shown that the Rap GTPases play an essential role in BCR- and chemokine-induced integrin activation, chemokine-induced migration, BCR-induced actin remodeling, and BCR-induced IS formation [55, 68, 82, 247-252]. Although Rap has an evolutionarily conserved role in establishing cell polarity, its role in regulating MTOC polarization during IS formation had not been studied.  74 1.7.2.1 The Rap GTPases  There are five mammalian Rap GTPases (collectively referred to herein as Rap), each encoded by a separate gene: Rap1a, Rap1b, Rap2a, Rap2b, and Rap2c. Based on sequence homology the Rap family is divided into Rap1 or Rap2 GTPase proteins. There is ~95% amino acid sequence identity between Rap1a and Rap1b. Rap2a, Rap2b, and Rap2c share ~90% amino acid sequence identity. All of the Rap proteins are ~21 kDa. Like other small GTPases, the Rap proteins contain a guanine nucleotide-binding pocket (reviewed in detail in [253]) as well as effector binding domains [254]. In B cells, Rap1b is expressed at much higher levels than Rap1a [255]. All three Rap2 isoforms are expressed in B cells (Immgen database, www.immgen.org) but their relative expression levels are not known.  Rap is activated by many extracellular stimuli including growth factors, cytokines, antigens, physical force, and osmotic pressure [246]. Rap proteins cycle between a GDP-bound inactive state and a GTP-bound active state, which binds multiple downstream effectors that mediate Rap-regulated cellular functions. Guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs) control the activation and functions of Rap. GEFs convert Rap GTPases to their active GTP-bound state by deforming the nucleotide-binding pocket of Rap resulting in the release of GDP [70, 256]. GTP, which is present in cells at 10-fold higher concentrations than GDP, then binds the nucleotide-binding pocket and displaces the GEF [254]. GAPs convert Rap GTPases to their inactive GDP-bound state by providing an essential catalytic group and thereby accelerating the intrinsic GTP-hydrolysis activity of Rap [70, 256, 257] (Figure 1.20). In its active state, Rap can interact with a number of effector proteins via its Ras-association (RA) and Ras-binding domains (RBD) [254].  75  Figure 1.20. Rap activation  Inactive Rap is bound to GDP. When the cell receives an extracellular stimulus, guanine nucleotide exchange factors (GEFs) are released from autoinhibition, or recruited to the plasma membrane where Rap is located, and then stimulate the exchange of GDP for GTP in the nucleotide-binding pocket of Rap. GTP-bound Rap is the active form, which can bind effector proteins that mediate the effects of Rap on cellular functions that such as adhesion, cytoskeletal remodeling, polarity, and gene expression. GTPase-activation proteins (GAPs) increase the GTP hydrolysis activity of Rap, resulting in the conversion of GTP to GDP, returning Rap to its inactive state.   1.7.3 Rap GEFs and GAPs GEFs and GAPs control the activation state of GTPases. The activity and subcellular localization of GEFs and GAPs can be regulated by four general mechanisms: post-translational modifications, protein-protein interactions, interactions with second messengers, and lipid 76 interactions [257]. These regulatory mechanisms induce subcellular translocation, release of autoinhibition, or allosteric changes in the catalytic domain [257]. Because Rap binds GDP with high affinity [258], the catalytic action of GEFs is required for external stimuli to rapidly activate of Rap [258]. RapGEFs contain a characteristic catalytic CDC25 homology domain, which mediates the nucleotide exchange activity, as well as a Ras exchange motif for catalysis [256]. There are multiple GEFs that that can activate Rap1 including C3G (RapGEF1), PDZ-GEF1 (RapGEF2), PDZ-GEF2 (RapGEF6), Epac1, (RapGEF3) Epac2 (RapGEF4), Repac (RapGEF5), CalDAG-GEF1 (RasGRP2) and Dock4 [246]. All of these GEFs are expressed in B cells (Immgen database). However, RasGRP2 has been implicated in BCR-induced Rap activation and will be described in more detail below. Although their catalytic mechanism is conserved, different RapGEFs contain unique domains that regulate their activation or localization [254]. For example, Epac is normally autoinhibited, but upon binding cAMP it undergoes a conformational change that relieves the autoinhibition and allows it to act as a RapGEF [259].  Because the intrinsic GTP hydrolysis rate of Rap is slow, with a half-life on the order of hours, GAPs accelerate the GTP hydrolysis rate and play a key role in limiting how long Rap remains in its active form. GAPs that inactivate Rap include RapGAPI, RapGAPII, Spa1, and SIPA1L1/E6TP1 [246]. All of these GAPs are expressed in B cells (Immgen database). Whether these different Rap GEFs and GAPs act on all five of the Rap proteins is not fully understood and in a cellular context may depend on their expression levels.   77 1.7.4 Rap GTPases localize to cellular membranes The trafficking and subcellular localization of the Rap GTPases are important regulatory mechanisms that control their subcellular activation and functions. Rap1 undergoes a series of posttranslational modifications that target it to different membranes and organelles. Prenylation is a post-translational addition of a farnesyl or geranylgeranyl group to a cysteine residue in the CAAX motif (A = aliphatic, X = terminal amino acid) at the C-terminus of Rap proteins [260]. After prenylation, the three C-terminal residues (AAX) are cleaved off [260, 261]. These acyl groups act as lipid anchors that target Rap proteins to different cell membranes [257]. Rap1 isoforms end in a CAAL motif that becomes geranylgeranylated [261, 262]. In addition, the subcellular localization of the Rap1 GTPases may also be controlled by a C-terminal polybasic region [261]. In contrast, Rap2 proteins lack a polybasic region, are farnesylated at the CAAX motif cysteine, and are palmitoylated at other cysteine residues that lie N-terminal to the CAAX box [261]. These differences in lipid modifications and C-terminal motifs between the Rap1 (geranylgeranylation and polybasic region) and Rap2 GTPases (farnesylation and palmitoylation) may target Rap1 and Rap2 proteins to different membranes or membrane domains within the cell so that they can carry out distinct functions.  In addition to being localized at the plasma membrane, Rap1 is also found in vesicular compartments that are specialized pools of membrane, which can be rapidly fused with the plasma membrane during processes such as neutrophil degranulation [263]. Other studies have detected Rap proteins in the Golgi, endosomes, perinuclear regions of fibroblasts and epithelial cells [264, 265]. In response to external polarity signals, Rap proteins often assume a polarized distribution within the cell. For example, Rap1 is asymmetrically localized to the leading edge of 78 migrating cells [70, 76, 266]. In epithelial cells, Rap2a is enriched at the apical face of the cell where it controls the polarized formation of a brush border [267].   1.7.5 Rap1 regulates cell adhesion and actin cytoskeleton remodeling Activated Rap1 binds multiple effector proteins that regulate cell adhesion, cytoskeletal dynamics, and cell polarity. Most Rap effectors interact with Rap via its RA or RBD domains. The major Rap effectors are RAPL (also called NORE1), Riam, AF-6, and Krit1, the RacGEFs Tiam1 and Vav2, and the RhoGAPs RA-RhoGAP and Arap3. Rap1 is an important component of the inside-out signaling that links receptor signaling to integrin activation in order to direct cell adhesion and migration. Integrin activation by the TCR, BCR, other immune cell activating receptors, integrins, G protein-coupled receptors (e.g. chemokine receptors), cytokine receptors, and TNF receptor family members all depend on Rap activation [268, 269]. Activated Rap recruits the effectors RapL, RIAM, and PKD to the cytoplasmic domains of integrins to form an “integrin activation complex” [254]. In T cells, activated Rap1 binds to RIAM, which can direct this complex to the cytoplasmic tails of β1 and β2 integrins [270] and then activate them by recruiting talin [271]. Activated Rap1 promotes the binding of talin to RIAM, which exposes an integrin-binding site in talin [271]. RIAM also binds to adhesion- and degranulation-promoting adapter protein (ADAP) and SKAP-55, both of which facilitate integrin activation by recruiting Rap1 to the T cell IS [272]. In addition to activating integrins via RIAM, activated Rap1 also binds to RapL after TCR or chemokine stimulation. RapL mediates activation of the LFA-1 integrins by recruiting Mst1 to LFA-1 [273]. RapL also interacts with Rap2b to regulate T cell migration independent of its role in mediating Rap1-dependent integrin adhesion [274].  79 Rap1 regulates actin cytoskeleton dynamics and organization in multiple ways. Rap1 can act via GEFs and GAPs for other GTPases that regulate actin dynamics. For example Rap can regulate Rac-dependent cell spreading by binding to the RacGEF Vav2 [275]. Other Rap effectors include the RacGEF Tiam and the RhoGAPs Arap3 and Ra-RhoGAP [254]. The Rap1 effector RIAM can also recruit Ena/VASP and profilin to regulate actin remodeling at the leading edge of migrating cells [270]. Additionally, Rap1 activation is important for controlling cell migration by linking integrin signaling to the formation of polarized actin structures at the leading edge of the cell [276]. Moreover, Rap is important for regulating actin turnover and dynamics at the leading edge, which controls 2D and 3D cell migration [276]. Activated Rap1 also increases the actin-severing activity of cofilin (see section 1.5.5), which we have previously shown is important for regulating actin-mediated B cell spreading and IS formation [55].  Activated Rap is also important for maintaining cell:cell junctions. The Rap effector AF6 acts in a Rap-dependent manner at adherens junctions to prevent E-cadherin internalization [277]. Another junctional effector protein is Krit1, which has several domains that allows it to interact with both Rap and Ras [254]. The Rap-Krit1 pathway stabilizes junctional integrity in endothelial cells by suppressing the formation of actin stress fibers and opposing thrombin-induced permeability of endothelial junctions [278].   1.7.6 Rap activation and cell polarity External polarity cues from adhesion receptors, antigen receptors, and chemokine receptors cause localized activation of Rap1. The polarity-regulating role of Rap1 was first identified in the budding S. cerevisiae yeast, where Rap1 is important for determining the site of bud formation relative to previous bud sites. During axial budding, the new yeast bud forms next 80 to the previous bud scar. Bud-forming proteins are recruited to that site and the actin cytoskeleton is locally reorganized [246]. The S. cerevisiae homologue of Rap1, Bud1p/Rsrl1p, is important for determining budding polarity in yeast. The loss of Bud1p in yeast leads to buds forming at random sites [279]. This indicates that Bud1 is important for determining cell polarity (i.e. bud site selection) but is not required for the assembling the machinery required for budding. Bud1 binds to the scaffold protein Bem1p, as well as Cdc24p, which is a GEF for the Rho GTPase Cdc42 [280]. The Rsr1p/Bud1p bud site selection machinery recognizes Bud3p, Bud4p, Axl1p and Axl2p, proteins that mark previous bud sites [281]. The formation of this Bud1 complex is important for positioning the bud site and creates internal polarity cues by promoting localized activation of Cdc42. Cdc42 functions downstream of Bud1 to establish cell polarity by recruiting actin polymerization machinery and septin ring formation machinery to the bud site [29, 246]. Interestingly, loss of either Bud1p GEF (Bud 5) or GAP (Bud2) results in random bud site selection [282, 283], suggesting that the cycling of Rap between its active and inactive states is important for its polarity-determining function. The polarity-determining function of Rap1 is evolutionarily conserved as Rap1 also controls cell polarity in mammalian cells. Rap1 is found at the leading edge of migrating lymphocytes [284] and is required for establishing chemokine-induced lymphocyte polarity [243, 284]. Additionally, Rap1 is important for determining axon polarity in neurons [244]. Rap1 is also important for establishing polarity in migrating tumour cells. In metastatic melanoma cells, Rap controls the polarized assembly of the actin cytoskeleton at the leading edge where new focal adhesions form [276]. Moreover, the cycling of Rap is important for regulating this polarized cell morphology as inactivating Rap and expressing a constitutively active form of Rap both induce the formation of stable actin structures and adhesions that are distributed throughout 81 the cell [276]. Importantly, this loss of Rap-mediated polarity correlates with impaired tumour cell invasion into 3D matrices and decreased metastasis [276]. Rap1 can also work in concert with, or upstream of, other polarity regulators including Cdc42 and Rac1 to establish cell polarity. In migrating T cells, Rap1 acts upstream of Cdc42 activation to promote the localization of the Par polarity complex to the leading edge of the cell and away from the uropod [284]. The authors of this study propose that activated Rap1 binds to the RacGEF Tiam1 and co-localizes Rap1-dependent Cdc42 activation with Tiam1-dependent Rac activation [284].   1.7.7 Rap1 in B cells We and others have shown that the Rap GTPases controls B cell adhesion, migration, cell polarity, BCR microcluster formation, IS formation, and BCR-induced reorganization of the actin cytoskeleton [68, 76, 243, 247-251, 276, 284, 285]. Signaling by the BCR, integrins, and chemokine receptors can all activate the Rap1 and Rap2 GTPases in B cells [68, 76, 247-249, 251, 276]. BCR-dependent Rap activation requires PLCγ2 and can be stimulated by the PLCγ2 product DAG [252] (Figure 1.21). The RapGEF RasGRP2 has been implicated in BCR-induced activation of Rap1 [286]. DAG is thought to activate PKC, which then phosphorylates RasGRP2 on sites that are critical for its activation [246, 286]. Although RasGRP2 contains a C1 domain, which was thought to bind DAG and mediate BCR-induced recruitment of RasGRP2 to the plasma membrane, the C1 domain of RasGRP2 does not to bind to DAG [286]. Instead, the localization of RasGRP2 at the plasma membrane may depend on F-actin [287], which can bind to the amino terminus of RasGRP2 [286].  In vitro, Rap activation is important for CXCL12- (also called SDF-1), CXC13-, and S1P-induced B cell migration and integrin-mediated adhesion [75, 251]. In vivo, Rap1 is 82 important for B cell homing to SLOs [288]. Rap1a knockout mice exhibit mild defects and immune functions are largely normal. B cells from Rap1a knockout (KO) mice exhibit decreased integrin-mediated adhesion to fibronectin and ICAM-1-coated surfaces [285], as well as decreased chemotaxis to homeostatic chemokines [289]. More striking defects in B cell function are observed in mice lacking Rap1b, which is more abundant than Rap1a in B cells [255]. Rap1b-deficient B cells exhibit decreased adhesion to ICAM-1, impaired chemotaxis towards CXCL12 and CXCL13, decreased in vivo lymph node homing, and defective T-dependent antibody responses [255].  Rap activation is required for lymphocyte polarization in response to chemokines and antigens [55, 245, 247]. In addition to its critical role in T and B cell chemotaxis, Rap activation is important for the formation of an IS in response to polarized arrays of antigen such as those presented on APCs, lipid bilayers, or beads. We have previously shown that activated Rap localizes to the contact site with anti-Ig-coated beads and that Rap activation is important for the development of a polarized actin morphology in which F-actin forms a ring around the bead contact site [247] or at the periphery of the B cell:APC contact site [55]. This Rap-dependent actin reorganization is important for B cells to spread on anti-Ig-coated surfaces and for optimal BCR signaling in response to particulate antigens but not to soluble antigens [247]. Our lab showed that the Rap-cofilin pathway drives B cell IS formation by controlling the remodeling of the actin cytoskeleton [55, 247]. Importantly, blocking Rap- and cofilin-dependent actin remodeling substantially reduced both BCR microcluster formation and BCR signaling initiated by APC-associated antigens [55].  Dysregulation of Rap1 may contribute to B cell-mediated autoimmunity and to the spread of malignant B cells. The loss of the Rap-specific GAP Spa-1 in mice results in the accumulation 83 of active Rap in B cells and these mice develop lupus-like nephritis due to autoantibody production by autoimmune B cells [290]. Rap activation is also important for the metastatic dissemination of B cell lymphomas in mice [76]. Lin et al. showed that blocking Rap activation in A20 B-lymphoma cells by overexpressing RapGAPII reduced their ability to undergo CXCL-12-induced directional migration as well as transendothelial migration in vitro [76]. Consistent with this finding, RapGAPII-expressing A20 cells that had been intravenously injected showed a greatly reduced ability to form tumors in the liver compared to control A20 cells [76]. Control lymphoma cells entered the liver parenchyma and formed large tumors whereas the RapGAPII-expressing cells remained in the vasculature, consistent with their inability to undergo transendothelial migration in vitro [75, 76].      84   Figure 1.21. BCR-mediated activation of Rap GTPases  BCRs that bind to cognate antigen initiate signaling, leading to the activation of PLCγ2, which cleaves the phospholipid PIP2 to form the second messenger DAG. DAG is important for the activation of the RasGRP2 GEF, which converts inactive GDP-bound Rap to its active form by facilitating the exchange of GDP for GTP. Rap, which is localized at the plasma membrane, then recruits a number of effectors that regulate cell polarity, actin remodeling, and integrin-mediated adhesion. Rap is inactivated by GAPs that increase the rate of GTP hydrolysis and convert Rap to its inactive GDP-bound form. See text for details. 85 1.8 Hypothesis and Specific Aims Based on previous work on the role of the MTOC during IS formation and function, and the roles of Rap in establishing cell polarity by regulating the cytoskeleton, my overall hypothesis is that:  The Rap GTPases coordinate actin cytoskeleton remodeling and MTOC reorientation during IS formation, and that this is important for antigen acquisition at the IS.   To test this hypothesis, I had the following specific objectives to determine if: 1. The Rap1-cofilin pathway coordinates actin cytoskeleton remodeling with MTOC reorientation during IS formation in B cells.  2. IQGAP1 and CLIP-170 bridge actin remodeling to MTOC polarization in B cells.  3. The Rap1-cofilin pathway is important for B cells to acquire antigens from APCs.  4. The HS1 protein is important for actin remodeling during IS formation in B cells.    86 Chapter 2: Methods 2.1 Cell isolation and culture 2.1.1 Primary B cell isolation and culture B cells were isolated from spleens of 6-12-week-old C57BL/6 or MD4 mice [291] (Jackson Laboratories, stock #002595) using a negative selection B cell isolation kit (Stemcell Technologies, catalogue #19854). The university animal care committee approved all protocols. B cells were cultured in RPMI-1640 supplemented with 10% heat-inactivated FCS, 2 mM glutamine, 1 mM pyruvate, 50 µM 2-mercaptoethanol, 50 U/ml penicillin, and 50 µg/ml streptomycin (complete RPMI medium). Primary B cells were used either immediately after isolation or were cultured for 6 hr in complete medium plus 5 µg/ml E. coli 0111:B4 LPS (Millipore Sigma, catalogue #L4391) prior to being transfected.   2.1.2 B cell lines and culture The A20 murine IgG+ B cell line, the RAMOS human IgM+ B cell line, and the WEHI-231 murine IgM+ B cell line were obtained from ATCC and cultured in complete RPMI medium. A20 cells expressing the HEL-specific D1.3 transgenic BCR [30] were a gift from F. Batista (Cancer Research UK, London, UK). A20 and WEHI-231 cells stably transfected with pMSCVpuro or pMSCVpuro-FLAG-RapGAPII have been described [68] and were cultured in complete RPMI medium with 4 µg/ml puromycin. A20 cells that were stably transduced with retroviruses encoding HS1-YFP in the pMSCV2.1 vector (a gift from J. Burkhardt, University of Pennsylvania) were generated as described in section 2.2 and cultured in complete RPMI medium. A20 cells expressing shRNA for IQGAP1 and CLIP-170 were generated by lentiviral 87 transduction as described in section 2.3.2 and cultured in complete RPMI medium with 4 µg/ml puromycin. 2.1.3 T cell isolation and culture Murine T cells were isolated from the spleens of OT-II transgenic mice, which express a TCR specific for chicken ovalbumin 323-339 peptide in the context of I-Ab (Jackson Laboratories, stock #004194) using the EasySep™ Mouse CD4+ T Cell Isolation Kit (Stemcell Technologies, catalogue #19852) according to the manufacturer’s instructions.   2.2 Transfections and transductions 2.2.1 Transient transfection of B cell lines Plasmids encoding WT cofilin, cofilin S3D, or cofilin S3A fused to mCherry have been described [55]. The CLIP-170-GFP, IQGAP1-WT-GFP, IQGAP1-CT-GFP (amino acids 1503-1657 of IQGAP1) constructs [228] were gifts from K. Kaibuchi (Nagoya University, Nagoya, Japan). LifeAct [292] fused to mCherry was from Addgene (Cambridge, MA; plasmid #54491). F-tractin-tdTomato and F-tractin-GFP [81, 226] were gifts from J. Hammer (NIH). Actin-GFP was a gift from R. Nabi (UBC). A20 cells or RAMOS cells (2 x 106) were transiently transfected with 2 µg of plasmid DNA using AMAXA nucleofector kit V (Lonza, catalogue #VCA-1003). Cells were cultured in complete RPMI medium and used 24 hr after transfection.    88 Table 2.1 Plasmid constructs used to express proteins in B cells.  Gene Plasmid Name Origin F-actin GFP-F-tractin Gifts from Dr. John Hammer (NIH) tdTomato-F-tractin Lifeact-mCherry Addgene plasmid #54491 Actin-GFP Gift from Dr. Robert Nabi (UBC) Cofilin mCherry-WT cofilin [55] mCherry-cofilin S3D mCherry-cofilin S3A CLIP-170 CLIP-170-GFP Gifts from Dr. Kozo Kaibuchi (Nagoya University, Nagoya, Japan) IQGAP1 IQGAP1-CT-GFP IQGAP1-WT-GFP   2.2.2 Retrovirus-mediated transduction  Retroviral transduction of A20 cells with pMSCVpuro, pMSCVpuro-FLAG-RapGAPII, or pMSCV2.1-hHS1-YFP [142] was performed as described previously [293]. Briefly, BOSC23 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% FCS, 2 mM L-glutamine and 1 mM sodium pyruvate (complete DMEM medium) until 85% confluent. For transfection of the BOSC23 cells, the medium was replaced with complete DMEM medium containing 25 μM chloroquine (Millipore Sigma, catalogue #C6628). For CaCl2-mediated transfection, 200 μl of transfection buffer (50 mM HEPES, pH 7.05, 10 mM 89 KCl, 12 mM dextrose, 280 mM NaCl, 1.5 mM Na2HPO4·H2O) was added drop-wise while vortexing to a tube containing 200 μl of 250 mM CaCl2 and 2 μg of DNA plasmid, and then immediately added to the cells. After 8-10 hr the medium was replaced with fresh complete DMEM. The medium was changed with fresh DMEM again after 24 hr and the virus-containing supernatant was collected after 44 hr. The medium was filtered to remove cell debris and added to A20 cells (5 x 105 cells in one well of a 6-well dish) along with polybrene at a final concentration of 12.5 μg/ml. The plate was then centrifuged for 30 min at 1800 rpm (730 × g) at 21°C. The A20 cells were then cultured at 37°C. At 16 hr post infection, the virus-containing medium was removed and the A20 cells were resuspended in fresh complete RPMI medium. At 48 hr post infection, 4 µg/ml puromycin was added to the medium to select for cells transduced with MSCVpuro-containing constructs. pMSCV2.1-hHS1-YFP-transduced cells were sorted by flow cytometry for YFP+ A20 cells.  2.3 siRNA- and shRNA-mediated gene silencing 2.3.1 siRNA knockdown   Table 2.2 siRNAs  siRNA Manufacturer Catalogue Number Mouse Rap1a Dharmacon, GE Life Sciences L-057058-01-0005 Mouse Rap1b L-062638-01-0005 Mouse Rap2c L-055004-01-0005 Mouse HCLS1 (HS-1) L-046134-01-0005 Non-Targeting Pool D-001810-01-05   90 ON-TARGETplus siRNA SMARTpools were used for protein depletion studies. A20 cells or LPS-treated primary B cells (2 x 106) were transduced with 2 µg of HS1 siRNA, 2 µg each of Rap1a siRNA and Rap1b siRNA, 2 µg Rap2c or 2 µg control siRNA using AMAXA Nucleofector Kit V (Lonza). Before being used for experiments, the cells were cultured for 18 hr in complete RPMI medium with the addition of the pro-survival cytokine BAFF (5 ng/ml; R&D Systems, Minneapolis, MN; catalogue #2106-BF-010) for primary B cells.   2.3.2 Lentivirus-mediated expression of shRNAs  Table 2.3 shRNAs  shRNA Vector Manufacturer Catalogue Number Mouse IQGAP1 pGipZ Thermo Fisher  #V3LMM_426631 Mouse IQGAP1 #V3LMM_426629 Mouse CLIP-170 #RMM4431-200324779 Mouse HCLS1 pLKO.1 Millipore Sigma #TRCN0000103625 Mouse HCLS1 #TRCN0000103626 Mouse HCLS1 #TRCN0000103629 TRC2 pLKO.5-puro Non-Mammalian shRNA Control Plasmid DNA #SHC202  The pGipZ-based plasmids encode both GFP and puromycin resistance in addition to the shRNA construct. The pLKO.1-based plasmids encode puromycin resistance. These plasmids were co-transfected with pCMV-VSV-G-M5 and pCMV-δR8.91 (a gift from D. von Laer, Medical University of Innsbruck, Innsbruck, Austria) into HEK293T cells (ATCC). Virus particles were collected at 12 hr and 36 hr post-transfection and added to 12-well plates containing A20 cells (2 x 105 per well), which were then centrifuged at 2000 rpm (929 x g) for 1 hr at 21°C. The cells 91 were then cultured for 48 hr before being resuspended in medium containing 4 µg/ml puromycin. When pGipZ-based plasmids were used, populations of GFP-expressing cells were isolated by FACS sorting.  2.4 Immunoblotting Cells were lysed in RIPA buffer (30 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Igepal, 0.5% sodium deoxycholate, 0.1% SDS, 2 mM EDTA, 1 mM phenylmethylsulphonyl fluoride, 10 μg/ml leupeptin, 1 μg/ml aprotinin, 25 mM β-glycerophosphate, 1 μg/ml microcystin-LR, 1 mM Na3VO4,) and analyzed by immunoblotting as described [55]. The antibodies used are described in Table 2.1. Bands were visualized using ECL (GE Life Sciences).    92 Table 2.4  Antibodies for immunoblotting  Antibody Name Manufacturer Catalogue Number Dilution Rabbit anti-IQGAP1 Santa Cruz sc-10792 1:1000 Rabbit anti-CLIP-170 sc-25613 1:3000 Mouse anti-β-Actin (C4) sc-47778 1:5000 Mouse anti-Rap 2a/b/c sc-136138 1:500 Rabbit anti-Rap1a/Rap1b Cell Signaling Technologies 4938 1:1000 Rabbit anti-HS1  (rodent specific) 4557 1:1000 Rabbit anti-phospho-HS1 (Tyr 397) 4507 1:1000 Horseradish peroxidase-conjugated goat anti-rabbit IgG Bio-Rad 170-6515 1:3000 Horseradish peroxidase-conjugated goat anti-mouse IgG 170-6516 1:3000  2.5 Assessing BCR-mediated endocytosis using flow cytometry A20 B cells (1 x 106) were incubated with 40 μg/ml goat anti-mouse IgG (Jackson ImmunoResearch, catalogue #115-005-008) for 30 min on ice, washed 3 times with ice cold PBS, and then incubated at 37°C for the indicated times. Following incubation, cells were immediately chilled and immunostained on ice for 30 min using Alexa 488 rabbit anti-goat IgG (1:300). After staining, the cells were resuspended in FACS buffer (PBS with 2% FCS and 0.1% sodium azide) containing 25 ng/ml DAPI (Life Technologies, catalogue #D1306) and analyzed by flow cytometry.   93 2.6 Cytoskeletal inhibitors 2.6.1 Pharmacological inhibitors B cells were resuspended in the assay buffer (as indicated in each experimental protocol) and incubated for 5 min at 37oC with latrunculin A, nocodazole or paclitaxel before being added to anti-Ig-coated coverslips, anti-Ig-coated beads, or APCs. To inhibit the Arp2/3 complex, B cells were resuspended in RPMI-1640 and incubated with CK-666 [294], or the control compound CK-689, for 30 min at 37oC. To inhibit PKCζ, B cells were resuspended in PBS with the 20 μM PKCζ-pseudosubstrate [85] for 1 hr at 37oC.   Table 2.5 Inhibitors targeting cytoskeletal regulators  Inhibitor Manufacturer Catalogue Number Target Final Concentration Latrunculin A Enzo Life Sciences BML-T119 Actin 1 μM CK-666 Millipore 182515 Arp2/3 complex 100 μM CK-689 182517 Control for CK-666 100 μM Nocodazole Millipore Sigma M1404 Microtubules 1-5 μM Paclitaxel Millipore Sigma T7402-1MG 5 μM PKCζ pseudosubstrate Enzo Life Sciences ALX-260-155-M001 PKCζ 20 µM    94 2.6.2 Cofilin inhibitory peptides Cell-permeable peptides that inhibit F-actin severing by cofilin [295] were synthesized by Biopeptide Inc. (San Diego, CA) and used as described previously [56]. The M and W peptides contain an actin-binding sequence from cofilin coupled to penetratin. In the control Q peptide, the sequence in the W peptide was changed to QSQM to ablate F-actin binding. B cells (2.5 x 105) were resuspended in 200 μl cold RPMI-1640 and incubated with 5 µM each of the M and W peptides, or with 5 µM of the Q peptide, for 1 hr on ice. The cells were then warmed to 37oC for 5 min before being used in experiments.  Table 2.6 Cofilin inhibitory peptide sequences  Peptide Peptide Sequence Q control CDYKDDDDKWAPESAPLQSQM M CDYKDDDDKMASGVAVSDGVIK W CDYKDDDDKWAPESAPLKSKM  See text for details. Bold text shows sequence differences between the Q and W peptides.   2.7 B cell interactions with anti-Ig-coated beads Polystyrene beads (4.5 µm diameter; Polysciences, Warrington, PA) were coated with goat anti-mouse IgM (Jackson ImmunoResearch, West Grove, PA; catalogue #115-005-020) or goat anti-mouse IgG as described [247]. Amino Beads (3 μm diameter; Polysciences) were activated overnight at 4°C with 8% glutaraldehyde before adding 20 µg Alexa 647 goat anti-mouse IgM (Jackson ImmunoResearch catalogue #112-545-175) to 5 x 107 beads for 4 hr at room temperature. B cells (2 x105) were resuspended in modified HEPES-buffered saline (mHBS; 25 mM HEPES, pH 7.2, 125 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM Na2HPO4, 0.5 mM 95 MgSO4, 1 mg/ml glucose, 2 mM glutamine, 1 mM sodium pyruvate, 50 µM 2-mercaptoethanol) and mixed with 106 beads at 37°C. The cells were then adhered to poly-L-lysine (PLL)-coated coverslips and fixed with 4% PFA for 10 min at room temperature, or with ice-cold methanol for 10 min at -20oC for pericentrin staining. For live cell imaging at 37oC, cells were added to coverslips coated with 2 µg/ml anti-MHC II antibodies (Millipore, catalogue #MABF33) and imaged by spinning disk confocal microscopy, or added to poly-D-lysine-coated glass bottom #1.0 dishes (MatTek, Ashland, MA) and imaged by laser scanning confocal microscopy.  2.8 B cell spreading on planar surfaces 2.8.1 B cell spreading on coated coverslips For spreading on anti-Ig-coated surfaces, coverslips were coated with 2 µg/cm2 goat anti-mouse IgG or goat anti-human IgM (Jackson ImmunoResearch; catalogue #109-005-043) as described [247]. Alternatively, coverslips were coated with 0.1 µg/ml PLL or 2 µg/ml anti-MHC II antibodies. B cells (104) were resuspended in mHBS and added to the coverslips. After various times at 37°C, the cells were fixed with 4% PFA for 10 min at room temperature, permeabilized, and stained as described in section 2.10. Alternatively, cells were transfected with fluorescent proteins to visualize tubulin or F-actin (described in section 2.10.2) and imaged in real time at 37oC using spinning disk confocal microscopy or TIRF microscopy. Image analysis and fluorescence quantification were performed using ImageJ software (NIH).  2.8.2 Mechanical stretching of B cells A20 cells were subjected to mechanical stretch as described previously for tumor cells [296]. Briefly, the A20 cells were added to 6-well flexible silicone rubber-bottom BioFlex culture 96 plates (Flexcell International Corporation, catalogue # BF-3001U) that had been coated with 2 μg/cm2 fibronectin (FN) to promote adhesion. The plates were placed on the BioFlex 25-mm loading posts of a FlexCell apparatus and the cells were subjected to equibiaxial stretch for 5 min by applying a 42.5 kPa vacuum pressure. The cells were then fixed and stained for α-tubulin and F-actin.  2.9 B cell:APC interactions 2.9.1 Antigen-presenting cells Cos-7 cells (ATCC) were grown in complete DMEM. Lipofectamine 2000 (Invitrogen) was used according to the manufacturer’s protocol for the transient transfection of Cos-7 cells with a plasmid encoding either a transmembrane form of HEL fused to GFP (mHEL) [66] or a transmembrane form of a single-chain anti-Igκ antibody [55, 297]. The cells were cultured for 18-24 hr before being detached using enzyme-free cell dissociation buffer (0.5 mM EDTA, 100 mM NaCl, 1 mM glucose, pH 7.4).    2.9.2 B cell interactions with APCs in suspension To assess MTOC reorientation or the acquisition of membrane-bound antigens by B cells, 2 x105 Cos-7 APCs expressing either mHEL antigen or the single chain anti-Igκ antibody were mixed in suspension with 4 x 105 HEL-specific or wildtype B cells, respectively, in 200 μl mHBS with 2% FCS (mHBS-FCS). After various times at 37°C, the cells were pipetted onto glass coverslips coated with 0.01% PLL, fixed with 4% PFA for 10 min at room temperature, permeabilized with 0.2% (W/V) Triton X-100, immunostained, and imaged by spinning disk confocal microscopy. Primary B cells that were used to assess antigen acquisition were pretreated for 6 hours with either 5 ng/ml BAFF or with 2.5 μg/ml of LPS prior to mixing with APCs.  97 2.9.3 B cell interactions with adherent APCs For experiments using adherent APCs, HEL-GFP- or anti-Igκ-expressing Cos-7 cells were resuspended in complete DMEM medium and added for 4-6 hr at 37oC to 18 mm coverslips that had been coated with 5 μg/ml FN. B cells were stained with CellMaskTM Far Red (Thermo Fisher; see Table 2.9). After allowing Cos-7 cells to spread, the DMEM medium was replaced with mHBS-FCS. B cells were then added to the APCs at different times and then fixed with PFA for immunostaining. For live cell imaging with HEL-GFP-expressing Cos-7 cells, 105 HEL-specific B cells were added to the APCs. The cells were then imaged in real time at 37oC by spinning disk confocal microscopy (see section 2.11.1 below).   2.9.4 T cell activation by APC-activated B cells Primary B cells from MD4 mice were isolated from spleens and transduced with Rap1a/b siRNAs or the control siRNA as described above in section 2.3.1. Primary T cells from OT-II mice were isolated from spleens as in section 2.1.3. Cos-7 cells expressing mHEL-GFP were allowed to adhere and spread on a tissue culture-treated dish for 4 hr before the adding B cells and culturing them with 5 ng/ml BAFF overnight. OT-II T cells were stained with 5(6)-Carboxyfluorescein N-hydroxysuccinimidyl ester CFSE (Thermo Fisher, catalogue #C1157, 1:1000 dilution), and added to the B cell:APC co-culture. After an additional 48 hr, the co-cultured cells were pelleted and then resuspended in FACS buffer (PBS with 0.5% FCS, 2 mM EDTA, 0.1% sodium azide). Cells were stained with the LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit (Thermo Fisher, catalogue #L34957) according to the manufacturer’s instructions before being fixed and permeabilized using Fixation/Permeabilization Solution (BD Biosciences, catalogue #554722) according to the manufacturer’s instructions. Fc receptors were blocked by 98 including 25 µg/ml of the 2.4G2 monoclonal antibody in the FACS buffer. The cells were then immunostained for mouse IL-2 (PE-conjugated, Thermo Fisher Scientific, catalogue #12-7021-41), CD3, and CD19 and analyzed by flow cytometry.   2.10 Cell staining for fluorescence microscopy 2.10.1 Immunostaining for fluorescence microscopy Cells that had been fixed onto coverslips were permeabilized for 3 min at room temperature with 0.2% (W/V) Triton X-100 in PBS, and then blocked with 2% bovine serum albumin (BSA) in PBS. Antibodies were diluted in PBS with 2% BSA and 0.1% (w/v) Triton X-100. Coverslips were incubated with primary antibodies (Table 2.7) for 40 min at room temperature, washed with PBS, and then incubated with secondary antibodies (Table 2.8) for 40 min at room temperature. Where indicated, F-actin was stained using fluorophore-conjugated phalloidins (Molecular Probes-Invitrogen, 1:300;Table 2.9). Coverslips were washed and then mounted onto glass slides using ProLong Diamond anti-fade mounting reagent containing DAPI (Molecular Probes-Invitrogen, catalogue #P36962).     99 Table 2.7 Primary antibodies used for immunostaining  Antibody type Target Species Manufacturer Catalogue Number Dilution Primary anti-α-tubulin Rabbit Abcam  ab52866 1:250 anti-α-tubulin Rat ab6161 1:350 anti-pericentrin Rabbit  ab4448 1:500 anti-Lamp1 Rabbit ab24170 1:300 anti-IQGAP1 Rabbit Santa Cruz sc-10792 1:250 anti-GFP Mouse Invitrogen A11120 1:400 anti-P-CD79a (Tyr 182) Rabbit Cell Signaling Technologies 5173S 1:250 anti-Phospho-HS1 (Tyr 397) Rabbit 4507 1:200 anti-EEA1 Rabbit 3288s 1:250    100 Table 2.8 Secondary antibodies used for immunostaining  SecondaryAntibody  Antibody Name Species Manufacturer Catalogue Number Dilution  Alexa 488 anti-rabbit IgG Goat Molecular Probes-Invitrogen A-11034 1:250 Alexa 568 anti-rabbit IgG A-11036 Alexa 647 anti-rabbit IgG A-21244 Alexa 488 anti-rat IgG A-11006 Alexa 647 anti-rat IgG A-21248 Alexa 568 anti-mouse IgG A-11031 Alexa 647 anti-mouse IgG A-21235 Alexa 647 anti-mouse IgM A-21238 Alexa 647 anti-goat IgG Donkey A-21447 Alexa 488 anti-rabbit A-21206   2.10.2 Fluorescent stains for visualizing cellular structures  Membrane stains were used to visualize the plasma membrane of B cells during live cell imaging experiments. LysoTracker® Red DND-99 (Thermo Fisher) were used for visualizing lysosomes. For CellMaskTM (Thermo Fisher) membrane stains, 106 cells were suspended in a 1:1000 dilution 101 of the reagent in PBS and incubated at 37oC for 10 min. The reaction was stopped by adding 5 ml of cold PBS with 2% FCS. After pelleting the cells, the cells were then washed 3 times with PBS with 2% FCS before being resuspended to 106 cells/ml in mHBS with 2% FCS. For staining with CELLVUE maroon membrane dye (Polysciences), 2 x 107 cells were suspended in 2 ml of Diluent C containing 4 μM of the dye for 2-5 min before adding 2 ml of PBS with 1% BSA and leaving the cells at room temperature for 1 min to stop the staining reaction. The cells were then washed 3 times with complete RPMI medium and resuspended to 106 cells/ml. The cells were kept on ice for up to 1 hr post staining. Staining F-actin with phalloidin was performed after fixing and permeabilizing the cells (see section 2.10.1, experimental procedures for immunostaining for fluorescence imaging).   Table 2.9  Cellular stains used for detecting cellular components  Cellular Stain Target Fluor Manufacturer Catalogue number Dilution CellMaskTM Deep Red Plasma Membrane Deep Red (far red) Thermo Fisher C10046 1:2000 CELLVUE Maroon Plasma Membrane Far red Polysciences 24847-1 4 µg/ml LysoTracker® Red DND-99 Acidic organelles (e.g. lysosomes) DND-99 (red) Thermo Fisher L7528 1:20000 Phalloidin F-actin Alexa 488 A12379 1:100-1:300 Phalloidin Alexa 532 A22282 Phalloidin Rhodamine R415 Phalloidin Alexa 647 A22287 102 2.10.3 Immunostaining for STED super resolution microscopy Cells that had spread on anti-Ig-coated coverslips were fixed for 10 min with 3% PFA plus 0.1% glutaraldehyde, then blocked and permeabilized for 10 min at room temperature in blocking buffer (PBS containing 3% BSA and 0.1% Triton X-100). Primary (Table 2.7) and secondary antibodies (Table 2.8; 1:100 dilution of each in blocking buffer) were added to the coverslips for 30 min at room temperature. Alexa 532-phalloidin (1:100) was added to the secondary antibody solution. Coverslips were mounted onto glass slides using ProLong Diamond anti-fade reagent (Molecular Probes-Invitrogen, catalogue #P36961).   2.10.4 Immunostaining for ground state depletion super-resolution microscopy imaging of microtubules Cell fixation, permeabilization, and staining procedures for ground state depletion (GSD) imaging of microtubules was carried out as described [298]. Briefly, A20 cells that had been allowed to spread on anti-IgG-coated #1.5 coverslips were fixed with PBS containing 3% formaldehyde and 0.1% glutaraldehyde for 10 min at room temperature. After quenching with PBS containing 0.1% sodium borohydride for 7 min, the cells were permeabilized with permeabilization/blocking buffer (3% BSA, 0.5% Triton X-100 in PBS) for 10 min. The samples were then stained with rat anti-α-tubulin for 30 min, followed by Alexa 647-goat anti-rat IgG, both of which were diluted 1:100 in permeabilization/blocking buffer. Each antibody staining step was carried out for 30 min and followed by 3 washes with PBS containing 0.2% BSA and 0.1% Triton X-100. Post-staining fixation was carried out for 10 min at room temperature using PBS containing 3% formaldehyde and 0.1% glutaraldehyde. The imaging medium consisted of 50 mM TrisHCl, pH 8.0, 10 mM NaCl, 10% w/v glucose, 143 mM 2-mercaptoethanol, and the 103 GLOX enzymatic oxygen scavenging mixture (1% v/v). GLOX was prepared by adding 10 mg glucose oxidase (Millipore Sigma, catalogue #G2133) and 50 µl catalase (Millipore Sigma, catalogue #C100) to 200 µl PBS and centrifuging the mixture at 13,000 rpm (11337 x g) for 1 min. 2.10.5 Proximity ligation assays The Duolink® In Situ Red Starter Kit (Millipore Sigma), which included donkey anti-mouse IgG and donkey anti-rabbit IgG antibodies coupled to complementary oligonucleotides, was used according to the manufacturer’s instructions for detecting protein-protein interaction sites by PLA.  2.11 Microscopy and image analysis 2.11.1 Confocal microscopy  Laser scanning confocal microscopy was performed using an Olympus IX81/Fluoview FV1000 confocal microscope based on an IX81 inverted microscope with a 100x NA 1.40 oil objective. Fluoview v3.0 and v4.0 software (Olympus) were used to analyze images and generate 3D reconstructions. ImageJ was used to quantify fluorescence intensities. Spinning disk confocal microscopy was performed using a system based on a Zeiss Axiovert 200M microscope with a 100x NA 1.45 oil objective and a QuantEM 512SC Photometrics camera for image acquisition (Quorum Technologies). For fixed cells, z-stacks were acquired in 0.3 µm increments. For live cell imaging, z-stacks were acquired at 0.5 μm increments every 12 seconds. Slidebook v5.5 and v6.0 software (3i Inc., Denver, CO) was used to analyze images and generate 3D reconstructions.   104 2.11.2 TIRF microscopy TIRF images were acquired at a 100 nm penetration depth using an Olympus cellTIRF 4-line microscopy system consisting of an Olympus IX83 Dual Deck motorized inverted microscope, a 100x NA 1.49 oil objective, and a Photometrics Evolve EM-CCD camera. MetaMorph software (Molecular Devices, Sunnyvale, CA) was used to acquire images.   2.11.3 STED microscopy Sample staining is described above in section 2.10.3. STED super-resolution microscopy was performed using a TCS SP8 laser scanning STED system (Leica) equipped with the 592 nm and 660 nm depletion lasers, a CX PL APO 100x NA 1.40 oil objective, and HyD high-sensitivity detectors (Leica). Acquisition was performed on the LASX software (Leica). Time-gated detection was set from 0.3–6 ns. Image deconvolution was performed using Huygens software (Scientific Volume Imaging, Hilversum, Netherlands).  2.11.4 Ground state depletion microscopy GSD/direct stochastic optical reconstruction microscopy (dSTORM) imaging of microtubules in fixed cells (see Section 2.10.4 above for the immunostaining protocol for GSD) was performed using the GSD followed by individual molecule return (GSDIM) microscope equipped with an EMCCD camera in TIRF mode set at 100 nm depth. Image deconvolution was performed using Huygens software (Scientific Volume Imaging, Hilversum, Netherlands).   105 2.11.5 Scanning electron microscopy B cells were mixed with anti-Ig-coated beads and then settled onto PLL-coated coverslips as in Section 2.7. The cells were then fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.37, and processed for scanning electron microscopy (EM) as described previously [56, 247]. After fixation, the samples were microwaved using a Pelco Biowave microwave processing system (Ted Pella, Redding, CA) at 100W under vacuum at 37ºC for four 2 min cycles, each separated by a 2-min rest period. The samples were then rinsed with 0.1 M cacodylate buffer pH 7.2 and microwaved at 22ºC without vacuum for 40 s at 300W, before being treated with 1% osmium tetroxide at 22ºC under vacuum. Samples were then microwaved again, rinsed with water and then microwaved a final time at 110W for 40 s. Samples were dehydrated by the sequential addition of 50%, 70%, and 95% ethanol and then three additions of 100% ethanol. Samples were microwaved under vacuum for 40 s at 110W after each addition. Coverslips were dried using a CPD 020 critical point dryer (Bal-Tec, Balzers, Lichtenstein) and covered with a ~25 nm layer of gold/palladium using a SEMPrep II sputter coater (Nanotech, Manchester, UK). Images were obtained using the Hitachi S-4700 field emission scanning electron microscope (University of British Columbia Bioimaging Facility).  2.11.6 Fluorescence recovery after photobleaching Fluorescence recovery after photobleaching (FRAP) was performed as described previously [55] using an Olympus FV1000 confocal microscope to image a 3.5-µm diameter circular region of interest (ROI). xy images of the confocal plane in which the cell was contacting the anti-Ig-coated coverslip were captured every 2-5 sec. After measuring the pre-bleach fluorescence signal, the ROI was photobleached using a 405-nm laser (100% intensity, 0.1 s). Fluorescence 106 recovery within the ROI was then imaged over 1 min. FluoView v1.6 (Olympus) software was used to quantify the fluorescence signal within the ROI, which was normalized to the pre-bleach intensity and the fluorescence background was subtracted using ImageJ.  2.12  Quantification of MTOC polarization The MTOC was visualized by immunostaining the tubulin network and defining the major point of microtubule convergence as the MTOC, or by staining for pericentrin, a component of the centrosome, which is the major MTOC in B cells. Confocal slices through the center of the B cell were used to assess the location of the MTOC relative to the contact site between the B cell and an anti-Ig-coated bead or an APC. For each B cell, the extent of MTOC polarization towards the contact site quantified by calculating a polarity index (PI) [85] as described in Figure 2.1. Distances were determined using Fluoview or ImageJ software.   107  Figure 2.1. Quantification of MTOC polarization  (A,B) The extent of MTOC polarization towards anti-Ig-coated beads was quantified by calculating a polarity index (PI). The PI is defined as the distance between the MTOC and the center of the bead (a, green) divided by the distance between the center of the B cell and the center of the bead (b, blue). PI <1 indicates that the MTOC is between the bead and the cell and PI <0.75 was selected as definitive polarization of the MTOC towards the bead. A random distribution of MTOC polarity indices would be reflected in an equal probability of the MTOC being at any location within the cell. To calculate the random probability that the MTOC PI in a given cell would be <1.0, <0.75, or <0.5, circles with their origin at the center of the bead (dotted lines) were drawn representing these different a/b ratios (B) and the area of overlap between a circle and the cell was calculated as according to the following formula (taken from http://mathworld.wolfram.com/Circle-CircleIntersection.html):   108 𝐴𝑟𝑒𝑎 = 𝑟2 𝑐𝑜𝑠−1(𝑏2 + 𝑟2 − 𝑅22𝑏𝑟) +  𝑅2 𝑐𝑜𝑠−1(𝑏2 + 𝑅2 − 𝑟22𝑏𝑅) − 12√(−𝑏 + 𝑟 + 𝑅)(𝑏 + 𝑟 − 𝑅)(𝑏 − 𝑟 + 𝑅)(𝑏 + 𝑟 + 𝑟)  where  r = radius of the circle representing a specific PI value = (PI)(b) b = the distance between the center of the cell and the center of the bead  R = radius of the cell R’ = radius of the bead  Two boundary conditions were considered, one in which the bead was just touching the edge of the cell (upper/left image in B) and one in which the center of the bead was at the edge of the cell (lower/right image in B). For the first condition, b = R + R’. For the latter condition b = R. The estimated values used for the diameters of the cells and beads were: 10.2 µm for A20 cells, 6.75 µm for primary B cells, and either 4.5 µm or 3 µm for the beads, depending on which were used. The arc cosine (i.e. inverse cosine) values were calculated in radians. The area of overlap between the PI circle and the cell was divided by the total area of the cell (πR2) to determine the percent of the cell area represented by the overlap area. For a cell population, this value is equal to the percent of cells that would randomly have their MTOC located within the area of overlap. The calculated values are shown in Table 2.10(for 4.5-μm beads) and Table 2.11 (for 3-μm beads). (C) The extent of MTOC polarization towards antigen-bearing APCs was quantified by calculating a PI that is defined as the distance along the z-axis between the MTOC and a plane created by the B cell:APC contact site (a) divided by the distance along the z-axis between the middle of the cell, i.e. the dashed line along the x-axis that passes through the center of the cell, and the B cell:APC contact site (b). A random distribution of MTOC localizations would be indicated by 50% of the cells having their MTOC being in the APC-proximal half of the cell (PI <1.0) and 37.5% of the cells having a PI ≤ 0.75.  (D) The extent of MTOC polarization towards anti-Ig-coated coverslips was quantified by calculating a PI that is defined as the distance along the z-axis between the MTOC and coverslip (a) divided by the distance along the z-axis between the middle of the cell and the coverslip (b). A random distribution of MTOC localizations would be indicated by 50% of the cells having their MTOC being in the lower half of the cell (PI <1.0), closer to the coverslip, and 37.5% of the cells having a PI ≤ 0.75.   109 Table 2.10 Percent of cells with polarity index that would randomly be ≤ 1.0, 0.75 or 0.5 when 4.5-μm beads were used  Polarity index 4.5-µm bead just touching 1o B cells (% of cells) 4.5-µm bead just touching A20 cells (% of cells) 4.5-µm bead with center at edge of 1o B cell or A20 cell (% of cells) 1.0 43.7 38.3 39.2 0.75 18.6 21.0 21.1 0.5 2.75 5.4 11.2  Values were calculated as described in Figure 2.1 using the formula in the figure legend. Table 2.11 Percent of cells with polarity index that would randomly be ≤ 1.0 or 0.75 when 3-μm beads were used.   Polarity index 3-µm bead just touching 1o B cells (% of cells) 3-µm bead with center at edge of 1o B cell (% of cells) 1.0 12.7 19.6 0.75 10.5 11.8 Values were calculated as described in Figure 2.1 using the formula in the figure legend.  2.12.1 Quantification of F-actin and IQGAP1 at bead:cell contact sites Fluorescence intensities were quantified using ImageJ. An 8.8-μm diameter circular region of interest (ROI) was drawn around the bead at the site of contact with the B cell and the total corrected fluorescence within the ROI was quantified as described [299] using the following equation: Integrated density – [(area of ROI) x (mean background fluorescence per unit area)], [299] where the integrated density is equal to [(area of ROI) x (mean fluorescence per unit area within the ROI)]. 110 2.12.2 Quantification of internalized antigen from APCs  Figure 2.2. Quantification of antigen internalization  Immunostaining was used to visualize the anti-Igκ surrogate antigen (Ag; green) as well as the periphery of the B cell, which was indicated by staining for the BCR (blue). For each B cell:APC conjugate, the amount of antigen that was inside the peripheral ring of BCR staining was quantified for each xy confocal slice through the B cell. As some antigen can remain on the B cell surface after being extracted from the APC, antigen fluorescence only within the peripheral ring of BCRs was included in the quantification as the amount of antigen internalized by the B cells. The sum of these values represents the total amount of internalized antigen (indicated in the figure as A). To take into account different densities of surrogate antigen presented on the surfaces of different APCs, the total amount of internalized antigen was normalized to the density of the surrogate antigen on the APC that the B cell was in contact with. The fluorescence intensity of the surrogate antigen in the first xy confocal slice through the APC was quantified and divided by the cell area in this confocal slice to yield the antigen density on the APC (indicated in the figure as B). The “normalized antigen internalization” is A divided by B. This measure of antigen internalization (uptake) excludes antigens that had been extracted from the APC but remained on the surface of the B cell because we could not rule out that these antigens were no longer associated with the APC.   111 2.13 Statistical analyses Student’s two-tailed paired t-tests were used to compare the mean values for matched sets of samples from multiple experiments. Unpaired t-tests were used to compare the means of data pooled from multiple experiments. Two-way ANOVA tests were used to compare mean differences between conditions that have been split on two independent variables. 112 Chapter 3: The Rap1-Cofilin-1 pathway regulates MTOC reorientation towards the B cell immune synapse   3.1 Introduction 3.1.1 Cytoskeletal reorganization drives B cell immune synapse formation In vivo, the differentiation of B-lymphocytes into antibody-producing cells is often initiated by APCs such as follicular dendritic cells. These APCs capture antigens and display them on their surface in an intact form that can be recognized by the BCR [8, 28, 38]. For such membrane-associated antigens, BCR-induced reorganization of the actin and microtubule cytoskeletons is critical for the two functions of the BCR, signal transduction and antigen internalization [29, 65, 300]. Initial BCR signaling at the B cell:APC interface promotes disassembly of the submembrane actin network and uncouples the actin meshwork from the plasma membrane [44, 53, 55]. This increases BCR mobility [44, 56] and allows antigen-bound BCRs to form microclusters that recruit signaling enzymes [9, 184, 301]. Concomitant actin polymerization at the cell periphery allows the B cell to spread across the antigen-bearing surface, encounter more antigens, and form additional BCR microclusters [64, 300]. Subsequent contraction of the B cell membrane [64] is accompanied by microtubule-dependent gathering of BCR microclusters into a cSMAC [190] that is characteristic of an immune synapse (IS) [213]. At the IS, B cells extract BCR-bound antigens from APCs [66]. These antigens are internalized and delivered to lysosomes and MHC II-containing vesicles via processes that involve both the actin and microtubule networks [29, 300]. The resulting peptide-MHC II complexes are presented to T helper cells, which then provide additional signals for B cell activation [29, 95]. Although complex changes in cytoskeletal organization play a key role in IS formation and 113 function, the mechanisms that coordinate these processes are not fully understood.   3.1.2 MTOC polarization supports immune synapse formation Reorientation of the microtubule network coordinates BCR organization and function at the IS and may therefore play an important role in B cell activation by mobile membrane-bound antigens. In response to localized BCR signaling initiated by antigen-bearing membranes or beads, the microtubule-organizing center (MTOC) moves towards the IS [29, 85, 220] such that microtubules extend along the inner face of the plasma membrane at the antigen contact site [190]. Dynein motor complexes that are recruited to antigen-bound BCRs then propel BCR microclusters along these juxtamembrane microtubules [190]. Dynein proteins move BCR microclusters centripetally towards the MTOC and concentrate the antigen-bound microclusters. As the antigen-bound BCR microclusters accumulate, this results in cSMAC formation [190]. Reorientation of the microtubule network also moves lysosomes and MHC II-containing vesicles towards the contact site so that extracted antigens can be efficiently delivered to these compartments [85, 222, 302, 303]. Thus, signaling pathways that link polarity cues initiating from the IS to direct MTOC reorientation are likely to be important for APC-mediated B cell activation.   3.1.3 Polarity proteins control the polarization of the MTOC  MTOC reorientation towards the IS is coordinated with remodeling of the actin cytoskeleton. MTOC polarization is preceded by the clearance of F-actin from the center of the IS and the accumulation of F-actin at the periphery of the IS [217, 304]. Microtubules associate with this peripheral ring of F-actin [211] and the movement of the MTOC towards the IS is then 114 driven by dynein motor complexes [210, 218, 226, 305]. Although the mechanisms that initiate the coordinated reorganization of the actin and microtubule networks at the IS are not fully understood, MTOC polarization towards the IS is controlled by evolutionarily conserved polarity proteins including the Cdc42 GTPase, PKCζ, and the Par polarity complex [85, 209, 217, 220, 227, 306, 307].  The Rap GTPases couple receptor signaling to cell polarity, cell adhesion, cell junction formation and remodeling of the actin cytoskeleton [244-246, 308]. The Rap proteins cluster into distinct Rap1 and Rap2 families. Rap1, which has been more extensively characterized than Rap2, plays a key role in establishing cell polarity. The yeast orthologue of Rap1, Bud1p/Rsr1p, controls bud site selection and establishes cell polarity by acting upstream of Cdc42 to initiate polarized actin polymerization and by promoting reorientation of microtubules towards the bud site [309-311]. Similarly, mammalian Rap1 acts upstream of Cdc42 to stabilize axons and establish neuronal polarity [244]. Rap1 also acts upstream of Cdc42 and the Par polarity complex to promote chemokine-induced polarization and directional migration in T cells [243, 284]. In B cells, Rap1 is activated by the BCR [252] and the active GTP-bound form of Rap1 accumulates in a polarized manner at the contact site with particulate antigens [247]. BCR-induced Rap1 activation is also required for the actin clearance and reorganization that is associated with IS formation [55, 247]. However, the role of Rap1 in MTOC polarization towards the IS has not been investigated and the role of Rap2 proteins in establishing cell polarity is not well understood.   115 3.1.4 The functions of Rap1 and Rap2 proteins Although the members of the Rap1 and Rap2 families have >90% sequence homology to each other, the sequence identity between the two Rap families is only 60-70%, suggesting that they could have different functions or regulation. The Rap GTPases function via effector proteins that bind to their active GTP-bound forms [254]. The switch regions of the Rap1 and Rap2 proteins, which are involved in effector binding, are highly conserved. However, amino acid changes in other regions of Ras and Rho family GTPases can cause dramatic changes in their function, at least with regard to cytoskeletal organization and cell morphology.[312] Taken together, this raises the possibility that the Rap1 and Rap2 proteins could have both overlapping as well as distinct functions. Overlapping functions could reflect the fact that Rap1 and Rap2 interact with many of the same effector proteins [254, 313]. In contrast, Rap1 and Rap2 have opposite roles in endothelial barrier function [314] due to the unique ability of Rap2 to bind several protein kinases that contain citron homology domains [315-317]. Similarly, Rap1 and Rap2 appear to have distinct roles in T cell adhesion and migration [274, 318]. Dissecting the roles of different Rap family members will help determine whether Rap proteins are important for MTOC polarization during IS formation.   3.1.5 Cofilin-mediated actin remodeling during B cell activation Cofilin is a member of the ADF/cofilin family of actin-severing proteins. Cofilin binds F-actin filaments and promotes actin severing in a non-enzymatic fashion by inducing a structural twist that destabilizes the interactions between actin monomers [160]. Cofilin-mediated F-actin severing promotes localized disassembly of the actin membrane-associated cytoskeleton while also creating new barbed ends where the Arp2/3 complex can nucleate branched actin 116 polymerization that drives cell spreading [55, 171, 319, 320]. The ability of cofilin to bind F-actin filaments and promote severing is regulated by its reversible phosphorylation on S3. Our lab showed that BCR engagement leads to the dephosphorylation and activation of the cofilin and that is dependent on Rap activation [55]. Importantly, we showed that Rap-dependent activation of cofilin is critical for the BCR-induced actin reorganization that is required for B cell spreading and IS formation [55].   3.1.6 Rationale and hypothesis IS formation is a cell polarization response that involves conserved polarity proteins such as Cdc42 and PKCζ [29, 222]. Rap1 is an evolutionarily conserved master regulator of cell polarity, which our lab has shown is essential for B cells to form an IS [55, 247]. Moreover, we showed that Rap1 acts via cofilin to induce the actin reorganization that drives B cell IS formation [55]. Reorientation of the microtubule network is also a critical element of cell polarization. BCR signaling causes reorientation of the MTOC towards polarized sites of antigen contact and the resulting juxtamembrane network of microtubules at the contact site is required for the centralization of BCR microclusters into the cSMAC of an IS [321]. Therefore, I tested the hypothesis that activation of the Rap-cofilin signaling pathway is required for BCR-induced MTOC reorientation towards the IS. I also investigated whether Rap1 and Rap2 proteins have distinct functions in BCR-induced MTOC reorientation. Using quantitative experimental systems for studying BCR-induced MTOC reorientation towards anti-Ig-coated beads or APCs expressing transmembrane antigens on their surface, I showed that Rap1, Rap2c, and cofilin are all required for BCR-induced MTOC polarization. Importantly, I showed for the first time that Rap1 coordinates actin remodeling and MTOC polarization at the B cell IS. 117 3.2 Results 3.2.1 Microtubules and Rap1 are required for cSMAC formation  When B cells bind artificial supported lipid bilayers that are embedded with antigens, BCR microclusters form at the periphery of the cell and then move centripetally along juxtamembrane microtubules to coalesce into a cSMAC [190]. To extend this finding to B cell:APC interactions, we performed real-time imaging of the contact site between A20 B-lymphoma cells expressing the hen egg lysozyme (HEL)-specific D1.3 BCR and Cos-7 surrogate APCs expressing a transmembrane form of HEL fused to GFP. Importantly, these APCs do not express other adhesion ligands (e.g. integrin ligands) for B cells as B cells do not form stable interactions with Cos-7 cells that do not express cognate antigen (data not shown). When we treated B cells for 5 min with the microtubule-disrupting drug nocodazole, which binds free tubulin heterodimer subunits and prevents them from polymerizing, BCR microclusters initially formed but did not coalesce into a cSMAC, as they did in control cells (Movie 1). By blocking Rap activation in B cell lines, our lab previously showed that activation of Rap GTPases is important for B cells to gather antigens into a cSMAC [55, 247]. I extended these findings to primary B cells from MD4 mice, in which all B cell express a HEL-specific BCR, by showing that siRNA-mediated knockdown of the two Rap1 isoforms, Rap1a and Rap1b (see Figure 3.5A), prevented BCR microclusters from coalescing into a cSMAC (Movie 2). Because silencing Rap1 impaired cSMAC formation, a process that depends on microtubules, we developed experimental systems for quantitatively assessing BCR-induced MTOC polarization and then tested whether Rap1 controls polarization of the MTOC and microtubule network towards the IS.     118   Movie 1. Microtubules are required for the centralization of antigen-bound BCRs into a cSMAC  A20 cells expressing the HEL-specific D1.3 BCR were stained with Cell Trace Far Red (blue) and pre-treated with either DMSO or 5 µM nocodazole for 5 min before being allowed to settle on adherent Cos-7 APCs expressing HEL-GFP (green). Time-lapse spinning disk confocal microscopy was used to acquire images every 12 s between 1 min and 12 min after adding the B cells to the APCs. The video is played back at 5 frames per second (60X real time).   blogs.ubc.ca/jiawangthesis/2016/05/10/chapter-2-movie-1/ Password: JWthesis    119   Movie 2. Rap1 is required for the centralization of antigen-bound BCRs into a cSMAC   Primary B cells from MD4 mice, which express a transgenic BCR specific for HEL, were transduced with control siRNA, or with Rap1a and Rap1b siRNAs. These cells were stained with Cell Trace Far Red (red) and added to adherent Cos-7 cells expressing HEL-GFP (green). Time-lapse spinning disk confocal microscopy was used to acquire images every 12 s between 1 min and 12 min after adding the B cells to the APCs. The video is played back at 10 frames per second (120X real time).   https://blogs.ubc.ca/jiawangthesis/2016/05/10/chapter-2-movie-2 Password: JWthesis   120 3.2.2 BCR clustering induces MTOC polarization towards anti-Ig-coated beads Antigens that are immobilized onto beads or planar coverslips, embedded in planar lipid bilayers or expressed on the surface of APCs have been used to create a polarized antigen contact site in order to model events that occur at the IS [55, 64, 85, 220, 247]. In all cases, B cells initially extend their plasma membrane across the antigen-coated surface (Figure 3.1). This requires Rap1-dependent reorganization of the actin cytoskeleton [55, 247]. In order for microtubules to orient juxtaposed to the IS membrane to act as tracks for the centripetal movement of antigen microclusters, reorientation of the MTOC must occur to bring the microtubule network to the cell:cell interface. To quantitatively assess the role of Rap1 in BCR-mediated reorientation of the MTOC, we initially used polystyrene beads coated with anti-Ig antibodies as a model of IS formation and to establish a well-defined polarized antigen contact site as previously described [85, 220, 247]. Importantly, this method of antigen presentation does not involve integrin ligands. Therefore, B cell adhesion to these antigen-presenting surfaces occur only via BCR-mediated antigen binding and allow us to study the BCR-induced regulation of the cytoskeleton in the absence of integrin signaling. We identified the MTOC by immunofluorescence as the point of microtubule convergence or by staining for the centrosomal protein pericentrin (see Figure 3.3). Real-time imaging of A20 cells expressing GFP-α-tubulin showed the rapid reorientation of the microtubule network, as well as movement of the MTOC to the antigen contact site (Movie 3). In some instances, the bead also moved along the surface of the cell. Upon MTOC reorientation, the MTOC remained docked at the bead:cell interface for the remainder of the interaction with the bead.  121    Figure 3.1. B cells extend lamellipodia across antigen-bearing surfaces  Images of B cells contacting anti-Ig-coated beads (A), APCs displaying antigens on their surface (B), or rigid planar surfaces (e.g. coverslips) coated with anti-Ig antibodies (C). In all cases, the B cells extended lamellipodia (indicated by white arrows) across the antigen-bearing surface. In panel A, primary B cells were mixed with anti-Ig-coated beads for 30 min, fixed, and imaged by scanning electron microscopy (SEM). In panel B, A20 B-lymphoma cells were added to Cos-7 APCs expressing the single chain anti-Igκ antibody on their surface. After 15 min, the cells were fixed and stained for F-actin (red), ezrin (green), and anti-Igκ (blue). A 3D reconstruction is shown. In panel C, A20 cells were plated on anti-Ig-coated coverslips. In the upper panel the cells were allowed to spread for 15 min, fixed, and then stained for F-actin (red) and tubulin (green). A representative 3D reconstruction is shown. In the lower panel, the cells were allowed to spread on anti-IgG-coated coverslips for 4 hr, then fixed, and imaged by scanning electron microscopy.  Movie 3. Polarization of the MTOC towards anti-Ig-coated beads   A20 cells that had been transfected with GFP-α-tubulin were mixed with 4.5-µm diameter anti-IgG-coated beads. Time-lapse confocal microscopy was used to acquire images every 2.5 s between 3 min and 8 min after adding the beads to the cells. The video is played back at 40 frames per second (100X real time).   https://blogs.ubc.ca/jiawangthesis/2016/05/13/chapter-2-movie-3/ Password: JWthesis 122 B cells form an F-actin-rich cup around anti-Ig-coated beads [247]. This F-actin ring structure resembles the IS where the region underlying the cSMAC is also depleted of F-actin [213]. The MTOC in B cells that had engaged anti-Ig-coated beads moved close to this actin-rich cup (Figure 3.2A). Three-dimensional reconstructions showed that the MTOC approached this zone of F-actin clearance when it polarized towards the anti-Ig-coated bead (Figure 3.2A, rightmost panel). Hence, MTOC polarization is coordinated with actin clearance at the center of the antigen contact site. We quantified the extent of BCR-induced MTOC polarization by calculating a polarity index (PI) [85] as described in Figure 2.1 where the distance between the MTOC and the center of the bead is divided by the distance between the center of the B cell and the center of the bead. A PI < 1 indicates that the MTOC was oriented towards the bead. A PI ≤0.75 was used as a definitive measure of MTOC polarization (Figure 2.1). Primary B cells initiated MTOC polarization towards anti-Ig-coated beads within 5 min and by 30 min >95% of the cells had a PI ≤0.75 (Figure 3.2). Approximately 20% of cells having a PI ≤0.75 would be expected for a random distribution of MTOC localizations (Figure 2.1). Similar results were obtained with A20 B-lymphoma cells (see Figure 3.5I, vector control cells), which have been used to study BCR-induced cytoskeletal reorganization and MTOC polarization [44, 55, 85, 220, 247].  In previous studies, our lab showed that culturing primary B cells for several hours or overnight with Toll-like receptor (TLR) ligands induces a primed state that is characterized by increased actin dynamics, relative to ex vivo B cells or B cells cultured in survival cytokines such as IL-4 or BAFF [56]. Furthermore, primary B cells are difficult to transfect and require pretreatment with LPS and survival cytokines to increase their transfection frequency and allow them to survive in culture until substantial siRNA-mediated protein depletion is achieved. 123 Therefore, I tested whether TLR priming of B cells affects BCR-induced MTOC polarization. Figure 3.2 shows that the kinetics and extent of BCR-induced MTOC polarization in lipopolysaccharide (LPS)-stimulated primary B cells were similar to that in ex vivo and IL-4-treated B cells, although the percent of cells with PI <0.75 was higher for the LPS-treated cells at the 15-min time point (Figure 3.2). Thus, BCR-induced MTOC polarization does not appear to be significantly enhanced in LPS-primed B cells.     124 Figure 3.2. Primary B cells reorient the MTOC towards anti-Ig-coated beads  (A,B) Ex vivo primary B cells were mixed with 4.5-μm anti-IgM-coated beads for the indicated times and then stained for α-tubulin, F-actin and nuclei (DAPI). Representative z-projections are shown along with an enlarged 3D reconstruction of the cell in the white box (A). Scale bar: 5 μm. (B) MTOC polarity indices (PI) (line graph) were calculated as in Figure 2.1 for 14–34 bead:cell conjugates per time point per experiment for a total of three experiments. Results are mean ± SEM for three experiments. *P <0.05, **P <0.01 compared to the <1 min time point. The percentage of cells with a PI ≤0.75 was determined in each experiment and the mean ± SEM is shown at each time point. Approximately 20% would be a random MTOC distribution; see Figure 2.1. (C) Primary B cells were cultured with 10 ng/ml IL-4 or 5 µg/ml LPS overnight. Ex vivo, IL-4-treated or LPS-treated B cells were then mixed with anti-IgM-coated beads for the indicated times. MTOC polarity indices were calculated for 15-32 bead-cell conjugates from two experiments. The mean ± s.d. is shown. The percentage of cells with a PI ≤0.75 is shown in the table. Student’s paired t-test showed that the values for the IL-4- and LPS-treated cells were not significantly different (P >0.05) from those for the ex vivo B cells.    3.2.3 Microtubule dynamics and PKCζ activity are important for BCR-induced MTOC reorientation   The movement of the MTOC that accompanies cell polarization (e.g. during cell migration) depends on microtubule dynamics as well as the interaction of microtubules with the cell cortex [193, 211, 226]. Consistent with this idea, MTOC reorientation towards anti-Ig-coated beads was abrogated when B cells were treated with either nocodazole, to deplete microtubules, or with paclitaxel, which stabilizes microtubules (Figure 3.3). The evolutionarily conserved polarity protein PKCζ supports MTOC polarization to the T-cell IS [322]. I found that treating B cells with a cell-permeable myristoylated PKCζ pseudosubtrate peptide blocked MTOC reorientation towards anti-Ig-coated beads (Figure 3.4), consistent with previous reports [85]. Hence BCR-induced MTOC reorientation requires dynamic changes in the microtubule network and involves evolutionarily conserved polarity proteins that mediate MTOC polarization in other 125 cell types. This led us to test whether Rap1, another conserved polarity regulator, plays a role in BCR-induced MTOC reorientation towards the IS.   Figure 3.3. Nocodazole and paclitaxel block BCR-induced MTOC polarization  (A) A20 cells were treated with DMSO or 5 µM nocodazole for 5 min before being mixed with anti-IgG-coated beads for 30 min. B cell:bead conjugates were stained for pericentrin and nuclei were visualized by DAPI staining. Polarity indices were calculated for >33 conjugates. Red lines indicate mean values. ****P <0.0001 using Student’s unpaired t-test. (B) Primary B cells were treated for 5 min with either DMSO or 5 µM paclitaxel before being mixed with anti-IgM-coated beads for 30 min. After staining the cells as in panel A, polarity indices were calculated for >96 126 conjugates from three experiments. Red lines indicate mean values. ****P <0.0001 using Student’s unpaired t-test. Scale bars: 5 μm.   Figure 3.4. PKCζ activity is required for BCR-induced MTOC polarization towards anti-Ig-coated beads  A20 cells were treated with 20 µM PKCζ pseudosubstrate inhibitor (PKCζ-PS) or an equivalent volume of DMSO for 1 h before being mixed for 30 min with anti-IgG-coated beads. B cell:bead conjugates were fixed and stained for α-tubulin and F-actin. Nuclei were visualized by staining with DAPI. The white arrows indicate the MTOC. Representative xy images are shown in (A). Scale bar: 5 μm. For each condition, polarity indices were calculated for >34 conjugates from one experiment (B). Red lines indicate mean values. ****P <0.0001 using Student’s unpaired t-test.   127 3.2.4 BCR-induced MTOC polarization towards anti-Ig-coated beads depends on Rap1 activation Rap1-dependent reorganization of the actin cytoskeleton is required for cSMAC formation, a process that also depends on MTOC reorientation towards the site of antigen contact. Therefore, we asked whether Rap1 is also important for MTOC reorientation. Indeed, we found that siRNA-mediated silencing of Rap1 blocked BCR-induced polarization of the MTOC towards anti-Ig-coated beads in both primary B cells (Figure 3.5A-D) and A20 cells (Figure 3.5E-G). To test whether BCR-stimulated activation of Rap GTPases is required for MTOC reorientation, we used B cell lines that overexpress RapGAPII (also known as RAP1GAP2), a GTPase-activating protein that converts Rap1 and Rap2 into their inactive GDP-bound forms [247]. Our lab has previously used this approach to show that Rap activation is essential for BCR-induced integrin activation and IS formation and for chemokine-induced migration [55, 68, 247, 250, 251]. I  found that RapGAPII expression abrogated the ability of anti-Ig-coated beads to stimulate Rap1 activation (Figure 3.6A) and inhibited MTOC polarization towards anti-Ig-coated beads in A20 cells (Figure 3.6B-C; Movie 4) and in the WEHI-231 IgM+ immature B cell line (Figure 3.6D-F). Altogether our results demonstrate that both Rap1 and Rap2 proteins are important for B cell MTOC reorientation.  128   Figure 3.5. Rap1 is important for MTOC polarization towards anti-Ig-coated beads   (A–D) LPS-stimulated primary B cells were transduced with control siRNA or with Rap1a/b siRNAs. Blots show Rap1 knockdown (A). The cells were mixed with 3-μm beads coated with Alexa Fluor 647-conjugated anti-IgM and then immunostained for α-tubulin. For cells mixed with beads for 30 min, representative 3D reconstruction images are shown (B) along with PIs for >100 conjugates from four experiments (C). ****P <0.0001. Panel D shows the full time course (mean ± SEM; four experiments with >16 cell:bead conjugates per time point per experiment). *P <0.05, **P <0.01, ***P <0.001 compared to control siRNA cells at the same time point. For 3-μm beads, having ∼11% of cells with a PI ≤0.75 would indicate a random MTOC distribution (see Figure 2.1 and Table 2.11). (E-G) A20 cells were transduced with control siRNA or with Rap1a/b siRNAs. Blots show Rap1 knockdown (E). The cells were mixed with 4.5-μm anti-IgG-coated beads for 30 min and stained for α-tubulin (green) and F-actin (red). Representative confocal xy slices overlaid on DIC images are shown in panel F. PIs and the percentage of cells (mean ± SEM) with a PI ≤ 0.75 are shown for >248 conjugates from four experiments (G). ****P <0.0001.  129     130 Figure 3.6. Rap activation is required for MTOC polarization towards anti-Ig-coated beads  (A) Vector control and RapGAPII-expressing A20 cells were mixed with anti-IgG-coated beads for the indicated times. The active GTP-bound form of Rap1, which was precipitated using a GST-RalGDS fusion protein (upper blot), as well as total Rap1 in cell lysates (lower blot), was visualized by immunoblotting with a Rap1 antibody. (B,C) Vector control and RapGAPII-expressing A20 cells were mixed with anti-IgG-coated beads. Cells were stained for pericentrin (green) and DAPI (blue). Panel C shows PIs and the percentage of cells with a PI ≤0.75 for each time point (mean ± SEM; 12-37 conjugates per time point for each of three experiments. *P <0.05, **P <0.01 compared to control cells at the same time point. Two-tailed paired t-tests were used. White arrows indicate the MTOC. (D) Vector control and RapGAPII-expressing WEHI-231 cells were mixed with anti-IgM-coated beads for the indicated times. Rap activation was assessed as in panel A. (E, F) Vector control and RapGAPII-expressing WEHI-231 cells were mixed with anti-IgM-coated beads for 30 min. Bead:cell conjugates were fixed and stained for pericentrin. Representative images of xy slices through the center of individual bead:cell conjugates are shown (E). Scale bar: 5 μm. In panel E, MTOC PIs were quantified for >105 conjugates from three experiments. ****P <0.0001, two-tailed unpaired t-test. The percent of cells with PI ≤0.75 is also shown. Scale bars: 5 μm.   Movie 4. Rap activation is required for MTOC polarization towards anti-Ig-coated beads  Vector control and RapGAPII-expressing cells A20 cells that had been transfected with GFP-α-tubulin (green) were mixed with 3-μm diameter anti-IgG-coated fluorescent beads (blue). Time-lapse spinning disk confocal microscopy was used to acquire images every 60 s between 1 min and 16 min after adding the beads to the cells. The videos are played back at 5 frames per second (300X real time).   https://blogs.ubc.ca/jiawangthesis/2016/05/13/chapter-2-movie-4/ Password: JWthesis  3.2.5 MTOC reorientation towards APCs depends on Rap1 In vivo, B cells are often activated by antigens that are displayed on the surface of APCs such as follicular dendritic cells and subcapsular macrophages [8, 323]. To assess MTOC reorientation towards APCs, we used Cos-7 cells expressing a transmembrane form of an anti-Igκ antibody that binds to the BCR [55]. When B cells contacted these APCs, the MTOC began 131 to move adjacent to the cell–cell contact site as early as 5 min and the majority of cells exhibited maximum MTOC polarization after 20 min (Figure 3.7A-B). Importantly, antigen-induced MTOC polarization towards the APC was inhibited by silencing Rap1 in primary B cells (Figure 3.7A–D) or by expressing RapGAPII in A20 cells (Figure 3.8). Thus, both Rap1 expression and its activation are important for B cells to reorient their MTOC towards polarized antigen arrays.    132   Figure 3.7. Rap1 is important for MTOC polarization towards APCs   (A–D) LPS-stimulated primary B cells that had been transduced with control siRNA, or with Rap1a and Rap1b siRNAs, were stained with CellTrace Far Red (pseudocolored blue) and mixed with APCs expressing anti-Igκ (antigen). B cell:APC conjugates were stained for antigen and α-tubulin. Representative images of B cells that were mixed with APCs for 20 min are shown in panel A. Arrows indicate the MTOC. Scale bar: 5 μm. PIs for >53 B cell:APC conjugates from four experiments (B; quantified as in Figure 2.1). ****P <0.0001. Panels C and D show the PI values and percentage of cells with a PI ≤0.75 for the full time course (mean±SEM; four experiments each with >45 conjugates per point). For APC experiments, 37.5% of cells with a PI ≤0.75 would be random MTOC distribution. *P <0.05, **P <0.01, ***P <0.001 compared to control siRNA cells at the same time point. Two-tailed unpaired t-tests (B) or paired t-tests (C, D) were used. 133   Figure 3.8. Rap activation is required for MTOC polarization towards APCs   (A) Vector control and RapGAPII-expressing A20 cells were mixed with anti-Igκ-expressing Cos-7 APCs for 30 min and then stained for α-tubulin (red) and the antigen (green). White arrows indicate the MTOC. Scale bar: 5 μm. PIs were calculated for >47 B cell:APC conjugates from three experiments (B). ****P <0.0001. Panels C and D show the full time course (mean ± SEM; three experiments, each with >35 conjugates per time point). *P <0.05, **P <0.01, ***P <0.001 compared to control cells at the same time point. Two-tailed unpaired t-tests (B) or paired t-tests (C, D) were used.  Unlike anti-Ig-coated beads, BCR ligands on APCs are mobile. When Rap1 is depleted or its activation is blocked, B cells interacting with antigen-bearing APCs formed microclusters but did not form a cSMAC (Figure 3.7A; Movie 2). We previously showed that blocking Rap1 activation reduced the amount of antigen clustering at the B cell:APC contact site to ∼30% of 134 that in control cells, and reduced BCR-induced phosphotyrosine signaling to a similar degree [55]. To investigate the relationship between antigen density at the contact site, the amount of BCR signaling generated, and the extent of MTOC polarization, 4.5-μm beads were coated with 40 μg/ml, 4 μg/ml, or 0.4 μg/ml goat anti-mouse IgG antibodies. To determine the relative amount of antigen on each set of beads, the beads were stained with Alexa 488-conjugated rabbit anti-goat IgG and analyzed by FACS (Figure 3.9A). This showed that the relative densities of anti-IgG on the beads coated with 40 μg/ml (the standard concentration used in all other experiments with anti-Ig-coated beads), 4 μg/ml, or 0.4 μg/ml were 1.0, 0.44, and 0.02, respectively (Figure 3.9B). For beads coated with a saturating amount of anti-IgG (i.e. the 40 μg/ml), 89% of the cells polarized their MTOC towards these “control” beads such that their MTOC PI was ≤0.75 (Figure 3.9B,C). Approximately, 20% of cells with PI ≤ 0.75 would be expected from a random distribution of MTOC localization (Figure 2.1). Compared to beads coated with the saturating amount of anti-Ig antibody, beads that had 44% of the surface density of anti-Ig antibody induced 35% as much early phosphotyrosine (P-Tyr) signaling at the B cell–bead contact site at 5 min (Figure 3.9C). Importantly, even though the early BCR signaling was reduced, this amount of P-Tyr signaling was still sufficient to cause substantial MTOC polarization after 30 min, with 53% of the cells having a PI ≤0.75 (Figure 3.9D, E). For beads coated with 0.4 μg/ml anti-IgG, the antigen density at the contact site was 10% that of the control beads and the amount of P-Tyr signaling at the contact site was barely detectable above background levels. Nevertheless, 31% of the cells polarized their MTOC towards the beads (MTOC PI ≤ 0.75), which is slightly above the ~20% expected from a random distribution of MTOC localizations (Figure 3.9A,B,D,E). In contrast, when APCs were used to present antigen to B cells, in which Rap1 was depleted (Figure 3.7B–D) or its activation blocked (Figure 3.8), 135 the percent of B cell–APC conjugates with a PI ≤0.75 did not exceed the 37.5% expected for a random distribution of MTOC localizations (note that PIs are calculated differently for APCs and beads; see Figure 2.2). Thus, the decreased antigen gathering and BCR signaling that occurs when Rap1 activation is blocked cannot account for the complete inhibition of MTOC polarization. Therefore, Rap1 must control additional processes that are important for MTOC reorientation. 136    137 Figure 3.9. Relationship of BCR signaling to MTOC polarization   (A) 4.5-μm beads were coated with 40 μg/ml, 4 μg/ml, or 0.4 μg/ml goat anti-mouse IgG (antigen). To determine the relative amount of antigen on each set of beads, the beads were stained with Alexa 488-conjugated rabbit anti-goat IgG and analyzed by FACS. The densities of goat anti-mouse IgG are shown, and are also expressed relative to the geometric mean of beads that had been coated with 40 μg/ml goat anti-mouse IgG (B). (C-E) A20 cells were mixed with the different sets of beads for either 30 min or 5 min. Cells that that had been mixed with anti-Ig-coated beads for 30 min, a time point when MTOC polarization is close to maximal, were stained for pericentrin to visualize the MTOC and calculate PIs. Cells that had been mixed with anti-Ig-coated beads for 5 min were stained for phosphotyrosine (P-Tyr), an indicator of proximal BCR signaling that peaks at this time point (C, left panel). The cells were also stained for F-actin in order to visualize the F-actin-rich membrane cup formed around the bead (C, right panel). Asterisks indicate the bead. Representative images of an A20 cell engaging a bead coated with 40 μg/ml of anti-IgG for 5 min are shown in panel C. Scale bar: 5 μm. The graph in panel D shows the percentage of cells with PI ≤0.75 (left y-axis, blue line; n >45 cells for each set of beads from three experiments) and the mean P-Tyr signal (right y-axis, orange line; 16-49 cells for each set of beads) when the B cells engaged beads coated with the indicated amounts of anti-Ig. The amount of P-Tyr signaling at the bead:cell contact site was quantified as described in the Materials and Methods. Briefly, an 11-μm diameter concentric circle was drawn around the bead and the total P-Tyr fluorescence intensity within this circle was quantified using ImageJ. Background corrections were carried out as described in the Materials and Methods. In panel E the MTOC PIs for >45 cells for each set of beads are shown as a dot plot and the percent of cells with MTOC PI ≤0.75 for each antigen density is shown.   138 3.2.6 Rap1 and Rap2 play non-redundant roles in BCR-induced MTOC reorientation We showed above that silencing the expression of Rap1a and Rap1b in B cells blocks MTOC polarization towards anti-Ig-coated beads and APCs, establishing a role for Rap1 proteins in MTOC polarization. We also showed that overexpressing the Rap-specific GAP, RapGAPII, blocks BCR-induced MTOC polarization. Because RapGAPII blocks BCR-induced activation of both Rap1 and Rap2 [250], this raised the possibility that Rap2 activation could also play a role in MTOC polarization in B cells.  To test the hypothesis that Rap2 proteins also participate in MTOC reorientation, we first investigated which Rap2 isoforms are expressed in B cells. The Immunological Genome project database (https://www.immgen.org), as well as our preliminary microarray and RT-PCR studies (data not shown), indicated that primary mouse splenic B cells express mRNA for all three Rap2 proteins. However, RT-PCR data suggested that A20 B-lymphoma cells express only Rap2c (data not shown). Indeed, transducing A20 cells with a Rap2c-specific siRNA caused ~95% depletion of the 21-kDa band detected by immunoblotting with a pan-Rap2 antibody (Figure 3.10A,B). This suggests that Rap2c is the predominant Rap2 isoform expressed in A20 cells. These experiments also showed that the Rap2c-specific siRNA did not cause significant depletion or upregulation of Rap1 proteins. Similarly, depleting Rap1a and Rap1b did not significantly alter Rap2 protein levels. Hence this approach allowed us to dissect the roles of Rap1 and Rap2 proteins in A20 cells. To assess the contributions of Rap1 and Rap2c to BCR-induced MTOC polarization, we used siRNA to selectively deplete either Rap1a/b or Rap2c in A20 cells and then mixed these cells with anti-IgG-coated beads to create a polarized antigen contact site. Depleting both Rap1a and Rap1b blocked BCR-induced MTOC reorientation (Figure 3.10C-E), as shown above. In 139 A20 cells transduced with control siRNA, the MTOC moved close to the bead contact site in the vast majority of the cells, with 86.6% exhibiting a PI ≤0.75. In contrast, the MTOC failed to reorient towards the anti-Ig-coated bead in the majority of Rap1-depeleted cells, with only 30.6% of the cells exhibiting a PI ≤0.75, compared to 20% for random MTOC localization. Importantly, Rap2c-depleted A20 cells also exhibited an impaired ability to reorient their MTOC towards anti-Ig-coated beads, although this defect was less profound than in Rap1-depleted cells (Figure 3.10C-E). Confocal imaging showed that some A20 cells transduced with Rap2c siRNA were able to move their MTOC close to the anti-Ig-coated bead whereas others did not (Figure 3.10C). This was reflected in the quantification, which showed that 51.3% of the cells transduced with Rap2c siRNA exhibited a PI ≤0.75, compared to 86.6% of control siRNA cells, 30.6% of the Rap1-depleted cells, and 20% for a random MTOC localization (Figure 3.10D-E). Binning the data (Figure 3.10E) showed that Rap2c depletion caused more cells to exhibit a PI between 0.5 and 1, compared to Rap1-depleted cells, but resulted in far fewer cells with a PI >1, i.e. completely non-polarized and in the distal half of the cell relative to the anti-Ig-coated bead. Nevertheless, these data indicate that Rap2c contributes to BCR-induced MTOC polarization and that normal levels of Rap1 cannot compensate for the loss of Rap2c in A20 cells.   140    141 Figure 3.10. siRNA-mediated depletion of either Rap1a/b or Rap2c impairs BCR-induced MTOC reorientation  A20 B-lymphoma cells were transduced with control siRNA, siRNAs targeting both Rap1a and Rap1b, or Rap2c siRNA. (A-B) Cell lysates were immunoblotted with antibodies that recognize all Rap1 isoforms or all Rap2 isoforms. β-actin was used a loading control. A representative western blot is shown (A). Protein densities were quantified for total Rap1 or total Rap2, normalized to the β-actin loading controls for the same sample, and expressed as values relative to the cells that had been transduced with control siRNA (defined as 1.0). Mean and SEM are shown for 3 independent experiments. (C-E) A20 cells were incubated for 30 min with anti-IgG-coated beads (red) before being fixed and immunostained for α-tubulin. Representative xy confocal images are shown (C). Scale bar: 5 μm. The dot plot shows the MTOC polarity indices for n >111 cells for each condition from 3 independent experiments (D). The red lines indicate the mean MTOC polarity index. ****P <0.0001, as determined using an unpaired 2-tailed t-test. The percent of cells with a polarity index ≤0.75 is shown. The distributions of MTOC polarity indices for each condition are shown in (E).   3.2.7 MTOC polarization requires cofilin-mediated actin severing  The actin and microtubule network are often coordinately regulated during processes that involve cell polarization [324]. Indeed, depolymerizing F-actin by treating B cells with latrunculin A (Lat A) abrogated BCR-induced MTOC reorientation (Figure 3.11). Hence an intact actin network is required to support MTOC polarization in B cells. A major function of Rap in lymphocytes is to promote actin remodeling at the IS by activating the actin-severing protein cofilin [55]. Cofilin is activated by dephosphorylation of S3 [164] and the ability of the BCR to stimulate cofilin dephosphorylation depends on Rap. Both blocking Rap activation using RapGAPII [55] and silencing the expression of Rap1a and Rap1b using siRNAs blocked BCR-induced dephosphorylation of cofilin (Figure 3.12). In addition to removing existing actin filaments, cofilin-mediated actin severing creates new barbed ends where the Arp2/3 complex can nucleate branched actin polymerization [320]. When B cells contact antigen-bearing surfaces, both Rap activation and cofilin-mediated actin severing are required for F-actin to be 142 cleared from the center of the contact site and accumulate at the periphery [55]. Because Rap activation and the actin network are important for BCR-induced MTOC reorientation, we asked whether cofilin-mediated actin severing is also important for polarizing the MTOC.     Figure 3.11. BCR-induced MTOC polarization requires an intact actin cytoskeleton  (A,B) Primary B cells were treated with DMSO or 2 μM Lat A for 5 min, mixed with 4.5-μm anti-IgM-coated beads for the indicated times, and then stained for pericentrin. Representative confocal xy slices are overlaid on DIC images (A). For the 30 min time point, PIs and the percent of cells with a PI ≤0.75 (∼20% would be random; see Table S1) were calculated for >156 conjugates from three experiments (B). ****P <0.0001. Scale bar: 5 μm. 143   Figure 3.12. Rap1 depletion inhibits BCR-induced cofilin dephosphorylation  LPS-activated primary B cells were transduced with either control siRNA or with Rap1a and Rap1b siRNAs and then stimulated with 20 μg/ml goat anti-IgM for 5 min. The levels of (A) cofilin that is phosphorylated on S3 (P-cofilin), (B) Rap1 and (C) β-actin as a loading control were assessed by immunoblotting. Band intensities were quantified using ImageJ. Protein levels are expressed relative to that for the corresponding control siRNA time 0 sample. Representative blots from one of three independent experiments are shown.   A B C 144 To assess the role of cofilin-mediated F-actin severing in MTOC polarization, A20 cells were transiently transfected with mCherry-tagged wild type (WT) cofilin, or with a mutant form of cofilin that had a phosphomimetic SD substitution at S3. This S3D mutant, which acts in a dominant negative manner to prevent BCR-induced actin reorganization [55, 56], significantly reduced the ability of A20 cells to polarize their MTOC towards anti-Ig-coated beads, compared to cells expressing WT cofilin-mCherry (Figure 3.13A,B). Similarly, treating B cells with cell-permeable cofilin inhibitory peptides that were derived from cofilin (peptides M and W) also abrogated MTOC polarization (Figure 3.13C,D). These peptides prevent endogenous cofilin from binding to and severing actin filaments [295] and thereby inhibiting cofilin-dependent actin dynamics in B cells [56]. MTOC polarization occurred normally in the presence of the control Q peptide in which key residues in the W peptide were changed so as to ablate F-actin binding [295]. Both the cofilin-inhibitory peptides [295] and cofilin S3D [325] bind F-actin and prevent severing by endogenous cofilin, raising the possibility that they also block the action of other actin-severing proteins. However, we have previously shown that cofilin is the major actin-severing protein in anti-Ig-stimulated B cells [55]. Importantly, our finding that siRNA-mediated depletion of cofilin blocked BCR-induced MTOC polarization (Figure 3.13E-G) indicates that both cofilin and its actin-severing functions are required for BCR-induced MTOC reorientation.  Because Rap1 controls cofilin dephosphorylation and severing activity in B cells [55], we asked whether cofilin acts downstream of Rap1 to promote MTOC polarization. Indeed, expressing the constitutively active cofilin S3A mutant restored BCR-induced MTOC polarization in RapGAPII-expressing A20 cells (Figure 3.13H,I). The ability of activated cofilin to bypass a defect in Rap1 activation argues that cofilin is the major downstream effector of Rap1 that couples BCR engagement to MTOC polarization. 145    146 Figure 3.13. Cofilin controls BCR-induced MTOC polarization  (A,B) A20 cells expressing either WT or cofilin S3D fused to mCherry were mixed with anti-IgG-coated beads for 30 min. Representative confocal images of α-tubulin and F-actin staining (A). PIs for >79 cells from three experiments (B). (C,D) Primary B cells were treated with the control Q peptide (5 μM) or the M and W cofilin-inhibitory peptides (5 μM each) and then mixed with anti-IgM-coated beads for 30 min. Representative images of pericentrin staining (C). PIs for >43 cells from three experiments (D). (E-G) A20 cells were transduced with control siRNA or cofilin siRNA. The blot shows cofilin knockdown (E). The cells were mixed with anti-IgG-coated beads for 30 min before being stained for pericentrin and for the anti-IgG on the beads. Representative confocal images are shown (F; the dotted line is the outline of the cell) along with PIs for >29 cells (G). ****P <0.0001. (H,I) RapGAPII-expressing A20 cells transfected with WT cofilin or the constitutively active cofilin S3A fused to mCherry were mixed with anti-IgG-coated beads for 30 min. Representative confocal images of α-tubulin and F-actin staining (H). PIs for >51 cells from three experiments (I). ***P <0.001, ****P <0.0001. Scale bars: 5 μm.    At the T cell IS and natural killer (NK) cell IS, the MTOC comes close to the plasma membrane, and this is associated with F-actin clearance at the center of the IS [217, 304, 326]. To test whether the MTOC approaches the membrane in response to BCR engagement, and whether this involves cofilin, we employed TIRFM. This allowed us to visualize only the actin and microtubule structures that were within 100 nm of the interface between a B cell and an anti-Ig-coated coverslip. A20 cells treated with the control Q peptide spread normally and formed a peripheral ring of F-actin surrounding a central region that was depleted of F-actin (Figure 3.14A). In these cells, the MTOC and microtubules moved into the 100-nm TIRF plane, with the MTOC in the center of the actin-depleted region (Figure 3.14A). When cofilin-mediated F-actin severing was blocked, B cell spreading was impaired, F-actin was not reorganized into a peripheral ring surrounding an actin-poor region, and the ability of the MTOC to approach the plasma membrane at the center of the contact site was significantly reduced (Figure 3.14A,B). siRNA-mediated depletion of cofilin also inhibited both actin reorganization and movement of 147 the MTOC into the TIRF plane at 15, 30 and 60 min (Figure 3.14C–E; Figure 3.15).  To resolve whether Rap1 and cofilin are important for the MTOC to move towards the antigen contact site, versus the MTOC being retained at that site (i.e. docking), we performed real-time imaging. When A20 cells expressing GFP-α-tubulin and the F-actin probe F-tractin-tdTomato were plated on anti-Ig-coated coverslips, the MTOC moved rapidly towards the coverslip and localized to the center of the region that was cleared of F-actin, with microtubules extending from the MTOC to the peripheral ring of F-actin (Movie 5, Movie 6, and Movie 7). In contrast, when Rap1 activation was blocked, resulting in impaired cell spreading and actin reorganization, the MTOC did not move close to the antigen contact site (Movie 5). For most RapGAPII-expressing cells, the MTOC did not move at all. In a few cells, the MTOC shifted slightly towards the antigen contact site but then moved back towards the center of the cell without ever coming close to the antigen contact site. Likewise, when B cells were treated with cofilin-blocking peptides, the cells did not spread, actin reorganization was impaired, and the MTOC did not move towards the antigen contact site (Movie 6, Movie 7). Kymographs depicting the time evolution of the fluorescence signals in the lowest xy plane of the cell also showed that the MTOC moved rapidly into this plane in control cells, but not in cells treated with cofilin-blocking peptides (Figure 3.14F). This shows that BCR-induced spreading, actin reorganization, and MTOC polarization are tightly linked and are coordinately regulated by Rap1 and cofilin.  148    149 Figure 3.14. Cofilin-mediated actin reorganization is required for the MTOC to approach the plasma membrane  (A,B) A20 cells were treated with 5 μM of the control Q peptide (A) or with 5 μM each of the M and W cofilin-blocking peptides and then allowed to spread on anti-IgG-coated coverslips for 15 min. Cells were stained for α-tubulin and F-actin and imaged by TIRFM with a 100-nm depth. Fluorescence intensity profiles along the dotted lines are plotted along with the percent of cells in which the MTOC was in the TIRF plane (B; mean ± SEM; >41 cells per condition in each of three experiments). **P <0.01. (C–E) A20 cells transduced with control siRNA or cofilin siRNA were allowed to spread on anti-IgG-coated coverslips for 15–60 min and imaged by TIRFM with a 100-nm depth. Images of cells at the 15-min time point are shown (C). The percent of cells with the MTOC within the TIRF plane (D) and the percent of cells that exhibited a peripheral F-actin ring surrounding a central actin-depleted region (E) are shown for each time point. (F) A20 cells expressing GFP–α-tubulin and F-tractin-tdTomato were treated with control (Q) or cofilin-inhibitory (M and W) peptides, added to anti-Ig-coated coverslips, and imaged in real time for 8 min by confocal microscopy. Still images of Movie 6 at 8 min are shown in the upper panels. The kymographs in the lower panels represent a time series of images for the confocal slice closest to the coverslip taken along the white line every 10 s. Scale bars: 10 μm.         150   Figure 3.15. Cofilin knockdown blocks actin reorganization and MTOC polarization   (A,B) A20 cells were transduced with control or cofilin siRNAs and allowed to spread on anti-Ig-coated coverslips for 15 min (A) or 60 min (B). The cells were then fixed, stained with anti-α-tubulin antibodies and rhodamine-phalloidin, and imaged by TIRFM with the TIRF plane extending 100 nm from the coverslip. Each dot on the graphs is an individual cell (n >39 cells for each condition). The graphs depict three parameters for each cell: 1. The extent of cell spreading as indicated by the cell area (y-axis), which was quantified using ImageJ. 2. Whether the MTOC had polarized towards the coverslip such that it was in the 100-nm TIRF plane (cells with their MTOC in the TIRF plane are above the x-axis; cells in which the MTOC was not in the TIRF plane are below the x-axis). MTOC polarization was characterized by the presence within the TIRF plane of a bright spot of tubulin staining at a point where the microtubules converge. Because the height of spread cells is ~5 μm, having the MTOC in the lowest 100 nm of the cell represents a very strong degree of polarization.  3. Whether the cells had reorganized their F-actin to form a distinct peripheral ring of F-actin surrounding an actin-depleted region in the center of the cell. Green dots represent cells that had reorganized F-actin in this manner. Red dots represent cells that were unable to reorganize their F-actin. Representative TIRFM images of F-actin organization in control and cofilin siRNA cells are shown in Figure 3.14C.   151   Movie 5. The actin and microtubule cytoskeletons are coordinately regulated in a Rap1-dependent manner during B cell spreading  Vector control and RapGAPII-expressing cells A20 cells that had been transfected with F-tractin-tdTomato (red) and GFP-α-tubulin (green) were allowed to attach to anti-Ig-coated coverslips. Time-lapse spinning disk confocal microscopy was used to acquire images every 40 s between 5 min and 15 min. 3D reconstruction of z-stacks (0.5 µm step size) with side and top views of the cells are played back at 5 frames per second (200X real time) Still images from the movies are shown above.   https://blogs.ubc.ca/jiawangthesis/2016/05/13/chapter-2-movie-5/ Password: JWthesis  Movie 6. Cofilin inhibitory peptides prevent MTOC reorientation towards the anti-Ig contact site  A20 cells expressing GFP-α-tubulin (green) and F-tractin-tdTomato (red) were treated with the control Q peptide (5 μM) or the M/W cofilin-inhibitory peptides (5 μM each) for 30 min and then added to anti-IgG-coated coverslips. Time-lapse spinning disk confocal microscopy was used to acquire xy images of the contact site between the cell and the anti-IgG-coated coverslip every 10 s between 1 min and 8.5 min after adding the cells to the coverslips. The videos are played back at 10 frames per second (100X real time). See Figure 3.14 for screen captures.  https://blogs.ubc.ca/jiawangthesis/2017/02/20/cofilin-inhibitory-peptides-prevent-mtoc-reorientation-to-the-anti-ig-contact-site/  Password: JWthesis 152   Movie 7. Cofilin inhibitory peptides prevent MTOC polarization towards the contact site with anti-Ig-coated coverslips  A20 cells expressing mTagRFP-tubulin (red) and GFP-F-tractin (green) were treated with the control Q peptide (5 μM) or the M/W cofilin-inhibitory peptides (5 μM each) for 30 min and then added to anti-IgG-coated coverslips. Starting at 5 min after adding the cells to the coverslips, time-lapse spinning disk confocal microscopy was used to acquire z-stacks at 0.5 μm step size once every 22 s for a total of 7 min for Q peptide-treated cells, and at 0.75 μm step size once every 45s for a total of 18 min for M/W peptide-treated cells. 3D reconstruction of z-stacks with side and top views of the cells are played back at 4 frames per second (~90X real time for Q peptide-treated cells and 180X real time for M/W peptide-treated cells). Representative still images from the movie is shown above.  blogs.ubc.ca/jiawangthesis/2017/02/20/cofilin-inhibitory-peptides-prevent-mtoc-polarization-towards-anti-ig-coated-coverslips/  Password: JWthesis     153 Because BCR-induced spreading on a planar surface correlated with MTOC polarization, we sought to separately assess the roles of BCR signaling and cell spreading. When A20 cells were plated on coverslips coated with anti-Ig antibodies, 75–80% of the cells polarized their MTOC towards the coverslip and had a PI ≤0.75 (calculated as in Figure 2.1). In contrast, when the cells were plated on coverslips coated with anti-MHC II antibodies or PLL, the percentage of cells with a PI ≤0.75 was similar to the 37.5% value (Figure 3.16A,B) expected for a random distribution of MTOC localizations (see section 2.12). We also assessed the effect of cell spreading in the absence of BCR signaling by plating A20 cells on a pliable fibronectin-coated substrate and using a FlexCell apparatus to apply radial stretch. When stretched, A20 cells flattened, and spread to approximately the same area as cells plated on a rigid anti-Ig-coated substrate (Figure 3.16B-D). The stretched cells, which developed abnormal actin structures, did not reorient their MTOC towards the substrate contact site (Figure 3.16C,D). The percentage of cells with a PI ≤0.75 did not exceed the 37.5% value expected for a random MTOC localization. Hence, spreading is not sufficient to cause MTOC polarization in the absence of BCR signaling. This, however, does not exclude a role for cell spreading in BCR-induced MTOC polarization.    154    Figure 3.16. Cell spreading alone is not sufficient to elicit MTOC polarization  (A) A20 cells were allowed to spread for 15 min on coverslips coated with anti-IgG antibodies, anti-MHC II antibodies, or 0.001% PLL. The cells were then fixed and stained with α-tubulin antibodies and rhodamine-phalloidin. Cell spreading area and MTOC polarity indices for >31 cells from three experiments per condition were quantified. To compare MTOC polarization in cells that had spread to the same extent, we determined the percent of cells with a PI ≤ 0.75 only for cells (dots below the horizontal dashed line) with an area of 100-300 μm2 (dots between the vertical dashed lines). For coverslips, a random distribution of MTOC localizations would result in 37.5% of the cells having a PI ≤0.75. (B-D) A20 cells were subjected to mechanical stretch, as we have done previously for tumor cells [296], before being fixed and stained for α-tubulin and F-actin. Panel B shows the MTOC PI values for A20 cells that were stretched or had been plated on anti-Ig, anti-MHC II, or PLL (for the latter three conditions, these are the same cells that are also plotted in panel A). The percent of cells with a PI ≤0.75 is shown; 37.5% would reflect a random distribution of MTOC localizations. In panel C, representative images show an xy view 155 of A20 cells that had spread on coverslips coated with anti-IgG for 15 min or had been stretched for 5 min as above. The lower panels depict xz slices along the dotted lines. This shows that the cells had spread and flattened to the same extent. Scale bars: 5 μm. To compare MTOC polarization in cells that had spread to the same extent, we determined the percent of cells with a PI ≤0.75 (dots below the horizontal dashed line) only for cells with an area of 100-300 μm2 (dots between the vertical dashed lines). A random distribution of MTOC localizations would result in 37.5% of the cells having a PI ≤0.75.  3.3 Discussion 3.3.1 Summary of findings The IS is a polarized cell structure that resembles the leading edge of a migrating cell or the yeast bud site. Establishing functional patterns of membrane protein organization and polarizing vesicular traffic towards these sites requires the coordinated reorganization of the actin and microtubule cytoskeletons. Rap GTPases are evolutionarily conserved regulators of cell polarity and actin reorganization [247]. Although Rap activation is essential for the actin reorganization that drives IS formation [247], its role in polarizing the microtubule network towards the IS had not been investigated. The microtubule network plays a critical role in both IS formation and function. In T cells and NK cells, reorientation of the microtubule network towards the IS directs the secretion of cytotoxic granules towards target cells [217, 229]. In B cells, movement of the MTOC towards the IS facilitates antigen acquisition [85], which is critical for B cells to elicit T cell help, a limiting step in B cell activation [64, 327]. Our finding that Rap1 is required for BCR-induced MTOC polarization towards antigen contact sites reveals an important new contribution of Rap in establishing cell polarity during IS formation.    We also report for the first time that Rap1 coordinates actin remodeling and MTOC reorientation at the IS via the actin-severing protein cofilin. BCR signaling induces Rap-dependent dephosphorylation of cofilin, which allows cofilin to bind F-actin and promote 156 localized actin severing [55]. This results in the clearance of F-actin from the center of the IS and the formation of a peripheral ring of branched F-actin. When we depleted B cells of both Rap1a and Rap1b, or expressed RapGAPII in order to inactivate Rap1 and Rap2, B cells were unable to reorient their MTOC towards the site of antigen contact and consequently BCR microclusters did not coalesce into a cSMAC. Moreover, we showed that blocking the Rap1-cofilin pathway also disrupted normal actin remodeling and B cell spreading. Therefore, Rap1 regulates the coordinated remodeling of the actin and microtubule cytoskeletons in response to BCR binding to polarized arrays of antigens. Additionally, we showed that the Rap1a/b and Rap2c proteins regulate MTOC polarization towards the site of antigen contact in a non-redundant fashion.   3.3.2 The role of Rap1 and Rap2c isoforms in MTOC reorientation Although Rap1b is expressed at higher levels than Rap1a in B cells (see Appendix A), and consequently has a more important role in B cell adhesion and migration [255, 288], I assessed the combined role of both Rap1 family proteins by using siRNAs to deplete both Rap1a and Rap1b. I showed that Rap1 family proteins are important for MTOC reorientation, but did not assess whether Rap1a and Rap1b have distinct or overlapping roles in establishing cell polarity in B cells. As a complementary loss-of-function approach for studying Rap protein function, I expressed RapGAPII in B cell lines and showed that this also abrogated BCR-induced MTOC reorientation and polarization. However, because RapGAPII converts both Rap1 and Rap2 to their inactive GDP-bound forms [76], it raised the possibility that Rap2 proteins also contribute to BCR-induced MTOC polarization. Indeed, in A20 cells, which express Rap2c but not Rap2a or Rap2b, siRNA-mediated depletion of Rap2c resulted in a partial inhibition of BCR-induced MTOC reorientation.  157 To our knowledge, this is the first demonstration that a Rap2 family member is involved in MTOC reorientation and in establishing cell polarity in lymphocytes. Although Rap1 and the yeast Rap orthologue Bud1p promote the actin remodeling that establishes cell polarity [283], there is only one previous report linking a Rap2 protein to cell polarity. Gloerich et al. showed that Rap2a, but not Rap2b or Rap2c, is required for intestinal epithelial cells to form a polarized microvilli brush border [267]. This unique role for Rap2a in the polarization of intestinal epithelial cells may reflect its distinct subcellular localization compared to the other Rap2 proteins. Although we have identified a role for Rap2c in BCR-induced MTOC polarization, at least in A20 cells, primary B cells express all three Rap2 isoforms. Future experiments could address whether Rap2a and Rap2b also contribute to B cell polarity and MTOC reorientation. Our findings suggest that Rap1 and Rap2c have non-redundant functions in BCR-induced MTOC polarization. This could reflect the binding of distinct effector proteins to Rap1a/b versus Rap2c or distinct subcellular localizations of these GTPases. The localization of GTPases to different cellular compartments or to different membrane domains is controlled in part by lipid modifications at their C-termini [261]. Rap1a, Rap1b, and Rap2b are geranylgeranylated whereas Rap2a and Rap2c are farnesylated at their C-terminal CAAX motifs [328]. This may account for the finding that Rap1 and Rap2 are present in different intracellular vesicles in migrating T cells [318]. The subcellular localization of the different Rap2 GTPases and, importantly, their activated forms in B cells is not known. We have previously shown that BCR engagement on A20 cells causes Rap2 activation [68], which we now know is the activation of Rap2c because this is the only Rap2 isoform expressed in these cells. It is not known whether Rap2a or Rap2b are activated in primary B cells, which express all three Rap proteins.  158 Many proteins are involved in MTOC movement and polarization, and these represent potential targets of Rap1 and Rap2c signaling.  Downstream targets of BCR signaling that are essential for MTOC reorientation include evolutionarily conserved polarity proteins such as Cdc42, the Par polarity complex, and protein kinase C ζ. [82, 85, 220] These proteins initiate cell polarization to create cellular asymmetry and define the site towards which the MTOC reorients. Reorientation of the MTOC also requires the capture of microtubule plus ends at the cell cortex via interactions with membrane-associated proteins or with the cortical actin cytoskeleton. Microtubule plus-end capture proteins include dynein motor complexes as well as actin-microtubule crosslinking proteins such as IQGAP1 and spectraplakins [196]. The anchoring of microtubules to the cell cortex by these plus-end capture proteins allows forces to be exerted on microtubules to move the MTOC towards the polarity cue. In the next chapter, I show that IQGAP1 is essential for BCR-induced MTOC reorientation and that Rap1-dependent actin reorganization promotes the accumulation of IQGAP1 at the periphery of the antigen contact site. Whether Rap2c controls the localization or function of other actin-microtubule crosslinking proteins or cortical capture proteins such as dynein is not known.  The unique role of Rap2c in BCR-induced MTOC polarization could also reflect the ability of Rap2, but not Rap1, to interact with TRAF2/Nck-interacting kinase (TNIK) [316, 329]. TNIK has an evolutionarily conserved role in remodeling the actin cytoskeleton and regulating actin-dependent processes such as cell migration [330]. TNIK phosphorylates the Arp2 subunit of the Arp2/3 complex, which allows Arp2/3 to initiate the formation of branched actin networks [330] and I show in Chapter 5 that Arp2/3 activity is required for BCR-induced MTOC reorientation.   159 In both B and T cells, remodeling of the actin cytoskeleton is required for antigen-induced MTOC reorientation [82, 217].  We have shown that Rap1-dependent activation of cofilin is required for BCR-induced actin reorganization, a necessary prerequisite for MTOC polarization in B cells [82]. Cofilin promotes actin remodeling both by severing existing filaments and by creating new barbed ends where the Arp2/3 complex can initiate actin branching [162, 331]. Thus, Rap1 could promote actin remodeling and MTOC polarization via cofilin whereas Rap2 does so via a TNIK-Arp2/3 pathway. siRNA-mediated depletion could be used to assess the role of TNIK in BCR-induced actin reorganization and MTOC reorientation. Although TNIK is required for CD40-induced activation of NFκB and c-Jun N-terminal kinase (JNK) in B cells [332] its role in BCR signaling and MTOC polarization have not been investigated. In summary, our findings suggest that Rap2c contributes to BCR-induced MTOC polarization in a manner that is distinct from that of Rap1.      3.3.3 The Rap1-cofilin pathway is important for MTOC reorientation BCR-signaling induces Rap1 activation, which is required for cofilin dephosphorylation and actin severing [55]. Although F-actin clearance is associated with MTOC reorientation towards the IS in T cells and NK cells [217, 304, 326], this is the first report that cofilin is involved in MTOC polarization towards the IS. We found a strong correlation between BCR-induced MTOC reorientation and the inter-related processes of actin reorganization and cell spreading, which are all dependent on cofilin. In both B and T cells we showed previously that Rap-dependent activation of cofilin is essential for the formation of a peripheral ring of branched F-actin that drives cell spreading, and for the concomitant depletion of F-actin from the center of the IS [55]. This pattern of actin reorganization is also associated with MTOC reorientation 160 towards the T cell and NK cell IS [217, 304, 326]. The inhibition of BCR-induced actin reorganization and MTOC polarization caused by depleting Rap1 or blocking Rap activation was phenocopied by depleting cofilin or inhibiting its function. Moreover, expressing an activated form of cofilin bypassed a block in Rap activation and restored BCR-induced spreading [55], actin reorganization, and MTOC polarization. Likewise, inhibiting Arp2/3, which disrupts normal antigen receptor-induced actin cytoskeleton remodeling at the periphery of the IS that mediates cell spreading [80], also impaired MTOC reorientation towards the IS. Taken together, these results demonstrate that MTOC reorientation relies on the critical function of Rap1 and cofilin-mediated actin remodeling. In the next chapter, I investigated how Rap coordinates the regulation of the actin and microtubule cytoskeletons during IS formation. Activation of the Rap1-cofilin pathway enhances BCR signaling by promoting B cell spreading, BCR microcluster formation, and the coalescence of BCR microclusters into a cSMAC [55]. Hence, the impaired MTOC reorientation in cells in which Rap activation is blocked could simply reflect decreased BCR signaling in these cells. To test this, we altered the density of anti-Ig coated onto beads and assessed the relationship between the level of BCR signaling and the extent of MTOC polarization. We found that even when BCR signaling was reduced by 65%, BCR-induced MTOC reorientation was still significantly greater than that caused by stimulating Rap1-depleted or Rap-inactivated cells with saturating concentrations of anti-Ig coated onto beads. Thus, the decreased BCR signaling caused by blocking Rap activation cannot by itself account for the nearly complete block in MTOC polarization.  This suggests that additional Rap- and cofilin-dependent mechanisms are required for MTOC polarization.  As described above (see section 1.6.1.2), there are several models that could explain how the actin cytoskeleton supports MTOC reorientation. Positioning of the MTOC requires 161 microtubule-mediated contacts with the cell cortex. Microtubule capture proteins associated with the plasma membrane or the actin cytoskeleton can mediate contacts between microtubule plus ends and the cell cortex. Remodeling of the actin cytoskeleton changes the pre-existing actin structures while creating new structures, which may move the MTOC by changing the spatial organization of interactions with microtubules. During cell spreading, the Rap1-cofilin pathway and Arp2/3-mediated actin nucleation drive the polymerization of dendritic actin networks that exert forces on the cell membrane and cause cell spreading. Because of the interactions between the cell cortex and the microtubules, outward forces exerted on the plasma membrane could be transmitted as pulling forces on the microtubules to reposition the MTOC towards the interface between the B cell and the antigen-presenting surface. However, when we mechanically stretched cells or allowed cells to adhere and spread on non-BCR substrates, the cells did not reorient their MTOCs towards the site of antigen contact. Importantly, cells that were mechanically stretched did not form peripheral dendritic actin networks and did not exhibit radial cell spreading, which occur frequently in B cells spreading on substrates that stimulate BCR signaling. Therefore, these results suggest that BCR-induced remodeling of the actin cytoskeleton and the resulting formation of a peripheral ring of F-actin is required for MTOC reorientation in response to polarized arrays of antigens.  Although cell spreading alone is not sufficient to drive MTOC repositioning, this does not exclude the role of actin reorganization in reorienting the MTOC. The remodeling of the actin cytoskeleton that mediates cell spreading may work in concert with other mechanisms to promote MTOC movement. Actin cytoskeleton dynamics and organization are controlled by a large number of proteins that regulate actin polymerization, capping, depolymerization, branching, and bundling, as well as interactions with membrane proteins and other structural 162 elements such as microtubules. Our finding that BCR-induced actin reorganization is critical for MTOC polarization suggests that many of these actin-regulatory proteins could modulate MTOC reorientation in B cells. Future studies could explore how different classes of actin regulators influence MTOC polarization in B cells. In Chapter 4, we explore how actin-microtubule crosslinking proteins coordinate Rap-dependent actin reorganization and MTOC reorientation.  During physiological B cell-APC interactions, B cells bind both cognate antigen and integrin ligands (i.e. adhesion molecules such as ICAM-1) on the surface of the APC. However in our studies we used anti-Ig-coated beads and APCs that do not express integrin ligands. Thus the interaction of the B cells with these polarized antigen sources is mediated entirely via the binding of BCR to cognate antigen. Rap activation is important for both BCR- and integrin-mediated cytoskeletal remodeling, and for the inside-out signaling that activates integrins [68, 77, 78]. This allowed us to investigate the mechanism of BCR-mediated MTOC reorientation independent of integrin-mediated adhesion. However, the contribution of integrin-mediated adhesion and signaling to B cell polarity responses will be an interesting area of investigation. It has previously been shown that B cell activation and IS formation can be induced by limiting amounts of antigen when LFA-1 integrin ligands are present [72]. Importantly, adhesion mediated by integrin ligands attached to antigen-bearing membranes synergizes with BCR signaling to promote B cell activation by low affinity antigens [69, 72]. Therefore, it would be interesting to determine whether the engagement of LFA-1 integrins synergizes with BCR signaling to polarize the MTOC during IS formation.   163 3.3.4 Perspectives B cell activation and BCR-mediated antigen presentation are critical for both normal immune responses and B cell-dependent autoimmune diseases. Hence, elucidating how the BCR controls the actin and microtubules dynamics involved in IS formation and MTOC reorientation may suggest new strategies for developing drugs aimed at enhancing immunity or limiting autoimmunity and acute graft rejection. For example, studies in a graft rejection model showed that targeting the RhoA effector p160ROCK in T cells inhibits chronic rejection of heart allografts in rats [194]. p160ROCK is important for linking centrioles in the centrosome and also facilitates MTOC repositioning in the cell [333]. Although p160ROCK has many functions in regulating the cytoskeleton, these reports suggest that MTOC polarization may play a role in T cell activation in vivo. Drugs that selectively target the signaling pathways that regulate MTOC polarization in immune cells, such as Rap-regulated pathways, could potentially be used to limit the inappropriate and excessive activation of immune cells in autoimmune diseases, inflammatory diseases, and allergy. In chapter 4 I identify key proteins that are important for the role of Rap in coordinating actin cytoskeleton remodeling with MTOC polarization to the B cell immune synapse.164 Chapter 4: IQGAP1 and CLIP-170 couple actin reorganization to MTOC polarization   4.1 Introduction The reorientation of the MTOC to the site of antigen contact is a key polarity event that promotes BCR microcluster coalescence and IS formation during B cell activation. In Chapter 3 I showed that the Rap1 GTPase and cofilin-mediated actin remodeling are required for BCR-induced MTOC reorientation. Live cell imaging showed that MTOC reorientation occurred concomitantly with actin remodeling and actin-dependent cell spreading. Consistent with the idea that actin reorganization is a pre-requisite for MTOC reorientation, I found that blocking Rap activation, preventing cofilin-mediated actin severing, and disrupting the actin cytoskeleton with latrunculin A, all prevented MTOC reorientation towards anti-Ig-coated beads [82]. It has been proposed that actin clearance at the cSMAC is coordinated with MTOC reorientation. In T cells, the formation of a peripheral ring of F-actin at the antigen contact site is accompanied by the clearance of F-actin at the center of the IS. This allows the MTOC to dock at the plasma membrane and support the directed secretion of cytotoxic granules towards the IS that forms at the T cell:target cell contact site [217]. Similarly, when B cells encounter antigen-bearing surfaces or APCs, F-actin accumulates at the periphery of the contact site and is cleared from the center of the contact site as the MTOC moves towards the plasma membrane at the antigen contact site [82]. I showed that when cofilin-dependent actin severing was blocked, actin clearance did not occur and, importantly, the MTOC did not move towards the plasma membrane [82]. Although actin and microtubule reorganization is coordinated during IS formation, the 165 mechanisms that coordinate the restructuring of these two cytoskeletons are not completely understood. At the periphery of the T cell contact site with an APC, the interaction of microtubules with the cell cortex may be important for establishing cell polarization by reorienting the MTOC towards the IS [211, 324]. Because F- actin accumulates at the periphery of the IS, where microtubules are captured, it was proposed that microtubules interact directly with the peripheral actin network and that actin-microtubule crosslinking proteins facilitate MTOC reorientation towards the IS. A number of actin-microtubule crosslinking proteins have been identified, including spectraplakins (e.g. ACF7) [334] and IQGAP1 [234]. IQGAP1 is a widely expressed, evolutionarily conserved scaffolding protein that contains multiple protein-interaction domains. Scaffolding proteins can coordinate and link cellular responses that are carried out by their binding partners [234]. IQGAP1 has been implicated in regulating cell polarity, actin cytoskeletal reorganization, microtubule dynamics and cell signaling [230-232, 234, 239, 335]. The role of IQGAP1 in cell polarity is illustrated by its importance for cell migration. In neutrophils, IQGAP1 is recruited to the leading edge of the cell in response to chemokine receptor signaling where it controls directional migration by interacting with the polarity regulator Cdc42 [336]. In fibroblasts, IQGAP1 at the leading edge mediates the capture of microtubule plus ends by forming a complex with the plus-end binding protein CLIP-170 as well as APC and the Cdc42/Rac1 GTPases, thereby connecting microtubules to the actin cytoskeleton [228, 234, 337, 338]. These interactions also stabilize captured microtubules, which facilitates reorientation of the MTOC [228, 234, 337, 338], and establishes polarity that is required for directional cell migration. Additionally, IQGAP1 regulates actin cytoskeleton remodeling via its interaction with Cdc42 and Rac1, which are important for controlling Arp2/3-mediated actin 166 polymerization [127, 232, 234, 339]. Although IQGAP1 is important for regulating cell polarity and cytoskeletal reorganization, which are critical processes during B cell IS formation, the role of IQGAP1 in B cells is not understood [234].  IQGAP1 has been implicated in the regulation of cell polarity, actin reorganization, and IS formation in T cells and NK cells. Similar to the localization of IQGAP1 at the leading edge during cell migration, during T cell IS formation, IQGAP1 relocates to the dSMAC, where F-actin accumulates [217]. This may allow the growing peripheral F-actin network at the dSMAC to exert forces on microtubules and move the MTOC towards the cSMAC, where the submembrane actin meshwork has been cleared [340]. In cytotoxic T cells, docking of the MTOC at the actin-depleted cSMAC is required for the directional release of cytotoxic granules at the IS towards the target cell [217]. Other studies in CD4+ T cells showed that IQGAP1 also controls the reorganization of the actin cytoskeleton at the IS [341]. In NK cells, which also form an IS with target cells, both Rap1b and IQGAP1 are important for the proper organization of the MTOC and the microtubule network [342]. However, the underlying mechanism was not elucidated.  Because I showed in Chapter 3 that reorientation of the MTOC occurs in a coordinated manner with actin remodeling during IS formation in B cells, I asked whether proteins that connect the actin cytoskeletons and microtubules are important for BCR-induced MTOC reorientation. IQGAP1 can associate directly with the actin cytoskeleton and mediate interactions with microtubule plus ends, and has been implicated in controlling MTOC polarization in migrating cells. Therefore, I tested the hypothesis that IQGAP1 regulates BCR-induced MTOC polarization by linking the actin and microtubule cytoskeletons. Because Rap1-regulated actin 167 reorganization is important for MTOC reorientation in B cells, I also tested whether IQGAP1 acts downstream of BCR-induced Rap1 activation to promote MTOC reorientation.   4.2 Results 4.2.1 BCR-induced MTOC polarization requires IQGAP1 and CLIP-170  To gain further insight into how actin reorganization promotes MTOC reorientation in B cells, we investigated actin-microtubule crosslinking proteins. IQGAP1 binds F-actin as well as the microtubule plus-end-binding protein CLIP-170 (also known as CLIP1) [228, 239]. During fibroblast migration, IQGAP1 links microtubules to the actin cortex and coordinates actin reorganization with MTOC reorientation [228, 239, 338]. We found that depleting either IQGAP1 or CLIP-170 in A20 cells (Figure 4.1 A,B) blocked MTOC polarization towards anti-Ig-coated beads (Figure 4.1 C-F) and towards APCs (Figure 4.1 G, H). Moreover, expressing the truncated C-terminal CLIP-170-binding domain of IQGAP1 (IQGAP1-CT), which lacks the N-terminal F-actin binding domain [228], also blocked MTOC polarization towards anti-Ig-coated beads (Figure 4.2). IQGAP1-CT expression in fibroblasts can disrupt the interaction between CLIP-170 and endogenous IQGAP1 presumably by competing with endogenous IQGAP1 for binding CLIP-170 at the plus ends of microtubules and thereby preventing CLIP-170-dependent microtubule interactions with the cell cortex [228]. Although this is consistent with the idea that IQGAP1–CLIP-170 interactions are important for BCR-induced MTOC reorientation, IQGAP1-CT could interfere with the ability of CLIP-170 or endogenous IQGAP1 to interact with other proteins. Nevertheless, BCR-induced MTOC polarization clearly requires both IQGAP1 and CLIP-170. 168   169 Figure 4.1. IQGAP1 and CLIP-170 are required for BCR-induced MTOC reorientation  (A, B) A20 cells were transduced with lentiviruses containing the empty pGipZ vector, IQGAP1 shRNAs or CLIP-170 shRNA. The blots show IQGAP1 (A) and CLIP-170 (B) expression. The cells were mixed with anti-IgG-coated beads for 30 min and stained for pericentrin (C, E). The graphs show MTOC PIs for control versus IQGAP1 shRNA-expressing cells (D; >127 cells from three experiments) or control versus CLIP-170 shRNA-expressing cells (F; >126 cells from three experiments). The percentage of cells with a PI ≤ 0.75 is indicated (∼20% would be random distribution). (G, H) Vector control, IQGAP1 shRNA and CLIP-170 shRNA cells were stained with CMFDA and then mixed with anti-Igκ-expressing APCs for 30 min. xy confocal slices of cells stained for α-tubulin and antigen (G). White arrows indicate the MTOC. For each cell population, MTOC PIs were quantified for >81 cells from three experiments (H). The percentage of cells with a PI ≤0.75 is indicated (37.5% would be a random distribution). ****P <0.0001. Scale bars: 5 μm.     Figure 4.2. Expressing the C-terminal fragment of IQGAP1 (IQGAP1-CT) inhibits BCR-induced MTOC reorientation  A20 cells that had been transiently transfected with IQGAP1-WT-GFP or IQGAP1-CT-GFP (A) were mixed with anti-IgG-coated beads for 30 min and then stained for α-tubulin and F-actin (B). Dotted circles indicate the bead. For each cell population, MTOC polarity indices were quantified for >45 cells from three experiments and the percent of cells with PI ≤ 0.75 is indicated (C). ****P <0.0001, two-tailed unpaired t-test. Scale bar: 5 μm.  170 4.2.2 IQGAP1 and CLIP-170 are located at microtubule-actin interfaces   To test the idea that IQGAP1 and CLIP-170 mediate BCR-induced MTOC reorientation by linking the actin and microtubule cytoskeletons, it was necessary to obtain higher resolution images of the organization of the two cytoskeletons at the antigen contact site. To do this, I used super-resolution microscopy to image the actin and microtubule networks at the antigen contact site with nanometer scale resolution. Super-resolution microscopy can improve the resolution of the microtubule array in B cells (see Appendix A  B). First, I used single-color GSD/direct stochastic optical reconstruction microscopy (dSTORM) microscopy to visualize the microtubule network (Figure 4.3A). In A20 cells that were allowed to spread on anti-IgG-coated coverslips, the microtubules emanated from a single point of convergence, which represents the MTOC, and extended outwards towards the cell periphery. At the periphery, many of the microtubules curled around the edges of the cell. Using stimulated emission depletion (STED) super-resolution microscopy I then visualized the F-actin architecture in detail at the contact site with an anti-Ig-coated coverslip. In these radially spread cells, F-actin was cleared from the center of the cell while a ring of branched F-actin formed a lamellipodial structure at the cell periphery (Figure 4.3B).  171  Figure 4.3. Super-resolution images of B cell microtubules and F-actin   A20 cells were allowed to spread on anti-IgG-coated coverslips for 15 min before being fixed and stained with -tubulin antibodies and Alexa Fluor 647-conjugated secondary antibodies (A) or Alexa Fluor 488-conjugated phalloidin (B). Cells were imaged using a GSD dSTORM microscope where the data were reconstructed from 138,749 frames at 60 frames per second (A), or using a STED microscope (B). Insets on the right of each image are 2X-enlarged images of the areas in the white boxes. Scale bar: 5 µm. 172 To simultaneously visualize the relative spatial organization of the actin and microtubule networks, and assess whether the subcellular localizations of IQGAP1 and CLIP-170 could allow them to link these two cytoskeletal networks, we plated B cells on anti-Ig-coated coverslips and then imaged the cells using multi-color STED microscopy combined with confocal microscopy. The STED images showed that microtubules projected radially from a central point and that the MTOC was positioned in the lowest confocal plane of the cell (i.e. close to the coverslip) in the center of the area depleted of F-actin (Figure 4.4A,B). As seen in Figure 4.3, a number of microtubules were in the lowest plane of the cell close to the antigen contact site and extended to the cortical ring of F-actin. Upon reaching the cortical F-actin ring, some of these microtubules curled so that they extended along the interface between the peripheral actin ring and the central zone that was cleared of F-actin (Figure 4.4 and Figure 4.5).  Real-time imaging of B cells expressing CLIP-170–GFP, which marks the plus ends of microtubules, showed that CLIP-170 moved towards the periphery of the cells as they spread on immobilized anti-Ig antibody (Movie 8, Movie 9). Some CLIP-170 clusters followed similar trajectories as the microtubules that they were associated with (Movie 10). CLIP-170 extended as far as the cortical ring of F-actin, but not to the edge of the cell (Figure 4.4 and 5B), suggesting that the F-actin ring limits microtubule extension. Indeed, using latrunculin A to deplete F-actin resulted in dramatic microtubule bundling and elongation (Movie 11). This suggests that microtubules interact with the peripheral ring of F-actin and that this association regulates microtubule dynamics and organization.  To test the idea that IQGAP1 links microtubule plus ends to the peripheral F-actin ring, we asked whether F-actin, IQGAP1, and CLIP-170 were co-localized at the cell periphery. STED imaging showed that IQGAP1 was closely associated with the peripheral F-actin network 173 (Figure 4.5A) and that the F-actin and IQGAP1 were interwoven with each other at the cell periphery (Figure 4.5B). CLIP-170 was embedded in the meshwork of F-actin and IQGAP1 at the inner edge of the peripheral F-actin ring (Figure 4.5B). Moreover, proximity ligation assays (PLA) detected multiple sites where IQGAP1 and CLIP-170 were closely associated (Figure 4.6). Almost all PLA spots either co-localized with a CLIP-170 cluster (yellow, Figure 4.6A,C) or were in contact with a CLIP-170 cluster (Figure 4.6A,C). Thus, IQGAP1 and CLIP-170 are in close proximity to each other at the inner face of the peripheral F-actin network, a site where they could link microtubules to the actin cytoskeleton. 174   175 Figure 4.4. STED super-resolution images of actin, microtubules, and CLIP-170 at the antigen contact site  A20 cells that had been transfected with CLIP-170–GFP were allowed to spread on anti-IgG-coated coverslips for 15 min and imaged by STED microscopy. Cells were stained with rhodamine–phalloidin and with anti-α-tubulin plus Alexa Fluor 532-conjugated secondary antibody. CLIP-170–GFP fluorescence was imaged directly. An enlarged merged image is shown (upper panel) along with the individual channels (lower panels). Scale bar: 10 μm.     176  Figure 4.5. Co-localization of IQGAP1, CLIP-170, and F-actin  A20 cells that had been allowed to spread on anti-IgG-coated coverslips for 15 min were imaged by STED microscopy. (A) Cells were stained with rhodamine–phalloidin, anti-IQGAP1 plus an Alexa Fluor 532-conjugated secondary antibody, and anti-α-tubulin plus an Alexa Fluor 488-conjugated secondary antibody. Scale bar: 5 μm. (B) A20 cells expressing CLIP-170–GFP were stained with rhodamine–phalloidin and with anti-IQGAP1 plus an Alexa Fluor 532-conjugated secondary antibody. CLIP-170–GFP fluorescence was imaged directly. A 7X enlargement of the area in the white box is shown. All images are representative of multiple cells.    177 Movie 8. CLIP-170 moves towards the periphery of B cells spreading on immobilized anti-Ig   A20 cells that had been transfected with CLIP-170-GFP to mark the plus ends of microtubules were allowed to spread on anti-IgG-coated coverslips for 10 min and then imaged by TIRF microscopy with a 100 nm TIRF depth. Images were acquired every 30 ms. The video is played back at 75 frames per second (2.25X real time). The CLIP-170-GFP fluorescence signal is shown as a heat map in which low fluorescence intensity is purple and high fluorescence intensity is white.   http://blogs.ubc.ca/jiawangthesis/2016/05/13/chapter-2-movie-6/ Password: JWthesis   Movie 9. CLIP-170 moves towards the peripheral ring of F-actin in B cells spreading on immobilized anti-Ig   A20 cells were transfected with CLIP-170 GFP (green) to mark the plus ends of microtubules, and with LifeAct-mCherry (red) to visualize F-actin. The cells were allowed to spread on anti-IgG-coated coverslips for 10 min and then imaged for 2.25 min using a spinning disk confocal microscope. Images were acquired every 0.9 sec. The video is played back at 10 frames per second (8X real time).   https://blogs.ubc.ca/jiawangthesis/2016/05/13/chapter-2-movie-7/  Password: JWthesis   Movie 10. CLIP-170 localizes at the plus end of microtubules  A20 cells expressing tubulin-RFP and CLIP-170 GFP were allowed to spread on anti-Ig-coated coverslips for for 10 min and imaged by TIRF microscopy, with a 90 nm TIRF depth. Images were acquired every 0.5 sec for a total of 4 min. The video is played back at 20 frames per second (10X real time).   https://blogs.ubc.ca/jiawangthesis/2017/05/10/clip-170-at-the-plus-ends-of-microtubules-in-a20-cells-spreading-on-anti-igg-coated-coverslips/ Password: JWthesis 178 Movie 11. Disrupting actin filaments alters CLIP-170 localization in B cells  A20 cells that had been transfected with CLIP-170-GFP were pre-treated with 2 μM latrunculin A for 5 min. The cells were then allowed to spread on anti-IgG-coated coverslips for 10 min and imaged by TIRF microscopy, with a 100 nm TIRF depth. Images were acquired every 30 ms. The video is played back at 75 frames per second (2.25X real time). The CLIP-170-GFP fluorescence signal is shown as a heat map in which low fluorescence intensity is purple and high fluorescence intensity is white. Note that this video is from the same experiment as Movie 8, which depicts a cell from the control sample that was not treated with latrunculin A.   https://blogs.ubc.ca/jiawangthesis/2016/05/15/chapter-2-movie-8/ Password: JWthesis  179   Figure 4.6. IQGAP1 and CLIP-170 localize at the cell periphery  (A,B) RAMOS IgM+ human B cells expressing GFP-CLIP-170 were allowed to spread on anti-IgG-coated coverslips for 15 min. Mouse anti-GFP and rabbit anti-IQGAP1 primary antibodies were used for immunostaining before adding oligonucleotide-coupled donkey anti-mouse IgG and donkey anti-rabbit IgG secondary and performing the PLA reaction. Representative xy images are shown in panel A. Scale bar: 10 μm. Each PLA signal (red) is presumably indicative of one interaction event. The graph in panel B shows the number of PLA spots (red) per cell for >11 cells for cells that had been stained with both anti-GFP and anti-IQGAP1 (PLA), and for control samples in which cells were stained with secondary antibodies (Ab) only or with anti-IQGAP1 only or anti-GFP only plus the secondary antibodies. ****P <0.0001 using unpaired two-tailed t-test. (C) RAMOS cells were allowed to spread on anti-IgG-coated coverslips for 45 min before immunostaining with primary antibodies, staining F-actin with Alexa-647-phalloidin, adding secondary antibodies, and carrying out the PLA reaction. Representative xy images are shown. Scale bar: 10 μm. Areas indicated by the white boxes in panels A and C were enlarged 4-fold. Yellow spots in the enlarged images indicate co-localization of the red PLA signal and the green CLIP-170-GFP signal. Images are representative of multiple experiments.   180 4.2.3 Rap-dependent actin reorganization promotes IQGAP1 accumulation at the periphery of the immune synapse   Because IQGAP1 is strongly associated with the F-actin at the periphery of the IS (Figure 4.4B,C), and Rap1 controls actin organization at the IS [55, 247], we asked whether Rap1 promotes the accumulation of IQGAP1 at antigen contact sites. Within 5 min of mixing primary B cells with anti-Ig-coated beads (Figure 4.7A,B) or anti-Igκ-expressing APCs (Figure 4.7E), both F-actin and IQGAP1 accumulated at the antigen contact site. When F-actin was depleted using latrunculin A, IQGAP1 was not present at the cell cortex and did not accumulate at the bead contact site (Figure 4.7A). This suggests that the localization of IQGAP1 is largely determined by its interaction with F-actin. Blocking Rap1 activation prevented the accumulation of both F-actin and IQGAP1 at the contact site with anti-Ig-coated beads (Figure 4.7B-D). Similarly, Rap1 knockdown prevented IQGAP1 accumulation at the B cell-APC contact site (Figure 4.7E). In both cases, IQGAP1 remained distributed uniformly around the cell cortex (Figure 4.7B,E).   The effect of Rap1 activation on the localization of IQGAP1 could be seen most clearly when B cells were allowed to spread on anti-Ig-coated coverslips. In control A20 cells, IQGAP1 localized to the peripheral F-actin ring and was depleted from the center of the cell where F-actin had been cleared (Figure 4.4B,C and Figure 4.7F,G). In RapGAPII-expressing A20 cells, which failed to spread, F-actin and IQGAP1 were not cleared from the center of the antigen contact site (Figure 4.7F,G). Nevertheless, as judged by calculating the Pearson and Manders co-localization coefficients, IQGAP1 still co-localized extensively with F-actin when Rap activation was blocked (Figure 4.7H,I). Thus, the colocalization of IQGAP1 and F-actin is not dependent on Rap1 or cofilin. Instead, our data support the idea that the Rap1–cofilin pathway promotes the 181 accumulation of IQGAP1 at the periphery of antigen contact sites by controlling the organization of the actin network. Consistent with the idea that IQGAP1 acts downstream of Rap1-dependent actin reorganization to promote MTOC polarization, depleting IQGAP1 did not inhibit B cell spreading or F-actin clearance from the center of the antigen contact site (Figure 4.8).  182    183 Figure 4.7. Rap1 promotes IQGAP1 accumulation at the IS by controlling actin organization  (A) A20 cells were treated with DMSO or 2 μM latrunculin A for 5 min, then mixed with anti-IgG-coated beads for 5 min. Cells were stained for IQGAP1, α-tubulin and F-actin. Confocal xy slices of bead:cell conjugates are shown. Dotted circles indicate the bead. (B–D) Vector control and RapGAPII-expressing A20 cells were mixed with anti-IgG-coated beads for 5 min and then stained for IQGAP1 and F-actin. For each conjugate, the corrected fluorescence intensity of F-actin (C) and IQGAP1 (D) within the white circle was quantified (in arbitrary units, AU) for >65 cells from two experiments. (E) LPS-activated primary B cells were transfected with control siRNA, or with Rap1a and Rap1b siRNAs, mixed with anti-Igκ-expressing APCs for 5 min, and then stained for IQGAP1, F-actin and antigen. Arrows show the contact site between the B cell and the APC. (F–I) Parental, vector control and RapGAPII-expressing A20 cells were allowed to spread on anti-IgG-coated coverslips for 15 min, then stained for IQGAP1 and F-actin and imaged by TIRFM or confocal microscopy. TIRFM images (F) and fluorescence profiles along the dotted lines are shown for representative cells (G). Confocal images were used to calculate the Pearson (H) and Manders (I) coefficients for co-localization of IQGAP1 and F-actin. ****P <0.0001. Scale bars: 5 μm.  184   Figure 4.8. IQGAP1 is dispensable for B cell spreading and actin reorganization  A20 cells were transduced with lentiviruses containing either the empty pGipZ vector or IQGAP1 shRNAs and then allowed to spread for 30 min on anti-IgG-coated coverslips. Cells were fixed and stained for F-actin using Alexa Fluor 647-phalloidin. (A) Representative images are shown. Scale bar: 10 μm. (B) Cell spreading area was quantified for >86 cells. The mean ± s.d. is shown. The mean area for IQGAP1 shRNA-expressing cells was not significantly different than that for the vector control cells, as determined using a two-tailed unpaired t-test.  185 4.2.4 BCR-induced MTOC polarization requires dynein activity  In T cells, dynein is essential for the movement of the MTOC to the IS [210, 218, 226, 305]. The minus-end-directed movement of dynein along microtubules, which is driven by the ATPase activity of dynein, allows membrane- or cortically-anchored dynein to reel in the microtubule network and move the MTOC towards the cell membrane. Dynein links BCR microclusters to juxtamembrane microtubules [190] but it is not known whether dynein activity is required for MTOC polarization in B cells. We found that treating A20 cells with either erythro-9-[3-(2-hydroxynonyl)] adenine (EHNA) ([343] or ciliobrevin D [344], both of which inhibit the ATPase activity of dynein, blocked BCR-induced MTOC polarization to the same extent as using nocodazole to depolymerize microtubules (Figure 4.9). Although EHNA concentrations 25–200 times greater than what we used have profound effects on actin organization [345], F-actin accumulation at the bead contact site occurred normally in B cells treated with the concentrations of dynein inhibitors that we used (Figure 4.9C).    186  Figure 4.9. Dynein is required for MTOC polarization  (A,B) A20 cells were treated with DMSO, 10 μM EHNA, or 5 μM nocodazole for 30 min and then mixed with anti-Ig-coated beads for 30 min. Representative confocal images of pericentrin and DAPI staining (A) are shown along with MTOC PIs for >30 cells (B). (C,D) A20 cells were treated with DMSO, 10 μM EHNA, or 20 μM ciliobrevin D for 40 min and then mixed for 30 min with 4.5-μm diameter beads. The beads were immunostained with Alexa Fluor 647-conjugated secondary antibodies. Representative xy confocal slices of pericentrin-stained cells (C, upper panels; dotted circles indicate the periphery of the cell) are shown along with 3D reconstructions of cells that had been stained for F-actin (C, lower panels). MTOC PIs for >33 cells are shown (D). The percentage of cells with a PI ≤0.75 is indicated in blue (~20% would be a random distribution). ***P <0.0001. Scale bars: 5 μm.  187 4.3 Discussion 4.3.1 Summary of findings In this chapter, I showed that IQGAP1 and the microtubule plus-end binding protein CLIP-170 are important for BCR-induced MTOC reorientation towards particulate antigens and APCs. Linking microtubules to the cortical actin cytoskeleton is essential for controlling microtubule organization and dynamics [228, 324]. IQGAP1 contains both an F-actin-binding domain and a CLIP-170-binding domain [228, 239]. My results are consistent with a model in which IQGAP1 and CLIP-170 bridge the actin and microtubule cytoskeletons, and link BCR-induced actin reorganization to MTOC polarization. I found that IQGAP1 and CLIP-170 were located in close proximity at the periphery of the IS. Furthermore, CLIP-170 was embedded in a meshwork of F-actin and IQGAP1. In Chapter 3 I showed that the Rap1-cofilin pathway coordinates the regulation of actin cytoskeleton remodeling with MTOC polarization in B cells. The findings in this chapter show that the subcellular localization of IQGAP1 is controlled by F-actin and that Rap1-dependent actin remodeling is important for IQGAP1 to accumulate at the IS where it can link the actin and microtubule cytoskeletons. Additionally, we also demonstrate that dynein is required for MTOC reorientation towards the B cell IS.   4.3.2 IQGAP1 localizes to the antigen contact site In migrating fibroblasts, the association between IQGAP1 and CLIP-170 at the leading edge increases microtubule interactions with the cell cortex, which supports MTOC polarization during directed migration [228, 338]. Consistent with these findings, our data show that IQGAP1 accumulates at the IS, a polarized structure that is similar to the leading edge of a migrating cell, and that IQGAP1 is required for BCR-induced MTOC reorientation. I also showed that when B 188 cells spread on anti-Ig-coated coverslips, IQGAP1 co-localized with F-actin at the periphery of the antigen contact site, where it was interwoven with actin structures and CLIP-170. This is consistent with previous findings in macrophages where IQGAP1is localized at the actin-rich phagocytic cup [346], a structure that is similar to an IS. Additionally, we found that the localization of IQGAP1 in B cells was linked to its association to F-actin. When we used latrunculin A to disrupt the actin cytoskeleton, IQGAP1 no longer associated with the cell cortex. This argues that the cortical actin cytoskeleton is responsible for the normal cortical distribution of IQGAP1. The finding that IQGAP1 accumulation at adherens junctions depends on F-actin and is ablated by actin-disrupting drugs [347] also supports the idea that F-actin controls the subcellular localization of IQGAP1. Importantly, our data showed that actin reorganization mediated by the Rap1-cofilin pathway promotes the accumulation of IQGAP1 at the periphery of antigen contact site. Moreover, IQGAP1 and F-actin co-localized even when Rap1 was inhibited, indicating that the subcellular localization of IQGAP1 is controlled primarily by F-actin, and hence only indirectly by Rap.  Our data are consistent with a model in which IQGAP1 is enriched at the peripheral F-actin ring, which forms in a Rap- and cofilin-dependent manner, and promotes MTOC polarization by binding to CLIP-170 and capturing microtubule plus ends. By directing actin organization, and thus the subcellular localization of IQGAP1, Rap1 may determine the sites at which IQGAP1 links the actin and microtubule networks at the IS. It is also possible that IQGAP1 is recruited to the IS via the ability of its IQ domain to bind active Rap [236]. Co-localization of Rap and IQGAP1 at the plasma membrane of MCF-7 epithelial cells has been demonstrated by fluorescence microscopy [236]. However, these studies only demonstrated a direct interaction in vitro and did not determine whether the subcellular colocalization was a 189 direct interaction, as opposed to being mediated by F-actin. Activated Rac1 and Cdc42 may also contribute to IQGAP1 localization at the IS or at the leading edge of migrating cells, as has been shown in fibroblasts [228]. To test the role of these Rho family GTPases in recruiting IQGAP1 to the B cell IS, one could test whether siRNAs or chemical inhibitors targeting Rac1 or Cdc42 prevent the accumulation of IQGAP1 at the contact site with anti-Ig-coated beads. However, Cdc42 and Rac1 are also important for actin dynamics and any resulting mislocalization of IQGAP1 could be a consequence of altering actin organization.  Although the actin cytoskeleton is important for IQGAP1 localization to the IS, our data show that IQGAP1 does not in turn control actin reorganization at the B cell IS. BCR-induced cell spreading and F-actin reorganization at the antigen-contact site were not altered when IQGAP1 was depleted. This suggests that in B cells IQGAP1 acts downstream of F-actin to regulate MTOC reorientation. IQGAP1 also localizes to the NK cell IS and is required for MTOC polarization and the polarization of cytotoxic granules towards target cells [217, 229]. In contrast, IQGAP1 is dispensable for MTOC polarization in T cells but instead regulates actin organization [341]. Depleting IQGAP1 causes excessive F-actin accumulation at the T cell IS and increases cell spreading. Thus, IQGAP1 may have different roles in cytoskeleton remodeling at the IS in different types of lymphocytes.   4.3.3 IQGAP1 may facilitate MTOC movement by capturing microtubule plus ends at the cell cortex The microtubule network is highly dynamic where microtubules can undergo cycles of polymerization and depolymerization in a process called dynamic instability. Therefore, a key question is how does the MTOC become reoriented in response to external stimuli given the 190 dynamic nature of microtubules? Microtubules can interact with the cell periphery or cellular organelles (e.g. Golgi apparatus. nucleus, etc.) and become transiently stabilized via these interactions. Movement of the MTOC occurs due to forces exerted on microtubules that have been captured and stabilized. Several mechanisms by which cortical capture can induce MTOC reorientation towards the IS have been proposed. As discussed in Chapter 1, these include the “cortical sliding”, “end-on capture shrinkage” and cell spreading models. In all of these models, cortical capture mechanisms tether microtubules to the cell cortex so that force-generating mechanisms can then pull the MTOC towards the IS. My findings are consistent with a model in which IQGAP1 and CLIP-170 link the actin and microtubule cytoskeletons at the cell cortex and mediate MTOC reorientation during B cell spreading. Outward forces that are generated by Rap1- and cofilin-dependent B cell spreading may reorient the MTOC towards the IS. However, my finding that dynein activity is also critical for BCR-induced MTOC polarization to the IS suggests that cortically anchored dynein motor complexes may also exert pulling forces on captured microtubules. Therefore, several mechanisms may act in concert are essential for the MTOC to move the MTOC towards the IS. In the next section I discuss in detail the cortical capture and force-generating mechanisms that may be involved in repositioning the MTOC to the B cell IS.   4.3.3.1 The cortical capture of microtubule plus ends  The coordinated regulation of microtubule capture and force-generating mechanisms is important for repositioning the MTOC within the cell. The interaction of microtubule plus ends with cortical capture proteins can overcome microtubule-depolymerization signals so that the captured microtubules are transiently stabilized. This stabilization of anchored microtubules 191 allows forces to then be transmitted along the microtubule network in order to move the MTOC. CLIP-170 is a +TIP that can promote the stabilization of microtubules via its interactions with the cell cortex [196]. We found that CLIP-170 is required for BCR-induced MTOC reorientation, perhaps through its capture at the periphery of the IS by IQGAP1. Indeed, my microscopy images showed that CLIP-170 at the plus ends of microtubules localized to the periphery of the antigen contact site to interact with the actin cytoskeleton and IQGAP1. Additionally, when we expressed the IQGAP1-CT mutant, which acts in a dominant negative fashion to prevent endogenous IQGAP1 from binding to CLIP-170, MTOC reorientation was abrogated. Although our results demonstrate a critical role for IQGAP1 and CLIP-170, we cannot exclude the contribution of other proteins in capturing microtubules during MTOC reorientation in B cells. Proteins that are involved in cortical capture include +TIPs that decorate the plus ends of microtubules (e.g. CLASPs, APC, CLIP proteins, EBs), scaffolding proteins that can link the actin and microtubule cytoskeletons (e.g. spectraplakins, IQGAP1, Diaphanous proteins such as mDia), and dynein/dynactin motor complexes that can bind +TIPs and generate pulling forces (reviewed in [348]).  The plus ends of microtubules are often bound by complexes consisting of multiple plus-end binding proteins that interact directly with microtubules or with other +TIPs [196, 208]. Most +TIPs are large multi-domain and multivalent proteins. This feature allows +TIPs to interact with one another in a dynamic fashion. Proteins that interact with CLIP-170 include EBs, SLAIN2, CLASPs, and LIS1/NDE1/dynein complexes [349]. Because different CLIP-170-associated +TIPs may have distinct functions, I discuss below how these proteins might contribute to MTOC reorientation in B cells.  CLIP-170 does not interact directly with microtubule plus ends but instead binds to other 192 +TIPs via its CAP-Gly domain [196]. EBs, which are autonomous tip trackers, bind directly to microtubules and recruit CLIP proteins to the microtubule plus ends. SLAIN2, which binds to EBs and has multiple +TIP-binding domains, may play a similar role in recruiting CLIP-170 and other +TIPs to the growing ends of microtubules [196]. Once it is associated with the plus ends, CLIP-170 recruits CLASP1 and CLASP2 and these proteins enhance the microtubule-stabilizing function of CLIP-170 [196]. In migrating cells, CLASPs are preferentially associated with microtubule plus ends at the leading edge of the cell, where they establish a polarized network of stable microtubules [196, 350] by binding to IQGAP1 at the leading edge [351]. This asymmetric distribution of CLASPs is important for vesicular transport to the leading edge during cell motility and is regulated by polarity signals at the leading edge. Because the IS mimics the structure of a leading edge, it is possible that CLASPs preferentially associate with the plus ends of microtubules that extend towards the antigen contact site. Indeed, BCR signaling inactivates the CLASP inhibitor, glycogen synthase kinase-3β (GSK3β) [350-353].  The binding of CLASPs to CLIP-170 can also mediate the cortical capture of microtubule plus ends, further enhancing microtubule stabilization by CLIP-170/CLASP complexes. CLASPs that are bound to CLIP-170 mediate the cortical capture of microtubules by binding to IQGAP1 as well as LL5β, a protein that binds the membrane phospholipid PIP3 [351, 354]. Whether these other +TIPs that interact with CLIP-170 play a role in MTOC reorientation in B cells could be addressed using loss-of-function approaches such as siRNA-mediated knockdown. In T cells, the interaction between CLIP-170 and the adaptor proteins lissencephaly 1 (LIS1) and NDE1 has recently been shown to be important for targeting microtubule plus ends to dynein motor complexes, the actions of which are essential for TCR-induced MTOC reorientation [355]. We showed that dynein is required for MTOC reorientation in B cells but the role of LIS1 and NDE1 193 in B cells has not been investigated. I discuss below their possible role in generating the forces that are required for MTOC polarization. In addition to IQGAP1, other actin-associated proteins can mediate the cortical capture of microtubule plus ends and thereby link the actin and microtubule cytoskeletons. In Drosophila, the actin-dependent motor protein myosin VI interacts with CLIP-190, which is the homolog of CLIP-170 [356]. Another cytoskeletal protein is the spectraplakin family member actin crosslinking factor-7 (ACF7), a scaffolding protein that can bind both microtubule plus ends and actin, and guides the growth of microtubules along actin bundles [196, 357-360]. In endodermal cells lacking ACF7, microtubules fail to grow along polarized actin bundles even though EB1 and CLIP-170 still localize to the plus ends of microtubules. In these ACF7-null cells, the microtubule network exhibits altered dynamic instability, the microtubules do not pause at cortical sites, and the cells fail to maintain cell polarity, which is required for directional cell migration [358]. Thus, as I showed in B cells, linking the actin and microtubule cytoskeletons so that they can be regulated in a coordinated manner is important for establishing cell polarity.  Real-time TIRFM imaging in fibroblast cells showed that microtubules tend to polymerize and grow towards the cell periphery along a similar trajectory as a predecessor microtubule [337, 361]. Live-cell imaging of CLIP-170-GFP, which marks the growing plus ends of microtubules, suggested that microtubules can grow along pre-existing tracks in spreading B cells.  The occurrence of such a phenomenon points towards a guiding system that directs microtubules towards specific sites. Indeed, the growth of microtubules towards specific sites at the cell cortex can be guided by actin filaments, with the aid of proteins such as ACF7 that can link microtubules to actin filaments [358]. STED imaging of B cells spreading on immobilized anti-Ig showed that in addition to the peripheral ring of dendritic F-actin there were 194 thin actin filaments that traversed the actin-poor region at the center of the antigen contact site. ACF7, or other actin-microtubule crosslinking proteins that are bound to these thin actin filaments, could provide the tracks that support microtubule growth towards the peripheral actin ring, where they can be captured. STED imaging of B cells expressing fluorescent ACF7 and tubulin, followed by staining F-actin with phalloidin, could reveal whether ACF7 was localized along thin actin filaments at sites where microtubules were closely associated to these filaments. siRNA-mediated depletion of ACF7 could be used to test whether ACF7 is important for delivering microtubules to cortical sites in B cells and, consequently, for MTOC polarization towards the antigen contact site. Tracking microtubule plus ends and examining the length of time that they spend interacting with the cell cortex in wild type versus ACF7-deficient cells would be a further test of whether ACF7 is important for capturing microtubules at the periphery of B cells that are spreading on anti-Ig-coated coverslips.  4.3.3.2 Force generation promotes MTOC reorientation and polarization Movement of the MTOC towards the center of the IS is mediated by forces that are exerted on the microtubules attached to it. When B cells spread across antigen-bearing surfaces, the outward movement of the peripheral F-actin ring could generate pulling forces on microtubules that are anchored to the cell cortex via IQGAP1 and CLIP-170. A similar force-generating mechanism is important for the appropriate positioning of the MTOC during ciliogenesis, where the centrosome moves towards the site of actin-dependent cell spreading [362]. In addition to actin-dependent MTOC reorientation that is driven by cell spreading, we also identified a critical role for dynein during BCR-induced MTOC reorientation, as has been shown in T cells [210, 218, 226, 305]. Dynein motor movement is the major force-generating 195 mechanism in both the “cortical sliding” and the “end-on capture shrinkage” models of MTOC reorientation [210, 211, 226]. Dynein can be recruited to the antigen contact site [211, 226] by binding to TCR [215] or BCR microclusters [190], and perhaps to other proteins that are located at the IS.  Dynein/dynactin complexes are often found at the plus ends of growing microtubules and the recruitment of dynein to microtubule plus ends is one of the major functions of CLIP-170 [351]. The Cap-Gly domain of p150glued, a component of dynactin, can bind to the zinc knuckle domain of CLIP-170 [350, 363]. This interaction may be controlled by LRRK1-mediated phosphorylation of CLIP-170 at T1384 within the zinc knuckle motif, which promotes the binding of p150glued [364]. CLIP-170 may also recruit dynein by binding to the dynein-associated protein LIS1, which promotes the attachment of dynein to microtubules and increases its motor function [350, 365]. Thus, the requirement for CLIP-170 in BCR-induced MTOC polarization could reflect both its ability to bind cortical IQGAP1 and its ability to recruit dynein to the cell cortex, where dynein can support the cortical capture of microtubules and exert forces that reposition the MTOC. Dynein forms two distinct complexes, one with NDE1 and LIS1, and another with the dynactin complex that contains p150Glued [366]. These two dynein-containing complexes may have distinct functions. In T cells, the NDE1/LIS1 complex is required for dynein to be recruited to the periphery of the T cell IS and for MTOC translocation whereas p150Glued is required for the directional secretion of cytotoxic granules [355]. In B cells, the p150Glued-containing dynein/dynactin complex binds to BCR microclusters via the adaptor proteins Cbl, Grb2 and Dok-3 and mediates their coalescence into a cSMAC [190]. Whether this requires LRRK1-mediated phosphorylation of CLIP-170 to promote the binding of p150Glued to microtubule plus ends is not known. Conversely, based on its role in T cells, it would be 196 interesting to test the hypothesis that it is the alternative dynein/NDE1/LIS1 complex that mediates the cortical capture of microtubule plus ends via CLIP-170 and supports MTOC polarization in B cells.   4.3.4 Perspectives  In B cells, MTOC reorientation is required for the formation of the IS, which supports both B cell signaling as well antigen acquisition from APCs, both of which are required for B cell activation. Therefore, it is likely that defects in IQGAP1 or CLIP-170 would impair multiple steps in B cell activation.  Impaired MTOC reorientation would prevent IS formation and could therefore decrease the amount of BCR signaling initiated by low densities of APC-bound antigen to the point that it may not exceed the threshold for BCR activation. Defective IQGAP1/CLIP-170-mediated MTOC reorientation could also dramatically reduce the ability of B cells to acquire antigens from APCs. Indeed, in Chapter 6 I show that the loss of CLIP-170 impairs the ability of B cells to acquire antigen from APCs. The acquisition of membrane-bound antigens occurs only after IS formation in naïve B cells [87] and often requires the reorientation of the MTOC to the antigen contact site [85, 109]. Therefore, the generation of plasma cells and memory B cells producing high affinity, class-switched antibodies is likely to be impaired when IQGAP1 and CLIP-170-dependent MTOC reorientation in naïve B cells is impaired. Additionally, even naïve B cells that are able to acquire antigens (e.g. soluble antigens, ICs) when MTOC reorientation is defective, may not form functional contacts with T helper cells as the reciprocal MTOC polarization of B and T cells towards one another is important for the delivery of T cell help [224].  197 Examining the roles of IQGAP1/CLIP-170-dependent MTOC reorientation in vivo using knockout mice in the context of an infection is an important extension of the work in this chapter and would reveal whether these proteins impact initial B cell activation as well as the generation of class-switched antibodies and memory B cells. Although the deletion of CLIP-170 in mice does not result in growth or developmental defects [350], immune cell populations were not examined in these knockout mice (Anna Akhmanova, Univ. of Utrecht, personal communication). IQGAP1 knockout mice display subtle defects in immune cell development [230] and T cells from these mice have aberrant signaling phenotypes in vitro [341]. However, BCR-induced IS formation, in vitro B cell activation by APC-bound antigens, and in vivo B cell activation by T-dependent and T-independent antigens have not been examined in these mice.  Although they are ubiquitously expressed, IQGAP1, CLIP-170, or associated proteins could be effective targets for drugs to limit B cell responses that contribute to autoimmunity. Indeed, in the experimental autoimmune encephalomyelitis (EAE) mouse model of multiple sclerosis, treating the mice with the microtubule-stabilizing drug paclitaxel delayed and reduced the incidence of autoimmune disease [367]. However, these pan-microtubule drugs act on all cell types and can induce global cell cycle arrest and off-target effects. In contrast, the absence of developmental and growth defects in mice that are deficient for either IQGAP1 and CLIP-170 suggests that targeting these proteins would have limited toxicity but could selectively block specific cellular responses such as B cell activation.   198 Chapter 5: HS1 regulates BCR-induced actin reorganization and IS formation  5.1 Introduction Actin cytoskeleton remodeling acts as a critical determinant of B cell activation by supporting the formation of the IS. BCR-induced actin cytoskeleton reorganization promotes BCR microcluster formation, B cell spreading, and IS formation. B cells that bind to cognate antigen initiate cofilin-mediated actin severing, which increases the mobility of BCRs and allows the formation of signaling BCR microclusters [55, 56]. Concomitant Arp2/3 activation drives actin polymerization, resulting in the the formation of dendritic/branched actin structures at the periphery of the B cell:APC contact site [55, 82, 247]. This provides pushing forces (see section 1.5.3) on the cell membrane so that the B cell spreads over the surface of the APC. B cell spreading increases the contact area with the APC, which is important for maximizing the chance that BCRs bind to antigen [64]. The actin reorganization that occurs during B cell spreading also promotes reorientation of the MTOC towards the site of antigen contact [82]. The coalescence of BCR microclusters into a cSMAC requires both the retrograde flow of the peripheral branched actin structures and MTOC reorientation, which allows juxtamembrane microtubules to act as tracks for dynein-mediated centralization of the BCR microclusters [80-82, 190]. Therefore, the dendritic actin network that forms at the B cell:APC contact site is important for multiple events during B cell IS formation.  HS1 is a homologue of the actin-binding protein cortactin that is expressed primarily in hematopoietic cells (see Section 1.5.3.1). Cortactin often localizes to sites of active actin assembly such as lamellipodia, endosomes, podosomes and invadopodia where it binds to and 199 promotes Arp2/3-dependent branched actin polymerization [368]. Like cortactin, HS1 also binds both F-actin and the Arp2/3 complex and is thought to stimulate Arp2/3-mediated actin polymerization and stabilize the resulting actin structures [139, 142, 148]. Consistent with its role in actin-dependent processes, HS1 is important for NK cell migration [145]. In T cells, HS1 is important for the accumulation of F-actin at the T cell IS and for sustaining the lamellipodial actin networks that drive T cell spreading [142]. In response to TCR signaling, HS1 is rapidly tyrosine phosphorylated by SFKs on Y378 and Y397 and recruited to the T cell:APC contact site [139, 142]. Phosphorylated HS1 can recruit the Rac/Cdc42 GEF Vav1 to the IS where it promotes actin polymerization [142, 319].  HS1-deficient mice appear to develop normally, but mature B cells fail to undergo antigen-induced B cell proliferation and apoptosis [154]. These mice also exhibit impaired antibody responses to T-dependent antigens [154]. These findings suggest that HS1 is involved in BCR signaling. HS1 is phosphorylated in response to BCR stimulation [369]. Additionally, when B cells bind anti-Ig-coated beads, HS1 localizes to the antigen contact site and recruits the Arp2/3 complex [136]. However, the role of HS1 in mediating actin remodeling at the B cell IS has not been studied. Because HS1 regulates actin dynamics in T cells in response to TCR signaling, I hypothesized that HS1 is important for the BCR-induced actin-reorganization that supports B cell spreading and IS formation. In this chapter I show that HS1 is important for B cell spreading, for the coalescence of BCR microclusters into a cSMAC, and for microcluster-based BCR signaling at the IS.    200 5.2 Results 5.2.1 BCR-induced tyrosine phosphorylation of HS1 depends on the Rap GTPases The phosphorylation of human HS1 on Y378 and Y397 is required for it to promote Arp2/3-mediated actin polymerization [139]. Consistent with a previous report showing that BCR stimulation induces the tyrosine phosphorylation of HS1 in human B cells [139], I found that treating the mouse A20 B cell line with soluble anti-Ig antibodies resulted in phosphorylation of HS1 on Y405, the equivalent of Y397 in human HS1 (Figure 5.1). Because the Rap GTPases are major regulators of actin dynamics in B cells, I asked whether Rap activation was important for BCR-induced phosphorylation and activation of HS1. To test this, I used A20 cells that overexpressed RapGAPII, which converts Rap to its inactive GDP-bound form [247]. Compared to control cells, I found that blocking Rap activation resulted in slightly reduced levels of HS1 phosphorylation at 2 min and, more strikingly, a failure to sustain HS1 phosphorylation past 5 min (Figure 5.1B). Therefore, BCR-induced HS1 phosphorylation, especially at later times, is at least in part dependent on Rap activation.  201   Figure 5.1. HS1 is phosphorylated upon BCR stimulation with soluble anti-Ig antibodies   A20 cells (A) as well as A20 cells that were stably transfected with either the empty MSCV vector or MSCV-RapGAPII (B) were stimulated with 20 µg/ml of anti-IgG for the indicated times. Cell lysates were probed with antibodies specific for HS1 that is phosphorylated at Y397 (P-HS1; upper panels) or total HS1 (lower panels). The blots shown are representative of 3 independent experiments. The topmost band in the P-HS1 blots aligns with the major band detected in the total HS1 blots.   202 5.2.2 Phosphorylated-HS1 is enriched at the antigen contact site The subcellular localization of HS1, and in particular P-HS1, may be important for it to regulate F-actin dynamics at the IS [142]. In T cells, P-HS1 is enriched at the IS and its phosphorylation on Y378 and Y397 is required for it to assume this polarized distribution [142]. Importantly, in HS1-null T cells, F-actin fails to accumulate at the contact site with an APC [142]. Because HS1 is phosphorylated upon BCR stimulation (Figure 5.1), I asked whether P-HS1 was also enriched at the B cell IS. To assess the subcellular localization of HS1, I expressed HS1-YFP in A20 cells. In resting cells, HS1-YFP was distributed throughout the cell but was somewhat enriched at membrane folds and ruffles, which are presumably F-actin-rich structures (Figure 5.2A, Movie 12). When A20 cells bound to anti-IgG-coated beads, HS1-YFP was still distributed throughout the cell (Figure 5.2B). However, immunostaining showed that P-HS1 accumulated at the bead contact site along with F-actin, and that it was largely absent from the remainder of the cell (Figure 5.2C). The accumulation of active, phosphorylated HS1 at the antigen contact site suggests that it could play a role in IS formation or function.   203  Figure 5.2. Phosphorylated HS1 is localized at the antigen-contact site with anti-Ig-coated beads  (A) Representative maximum intensity projection image of a resting A20 cell stably expressing HS1-YFP. (B) A20 cells stably expressing HS1-YFP were mixed with anti-IgG-coated beads for 5 min before being fixed and immunostained with Alexa Fluor 647-conjugated anti-goat IgG to visualize the beads. Representative 3D reconstruction images are shown. (C)  A20 cells were mixed with anti-IgG-coated beads for 5 min. Cell:bead conjugates were immunostained with anti-P-Y397-HS1 antibodies and with rhodamine-phalloidin to visualize F-actin. A representative image is shown. Scale bar: 5 µm. 204 Movie 12. HS1-YFP distribution with anti-Ig stimulation  Three-dimensional reconstruction of an A20 cell stably expressing HS1-YFP (green) bound to an anti-Ig-coated bead (red) for 5 min. The same cell is shown in Figure 5.2B, upper left panel.  https://blogs.ubc.ca/jiawangthesis/2017/07/24/movie-hs1-yfp-3d/ Password: JWthesis   To examine the localization of P-HS1 at the antigen contact site in more detail, A20 cells were allowed to spread on anti-Ig-coated coverslips in order to generate a larger and more stable antigen contact site. As observed previously, the F-actin at the contact site was reorganized into a peripheral ring and was largely cleared from the center of the cell. P-HS1 exhibited a similar distribution and co-localized extensively with the peripheral F-actin network (Figure 5.3). The Pearson coefficients for 5 min (0.8) and 15 min (0.77) showed high overlap between actin and HS1 staining (Figure 5.3). Furthermore, the overlap of HS1 with F-actin remained consistently high as indicated by the Manders’ coefficients (0.83) (Figure 5.3). Manders’ coefficients also showed that although there was high overlap of F-actin with HS1 at 5 min (0.7), this overlap decreased at 15 min (0.53). This suggests that while most of the HS1 is associated with F-actin structures during B cell spreading, some actin structures that formed at 15 min did not contain HS1.    205  Figure 5.3. Phosphorylated HS1 is localized at the cell periphery in cells spreading on immobilized anti-Ig antibodies  A20 cells were allowed to spread on anti-IgG-coated coverslips for 5 min (A) or 15 min (B). The cells were stained with P-HS1 antibodies and with rhodamine-phalloidin to visualize F-actin. Representative confocal xy images are shown. Scale bar: 10 µm. The Pearson’s coefficient (C) and Manders’ coefficients (D) were determined for >33 cells from one experiment.  206 5.2.3 The localization of HS1 at the cell periphery depends on F-actin In B cells forming contacts with anti-Ig-coated beads, both HS1 and F-actin were enriched at the bead contact site (Figure 5.2). Likewise, in B cells spreading on anti-Ig-coated coverslips, HS1 was localized at the periphery of the antigen contact site, where it overlapped extensively with F-actin (Figure 5.3). Therefore, I asked whether the localization of HS1 was controlled by F-actin and actin dynamics. To do this, I stably expressed HS1-YFP in A20 cells and allowed them to spread on anti-Ig-coated coverslips. As the cells spread, the HS1-YFP formed punctate structures at the periphery of the cell that moved inwards towards the center of the cell (Movie 13). This peripheral distribution and retrograde movement of HS1-YFP was abolished when the actin cytoskeleton was depleted using latrunculin A and HS1 rapidly became dispersed throughout the cell (Movie 13). Hence, the subcellular distribution and movement of HS1 at the periphery of spreading B cells is consistent with its role in organizing the peripheral actin cytoskeleton    Movie 13. HS1 dynamics are controlled by an intact actin cytoskeleton  A20 cells expressing HS1-YFP were allowed to spread on anti-IgG-coated coverslips for 15 min. Cells were then imaged using TIRF microscopy at a 90 nm TIRF depth where 1 frame was captured every 0.5 s. After 50 s of imaging, latrunculin A was added to a final concentration of 1 µM. Video is played back at 29 frames per second at 14X real time.  https://blogs.ubc.ca/jiawangthesis/2016/09/22/hs1-yfp-lata/ Password: JWthesis  207 5.2.4 HS1 is important for BCR signaling at the immune synapse  Effective B cell activation requires the reorganization of the actin cytoskeleton and the formation of an IS. BCRs that engage membrane-bound antigen on the surfaces of APCs aggregate into microclusters that recruit signaling proteins and assemble a signalosome that activates multiple downstream signaling pathways [370]. By regulating the activity of cofilin and ezrin [50, 53, 55], initial BCR signaling removes actin-based diffusion barriers that limit BCR mobility.  This allows increased BCR aggregation into microclusters, and hence increased BCR signaling. Additionally, actin remodeling at the cell periphery propels the cell membrane outwards so that the B cell can contact more cognate antigen, which further increases B cell signaling [44, 53, 64]. To test whether HS1 is important for BCR microcluster formation and BCR signaling, I generated A20 and A20/D1.3 cell lines that were stably transduced with an HS1 shRNA. Single cells were seeded into individual wells to isolate HS1 shRNA clones, which were identified by immunoblotting (Figure 5.4). Compared to control cells (scrambled shRNA sequence), cells in which HS1 was partially depleted gathered less antigen into clusters at the B cell:APC contact site (Figure 5.5A,B), similar to what was observed in A20/D1.3 cells that were transiently transduced with HS1 siRNA (Figure 5.6). Importantly, proximal BCR signaling was lower in HS1-depleted B cells than control cells, as quantified by the amount of phosphorylated CD79a and CD79b (collectively referred to as P-CD79) that was detected by immunostaining (Figure 5.5A,C). The amount of P-CD79 signal was directly proportional to the amount of antigen that was gathered (Figure 5.5D). Thus, for membrane-bound antigens, HS1 is required for maximal BCR-mediated antigen accumulation at the IS and for the resulting early BCR signaling events.   208   Figure 5.4. Clones of A20 and A20/D1.3 cells transduced with HS1 shRNA   Single cells were isolated from a mixed population (HS1 shRNA3 cells). HS1 levels in A20 clones (A) and A20/D1.3 clones (B) were assessed by immunoblotting. β-actin was used as a loading control. “Scrambled control” refers to a population of cells transduced with the scrambled sequence.  209   Figure 5.5. CD79 phosphorylation at the antigen contact site is dependent on HS1  (A) A20 cells expressing either the control scrambled sequence or HS1 shRNA were added to Cos-7 cells expressing the single chain anti-κ light chain antibody (i.e. surrogate antigen). After 5 min, the cells were fixed and immunostained with anti-rat IgG to visualize the surrogate antigen and with anti-P-CD79a antibodies to visualize phosphorylation of the ITAMs in the CD79a/b subunit of the BCR. Representative images are shown. Scale bar: 5 µm. Dot plot graphs show the fluorescence intensity of antigen clusters (B) and P-CD79 (C) at the contact site between A20 B cells and the Cos-7 APCs. Fluorescence intensities were corrected for background fluorescence. Note that antigen monomers and small clusters likely do not exceed the background fluorescence values and that only larger antigen clusters (i.e. BCR microclusters) are quantified. (D) For each cell, the ratio of P-CD79 fluorescence intensity divided by the gathered antigen fluorescence intensity was calculated. Red lines indicate the median values. n > 51 cells from 4 independent experiments. ****P <0.001 by unpaired t-test.  210 5.2.5 HS1 is required for APC-induced cSMAC formation BCR microclusters that form when BCRs bind antigen move centripetally and coalesce into a cSMAC at the center of the cell:cell interface. HS1 is important for promoting the formation of peripheral actin structures at the T cell:APC contact site that drive T cell spreading. The retrograde movement of these actin structures at the cell periphery promotes centralization of TCR microclusters into a cSMAC [81, 371]. Therefore, I asked whether HS1 is also important for cSMAC formation when B cells contact APCs. To test this, A20 B cells expressing the HEL-specific transgenic D1.3 BCR were added to Cos-7 APCs expressing a transmembrane form of HEL-GFP. In control A20/D1.3 cells, antigen clusters formed within 60 s and began to accumulate into several larger clusters, which eventually coalesced into a central cluster, or cSMAC, by 10 min (see Figure 5.6B). When HS1 was depleted using either siRNAs (Figure 5.6A) or shRNA (Figure 5.4) targeting the mouse HCLS1 gene, the B cells formed small antigen clusters throughout the B cell-APC contact site but the rate of cluster formation was slower than in control cells (Figure 5.6B,C). At 1 min after adding the B cells to the APCs, a time at which the number of small antigen clusters that had formed in control cells was maximal, the HS1-depleted cells formed 6-fold fewer small antigen clusters (Figure 5.6B,C). Although the number of antigen clusters increased in the HS1-depleted cells over time (Figure 5.6D), the amount of antigen that had accumulated at the cell:cell interface was much lower than in control siRNA-transfected cells (Figure 5.6E). Therefore, HS1 is important for the gathering of antigen into BCR microclusters and for the coalescence of antigen-bound BCR microclusters into a cSMAC.   211  **** **** 212 Figure 5.6. HS1 depletion blocks the formation and coalescence of BCR microclusters  (A) Immunoblots of A20/D1.3 cells that had been transfected with control siRNA or HS1 siRNA. (B-E) A20/D1.3 cells that had been transfected with control of HS1 siRNA were stained with CellMask Deep Red (blue) and added to HEL-GFP-expressing Cos7 cells. Still images that were taken at the indicated times (in seconds) show antigen gathering by representative control and HS1-depleted cells (B,C). Time 0 indicates the time at which the B cell first contacted the APC. Scale bar: 10 µm. The average number of antigen clusters per cell (D) and the average total GFP (antigen) fluorescence that was gathered into clusters at the B cell:APC interface (E), are shown for >6 cells from two experiments. The gray curves represent control cells and the red curves represent HS1 siRNA-expressing cells. ****P <0.0001 using the Mann-Whitney U test.    Movie 14. Control B cells interacting with APCs form a cSMAC  See legend for Figure 5.6. The cells were imaged every 11 s for 10 min. The movie is played back at 9 frames per second (100X real time).   https://blogs.ubc.ca/jiawangthesis/2016/06/10/movie-1/ Password: JWthesis  Movie 15. HS1 depletion blocks cSMAC formation on APCs  See legend for Figure 5.6. The cells were imaged every 11 s for 10 min. The movie is played back at 9 frames per second (100X real time).   https://blogs.ubc.ca/jiawangthesis/2016/06/10/hs1-movie-2/ Password: JWthesis   213 5.2.6 HS1 is important for the assembly of the peripheral actin cytoskeleton during B cell spreading The formation of BCR microclusters and their coalescence into a cSMAC during IS development is regulated by the remodeling of the local actin cytoskeleton [9, 32, 55, 65]. At the T cell IS, HS1 regulates actin dynamics and is important for establishing actin structures that control TCR organization and T cell activation [142]. Hence, I asked whether HS1 is also important for generating peripheral F-actin structures at the B cell IS. To test this, A20 cells were allowed to spread on anti-IgG-coated coverslips for 15 min before imaging the F-actin. Compared to control cells, I found that HS1-deficient cells exhibited impaired spreading, with a cell area that was ~50% that of control cells (Figure 5.7A,B). The control cells exhibited a dense ring of peripheral F-actin, which resembled the lamellipodial actin structures of migrating cells [372], as well as a central actin-depleted region (Figure 5.7A). The HS1-deficient cells cleared F-actin from the center of the contact site with the anti-IgG-coated coverslip to some extent but exhibited a greater number of punctate actin structures throughout the cell whereas control cells had only sparse puncta in this region (Figure 5.7A). Notably, in the HS1-depleted cells, the ring of peripheral F-actin was thinner than in control cells (average of 0.7 µm thick compared to 1.5 µm thick in control cells), its total area was <50% of that in control B cells, (Figure 5.7A,C), and the total F-actin fluorescence intensity in this peripheral ring was less than that in the control cells (Figure 5.7D). Even when normalized to the cell area, the ratio of peripheral F-actin area to total cell area was significantly lower in HS1 knockdown cells than in control cells (Figure 5.7E). Therefore, HS1 is important for establishing the peripheral F-actin structures that promote the spreading of B cells on an antigen-bearing surface.  214  215 Figure 5.7. HS1 is important for B cell spreading and for the formation of peripheral F-actin structures  The A20 scrambled shRNA control cell population or the indicated A20 HS1 shRNA clones were allowed to spread on anti-IgG-coated coverslips for 15 min before being fixed and stained with Alexa 488-phalloidin to visualize F-actin structures. Representative confocal xy images are shown of control (scrambled sequence) or HS1 shRNA-expressing cells (A). Scale bar: 10 µm. The total cell area (B) and the area of the peripheral F-actin ring (C) were quantified for control cells and for 3 different HS1 shRNA KD clones that had been allowed to spread for 15 min. The total F-actin fluorescence intensity in the whole cell is shown in D. The graph in E shows the ratio of peripheral F-actin area to the total cell area. For each condition, n >72 cells from 3 independent experiments were analyzed. *P <0.05, ****P <0.0001 using unpaired t-tests.   216   Figure 5.8. HS1 promotes the assembly of the peripheral F-actin structures   A20 cells transduced with control scrambled sequence or HS1 shRNA 3 were transiently transfected with GFP-actin and allowed to spread on anti-IgG-coated coverslips for 15 min. Representative confocal images are shown. Scale bar: 5 µm. The white double-headed arrows show the width of the peripheral F-actin ring. For each cell, the width of the peripheral F-actin was measured at >10 sites selected randomly.  The dot plot shows the mean width of the peripheral F-actin structures for each cell, with n >26 cells per condition. **** P <0.0001 using an unpaired t-test.   217 5.2.7 HS1 promotes the formation of sustained membrane protrusions at the B cell periphery In migrating cells, branched actin polymerization mediated by the Arp2/3 complex exerts outward force on the membrane to create membrane protrusions. By regulating Arp2/3 activity, the HS1-related protein cortactin promotes branched actin assembly at the leading edge of migrating fibroblastic sarcoma cells and is required for Arp2/3-dependent lamellipodial protrusions to persist [373, 374]. Because HS1 KD B cells exhibited decreased radial spreading on anti-IgG-coated coverslips, I first asked whether this was due to a decreased rate of F-actin assembly at the cell periphery. To quantify local rates of F-actin dynamics I used fluorescence recovery after photobleaching (FRAP). Control and HS1-deficient A20 cells that had been transfected with actin-GFP were allowed to spread on anti-IgG-coated coverslips for 15 min before photobleaching peripheral ROIs. The recovery of actin fluorescence was used as a measure of actin dynamics. However, I found that there was no significant difference in the fluorescence recovery post photobleaching between control and HS1 shRNA-expressing B cells (Figure 5.9, Movie 16). Therefore, the reduced cell spreading in HS1-deficient cells is due to factors other than a defect in actin dynamics.  218  Figure 5.9. HS1 knockdown does not affect actin dynamics   Actin-GFP FRAP kinetics for peripheral ROIs in control and HS1-depleted A20 cells. (A) Populations of A20 cells that had been stably transduced with control shRNA or HS1 shRNA were transiently transfected with actin-GFP. (B) A20 cells were transiently co-transfected with actin-GFP plus either control siRNA or HS1 siRNA. The cells were allowed to spread on anti-IgG-coated coverslips for 15 min. Circular 3.5-µm diameter ROIs at the periphery of the cell were photobleached and fluorescence recovery was imaged in real time. Confocal xy images were captured every 5 s (A) or every 2 s (B) post-bleaching. Fluorescence recovery of the actin-GFP within the ROI was quantified as described in the Methods. For each cell analyzed, a single ROI was imaged. For each time point, the mean ± SEM for ROIs from 19 cells in two independent experiments is shown.  219 Movie 16. FRAP for actin-GFP in A20 cells stably transduced with control shRNA or HS1 shRNA  The cells were transiently transfected with actin-GFP and allowed to spread on anti-IgG-coated coverslips for 15 min. A 3.5-μm diameter peripheral ROI was photobleached and the recovery of actin-GFP fluorescence was imaged for 85 s. The movie is played back at 4 frames per second (20X real time).   https://blogs.ubc.ca/jiawangthesis/2017/01/03/hs1-actingfp-frap/ Password: JWthesis    Because the rate of actin assembly was not significantly altered by HS1 depletion, I tested whether HS1 was important for the persistence of peripheral actin structures during cell spreading. A20 cells expressing F-tractin-Dendra were allowed to adhere to anti-IgG-coated coverslips for 3 min before they were imaged using TIRF microscopy in order to capture the cell spreading events at the antigen contact site. Consistent with previous observations that cortactin is required for the persistence of F-actin structures at the lamellipodia, HS1 deficiency resulted in transient peripheral protrusions that were short-lived compared to those in control cells (Figure 5.10, Movie 17, Movie 18). The membrane protrusions in control cells had an average lifetime of ~30 s before they retracted whereas the membrane protrusions in HS1-deficient cells were more transient with a lifetime of ~15 s (Figure 5.10). Thus, HS1 is important for forming stable actin-based protrusions that promote B cell spreading.   220  Figure 5.10. Peripheral actin protrusions are less stable in HS1-depleted B cells  Scrambled control (A) and HS1 shRNA-expressing (B) A20 cell populations were transiently transfected with F-tractin-Dendra and then allowed to spread on anti-IgG-coated coverslips for 3 min. The cells were then imaged by TIRF microscopy at a 90 nm TIRF depth for an additional 3 min, with a capture rate of one frame each 0.5 s. The blue bars on the cells in the upper panels indicate the areas used to create the representative kymographs (lower panels), which show peripheral actin dynamics over the 3-min imaging period. The dotted lines represent the edge of the cell. Lower right-hand panels of A and B show grayscale kymographs and the lower left-hand panels show cyan hot-pseudocolored and re-sliced kymograph images.    221 Movie 17. Actin dynamics in spreading control cells   Scrambled shRNA control A20 cells were transiently transfected with F-tractin-Dendra and then allowed to spread on anti-IgG-coated coverslips for 3 min. The cells were then imaged by TIRFM at a 90 nm TIRF depth for an additional 3 min, with a capture rate of one frame each 30 ms. The video is played back at 20 frames per second (12.5X real time) The yellow line indicates the area of cell that was used to generate the kymograph in Figure 5.10.  https://blogs.ubc.ca/jiawangthesis/2016/09/20/controlftractindendra-tirf/ Password: JWthesis   Movie 18. Actin dynamics in spreading HS1-depleted cells   HS1 shRNA-expressing A20 cells were transiently transfected with F-tractin-Dendra and then allowed to spread on anti-IgG-coated coverslips for 3 min. The cells were then imaged by TIRF microscopy at a 90 nm TIRF depth for an additional 3 min, with a capture rate of one frame each 30 ms. The video is played back at 20 frames per second (12.5X real time). The yellow line indicates the area of cell that was used to generate the kymograph in Figure 10.   https://blogs.ubc.ca/jiawangthesis/2016/09/20/hs1-tirfcellspreading/ Password: JWthesis  5.2.8 HS1 controls the actin organization around BCR microclusters Because actin reorganization is important for driving the coalescence and centralization of BCR microclusters, I next tested the hypothesis that HS1 is involved in the generation of actin structures that form immediately after B cells contact APC-bound antigens. To simultaneously image antigen gathering by BCR microclusters and the formation of F-actin structures, scrambled control and HS1 shRNA-expressing A20/D1.3 cells were transiently transfected with tdTomato-F-tractin and then added to Cos-7 APCs expressing HEL-GFP on their surface. Upon contacting the APC, the control A20 cells rapidly formed distinct antigen-BCR microclusters that were initially surrounded by F-actin clusters, which then assembled into a ring structure around the antigen clusters. These F-actin ring structures then reorganized around additional 222 antigen-BCR clusters that formed at the periphery, which merged together with the initial BCR microclusters (Figure 5.11). When HS1 was depleted, the F-actin structures that formed at the antigen contact site were filamentous and elongated (Figure 5.11B). Unlike in control cells, these actin structures did not assemble around the BCR-antigen clusters that formed at the B cell:APC interface. Rather, these actin structures were interspersed among the BCR-antigen puncta, which remained small compared to those in control cells at the same time points. Some BCR-antigen clusters in the HS1-depleted cells did move towards the center of the cell, apparently along some of the elongated actin fibers. Although some of these small BCR-antigen clusters fused into larger clusters, they did not coalesce into a cSMAC, similar to what was observed for HS1-deficient cells in Figure 5.6. Therefore, actin-dense structures form around microclusters and reorganize around coalescing antigen clusters, possibly to maintain the structure and integrity of the microclusters [53].   223   224 Figure 5.11. Actin-dense clusters form around BCR-antigen clusters at the interface between a B cell and an APC   Scrambled control (A) and HS1 shRNA 3 (clone 15)-expressing (B) A20/D1.3 cells expressing tdTomato-F-tractin were added to the Cos-7 cells expressing mHEL-GFP and imaged in real time. Representative xy images that were captured at the indicated times after the B cells were added to the APCs are shown. The scale bars in the upper panels of (A) and (B) are 5 μm. Scale bars in insets are 1 μm.   Movie 19. Actin dynamics during BCR-antigen cluster formation in control cells  Vector control-expressing A20/D1.3 cells were transiently transfected with tdTomato-Ftractin before adding to the top of mHEL-GFP-expressing Cos-7 APCs that had spread on FN-coated coverslips. Images were captured every 12 s over 12 min for a total of 60 frames. The movie is played back at 10 frames per second (120X real time).   https://blogs.ubc.ca/jiawangthesis/2017/07/25/hs1-f-actin-dynamics-during-antigen-cluster-formation/ Password: JWthesis  Movie 20. Actin dynamics during BCR-antigen cluster formation in HS1 KD cells.   Vector control-expressing A20/D1.3 cells were transiently transfected with tdTomato-Ftractin before adding to the top of mHEL-GFP-expressing Cos-7 APCs that had spread on FN-coated coverslips. Images were captured every 12 s over 12 min for a total of 60 frames. The movie is played back at 10 frames per second (120X real time).   https://blogs.ubc.ca/jiawangthesis/2017/07/25/hs1-f-actin-dynamics-during-antigen-cluster-formation/ Password: JWthesis     225 5.2.9 HS1 is important for MTOC reorientation towards the B cell IS Because HS1 is important for actin remodeling and IS formation in B cells, I tested the hypothesis that HS1 is also important for the polarization of the MTOC towards the IS, a process that depends on actin remodeling (see Chapter 3). As has recently been shown [136], shRNA-mediated KD of HS1 in A20 cells prevented MTOC reorientation towards anti-IgG-coated beads. Because HS1 works in concert with the Arp2/3 complex, I asked whether Arp2/3 is also important for BCR-induced MTOC reorientation. Consistent with the finding that HS1 is required for BCR-induced MTOC reorientation, shRNA-mediated knockdown of HS1 in A20 cells prevented MTOC polarization towards anti-Ig-coated beads (Figure 5.12). Blocking Arp2/3-initiated actin polymerization using the Arp2/3 inhibitor CK-666 [294] also prevented MTOC reorientation towards the site of antigen contact (Figure 5.13). These findings are consistent with the idea that HS1 promotes cSMAC formation by promoting Arp2/3-dependent actin remodeling, which is essential for reorienting the microtubule network towards the site of antigen contact.   226  Figure 5.12. HS1 is required for BCR-induced MTOC reorientation towards the B cell IS   Scrambled control and HS1 shRNA-expressing A20 B cells were mixed for 30 min with 4.5-µm diameter beads that were coated with anti-IgG antibodies. The cells were then fixed and immunostained with α-tubulin antibodies and rhodamine-phalloidin to visualize F-actin. (A) Representative images of B cell:bead conjugates are shown. Scale bar: 5 µm. (B) The dot plot shows the MTOC polarity indices for > 52 cells from 2 independent experiments. The red bars indicate the mean values. ** P <0.01, unpaired t-test.  227  Figure 5.13. The Arp2/3 inhibitor CK-666 blocks BCR-induced MTOC polarization   Ex vivo B cells were treated with the control CK-689 compound or 100 μM of the Arp2/3 inhibitor CK-666 before being mixed for 30 min with 4.5-μm diameter anti-IgM-coated beads. (A) Cells were fixed and stained with rhodamine-phalloidin (red), α-tubulin antibodies (green), and DAPI (blue). Representative confocal xy slices are shown. (B) MTOC PIs for >43 cells from two experiments are shown. The red bars indicate the mean values. ****P <0.0001 using an unpaired t-test. Scale bar: 5 μm.   228 5.3 Discussion 5.3.1 Summary of findings Actin remodeling at the site of antigen contact during B cell activation is crucial for forming an IS that optimizes BCR-mediated signaling and promotes APC-induced B cell activation. The molecular events leading to BCR microcluster formation and gathering into an IS are coordinated by cytoskeletal remodeling. A critical event driving B cell activation is the formation of BCR microsignalosomes. In vivo, B cells encounter cognate antigen in limited quantities on the surfaces of APCs. Therefore, forming as many BCR microclusters as possible helps B cells achieve sufficient BCR signaling for activation [9]. This is achieved by antigen-induced spreading of the B cell membrane across the surface of the APC, which increases the number of APC-bound antigens that the B cell encounters, leading to the formation of additional signaling microclusters [64]. Multiple BCR signaling pathways that target actin-remodeling proteins promote the reorganization of the actin cytoskeleton that drives B cell spreading and IS formation [9]. In this chapter I describe a role for HS1 in these processes. I found that HS1 is important for the development of peripheral actin structures that drive B cell spreading, promote BCR microcluster formation, and support the centralization of BCR microclusters into a cSMAC.   5.3.2 HS1 phosphorylation HS1 is rapidly phosphorylated upon antigen receptor engagement. This converts HS1 into an active form that can interact with nucleation factors to regulate actin polymerization and organization. I showed that HS1 is phosphorylated within minutes of stimulating B cells with soluble anti-Ig antibodies and this was, in part, a Rap-dependent process that was impaired in 229 cells expressing RapGAPII. This may be explained by the critical role of Rap in cofilin-mediated actin severing [55], which allows increased BCR mobility and BCR-BCR collisions, and hence increased BCR signaling. We showed previously that blocking Rap activation did not inhibit the later BCR-induced phosphorylation of ERK or Akt, which are important for B cell activation, in A20 cells [247]. Because Rap activation controls actin dynamics and organization in B cells, the phosphorylation or activation of actin-associated proteins such as HS1 may exhibit a stronger dependence on Rap than other targets of BCR signaling.  I found that the phosphorylated form of HS1 was localized with F-actin at the antigen contact site. In T cells, it is the phosphorylated form of HS1 that promotes actin polymerization at the IS [142, 319]. Rap-dependent actin remodeling is critical for BCR microcluster formation and signaling when B cells encounter polarized arrays of antigens (e.g. on APCs) [55]. Because Rap is important for HS1 phosphorylation, further studies should assess BCR-induced HS1 phosphorylation in the context of B cell:APC interactions. Therefore, I hypothesize that HS1 phosphorylation will exhibit a much stronger dependence on Rap activation when B cells encounter polarized arrays of antigen, such as on APCs, where Rap-dependent actin remodeling is critical for BCR microcluster formation. If so, it would support the idea that there is a positive feedback loop in which initial antigen binding activates Rap and stimulates HS1 phosphorylation. Subsequent Rap- and HS1-dependent actin remodeling, B cell spreading, and the formation of BCR microclusters that amplify BCR signaling could lead to the increased and sustained phosphorylation of HS1.   230 5.3.3 HS1 controls actin dynamics in B cells My observation that HS1 phosphorylation and activation is regulated by Rap suggests that HS1 could be an additional downstream target by which Rap regulates actin dynamics and organization upon BCR signaling. Similar to its role in T cell spreading [142], I found that HS1 is also important for B cells to spread maximally on immobilized anti-Ig. HS1-deficient cells formed a thinner ring of peripheral actin than control cells and never achieved the same cell spreading area as control cells. T cells spreading on lipid bilayers form a peripheral ring of F-actin that is 2 µm in width [371]. Consistent with this, I found that untransfected A20 B cells formed a peripheral actin ring that was ~1.5 µm thick. In contrast, the peripheral actin ring that formed in HS1-deficient cells was, on average, only half as thick. This may reflect a role for HS1 in the formation of the peripheral branched actin network, which pushes the membrane outward and drives cell spreading. Possibly, this may be due to the role of HS1 in stabilizing peripheral actin structures [142, 319, 375]. The less stable peripheral actin structures in HS1-deficient cells may not generate sufficient force to propel the membrane outward to the same extent.  The peripheral actin structures and their movement drive both B cell spreading and IS formation. Therefore, it was pertinent to examine the subcellular localization of HS1 and the active, phosphorylated form of HS1. When B cells spread on immobilized anti-Ig, HS1 co-localized with the peripheral F-actin ring and HS1-YFP exhibited retrograde movement from the cell periphery towards the center of the cell, much like the retrograde movement of the actin cytoskeleton. This suggests an interaction between HS1 and the peripheral actin structures. This idea is supported by my finding that the centripetal movement of HS1 was disrupted by adding latrunculin A to depolymerize F-actin. Moreover, this suggests that the movement of HS1 depends on actin retrograde flow. Importantly, immunostaining for P-HS1 showed that the 231 majority of active HS1 at the antigen contact site was localized at the periphery of the cell, and was closely associated with the dense ring of F-actin. Because P-HS1 enhances the activity of the Arp2/3 complex, immunostaining combined with super-resolution microscopy, as well as proximity ligation assays, should be carried out to determine whether P-HS1 and the Arp2/3 complex are in close proximity to each other at the peripheral ring of F-actin. I also found that some P-HS1 was also located in the central region of the cell, which is devoid of dense actin structures. Higher resolution imaging could determine whether P-HS1 is associated with actin arcs that form near the interface between the peripheral actin ring and the central actin-poor region. These actin arcs are curved bundles of actin filaments that are important for the centralization of TCR microclusters during cSMAC formation [80]. In non-lymphoid cells, the HS1 homologue cortactin is important for forming actin structures at the lamellipodium where it is thought to remodel a subset of dendritic actin filaments into transverse bundles of actin arcs [376]. Based on my finding that HS1 is important for BCR microclusters to coalesce, it would be interesting to determine if HS1-deficient B cells failed to form these actin arcs.   Consistent with its localization at the of the peripheral actin ring, early spreading events in HS1-depleted cells were altered such that irregular and transient actin-based protrusions were formed at the cell periphery. The inability to form stable peripheral actin structures may prevent the formation of the thick, radial lamellipodial ring of F-actin as described above. This is similar to what has been observed in T cells, where HS1-deficient cells form unstable lamellipodia during spreading on anti-CD3-coated coverslips [142]. The lamellipodium is comprised of a dendritic network of branched actin filaments that are formed via WASP- and WAVE-dependent activation of the Arp2/3 complex, [319]. In T cells, HS1 supports Arp2/3-mediated actin polymerization and stabilizes branched actin structures [142, 319]. Hence, the inability of HS1-232 deficient B cells to form stable, sustained actin-based protrusions at the cell periphery may reflect a critical role for HS1 in stabilizing Arp2/3-generated actin structures. The defects in BCR microcluster formation and centralization in HS1-deficient B cells may reflect the inability of these cells to stabilize both the peripheral branched actin network and actin arcs, whose retrograde motion drive the centripetal movement of BCR microclusters (Figure 5.14). High-resolution live imaging tools such as TIRF structured illumination microscope or Airy disk confocal microscopy [377] could reveal whether alterations in the organization or movement of these actin structures in HS1-deficient cells corresponds with the defect in BCR microcluster movement. 233  Figure 5.14. Model for the role of HS1 in IS formation.   BCRs that bind cognate antigen on the surfaces of APCs initiate BCR signaling. BCR-activated SFKs such as Lyn then phosphorylate HS1, which induces a conformational change that allows HS1 to interact with the Arp2/3 complex. The Arp2/3 complex then nucleates the assembly of dendritic actin at the periphery of the antigen contact site. This drives radial membrane spreading across the surface of the APC. Antigen clusters that form at the cell periphery are initially moved inwards towards the center of the antigen contact site via the retrograde flow of the peripheral dendritic actin. Additionally, actin arcs, which form after the B cell spreading has initiated, may sweep BCR microclusters inwards so that they can coalesce into larger clusters. Actin clusters form around these larger BCR microclusters in an HS1-dependent manner. This may prevent 234 individual BCRs from dissociating from the microclusters or promote the further coalescence of microclusters into larger clusters. At the same time, Rap and HS1-dependent cell spreading facilitates the reorientation of the MTOC and the microtubule network (orange) towards the site of antigen contact in order to further drive the centralization of BCR microclusters via dynein-dependent mechanisms.   5.3.4 HS1 regulates IS formation in B cells In T cells, HS1 is important for the organization and stability of actin structures that form at the IS [142]. The loss of HS1 in T cells results in impaired calcium mobilization and IL-2 secretion, which require IS formation and are critical steps during T cell activation [142]. Because actin remodeling also controls IS formation in B cells, I examined whether HS1 controls actin dynamics and BCR microcluster formation, movement, and organization when antigen is presented on APCs. I showed that depleting HS1 decreased the number of BCR-bound antigen microclusters that formed when B cells contacted antigen-bearing APCs. This could be due, at least in part, to the fact that the peripheral actin structures that formed at the antigen contact site in HS1-deficient B cells were not sustained, resulting in transient membrane protrusions and impaired cell spreading, as has been seen in T cells [142].  I found that depleting HS1 in B cells prevented the coalescence of BCR microclusters into a cSMAC. The coalescence of antigen-bound receptors into a cSMAC is driven by actin-based processes (Figure 5.14) [55]. In both B and T cells, many antigen receptor microclusters form at the periphery of the cell and then move centripetally to form a cSMAC in the central actin-depleted region of the cell [371]. In T cells, the rate of TCR microcluster centripetal movement is similar to the rate of actin retrograde flow, suggesting that actin retrograde flow drives the initial movement of the microclusters towards the center of the IS [81]. This retrograde actin flow has been linked to the Arp2/3-dependent formation of branched actin structures at the periphery of the cell [80, 226]. Additionally, high resolution real-time TIRF-SIM imaging 235 showed that actin arcs that lie within the ring of dendritic actin sweep the TCR microclusters inwards towards the actin-poor central region of the IS and that this leads to formation of the cSMAC [80]. Therefore, the radial lamellipodium-like organization of the actin cytoskeleton is important for TCR, and likely BCR, microcluster centralization. The thinner peripheral actin ring in HS1-deficient B cells, compared to control cells, could contribute to the impaired centralization of BCR microclusters. During lymphocyte spreading, both lamellipodial and filopodial protrusions can form. As with Arp2/3 inhibition [378], HS1 depletion did not inhibit the formation of filopodial structures at the periphery of the APC contact site, which acted as tracks for moving some peripheral antigen clusters from the edges of the cell towards the center of the cell (see Figure 5.11). Because HS1 and Arp2/3 often act cooperatively to assemble and maintain actin networks, the defects in antigen microcluster formation and centralization in HS1-depleted cells may result from impaired Arp2/3 functions. Although the role of HS1 in enhancing Arp2/3 function has been demonstrated biochemically, the effects of HS1 on Arp2/3 function in intact cells have not been examined. If HS1 acts primarily via the Arp2/3 complex, then treating B cells with the Arp2/3 inhibitor CK-666 should reproduce the defects in actin organization and BCR centralization that I observed in HS1-depleted cells. Moreover, because both HS1 and Arp2/3 are important for BCR-induced MTOC reorientation, the defect in cSMAC formation could also be due to impaired microtubule-dependent BCR microcluster centralization, which relies on the reorientation of the MTOC towards the site of antigen contact. I also found that actin-dense structures surrounded BCR microclusters in control cells but that these were either absent or more diffuse in HS1-depleted cells. It is not known whether Arp2/3 is required for forming these actin structures that encircle BCR microclusters or whether other HS1-dependent mechanisms 236 are involved. Although these actin structures may maintain the integrity of BCR microclusters [53], their role in cSMAC formation is not known. Using live imaging tools to simultaneously image the reorganization of the actin and the microtubule networks along with BCR microcluster movement in control, HS1-depleted, and Arp2/3 inhibitor-treated B cells would provide further insights into how HS1 and Arp2/3 coordinate these processes.   5.3.5 HS1 regulates BCR signaling B cell spreading over antigen-bearing surfaces allows B cells to maximize the contact area with APCs in order to increase the binding of cognate antigen [64, 65]. Defects in cell spreading could lead to decreased BCR signaling and hence diminished B cell activation and impaired antibody responses [65]. Actin-driven B cell spreading may be especially critical for B cell activation during immune responses to limited amounts of antigen. Indeed, I found that HS1 is important for the antigen-induced formation of BCR microclusters and, hence, total BCR signaling at the IS. In HS1-deficient B cells, the decrease in the number of microclusters was directly correlated with the decrease in CD79a phosphorylation. This suggests that BCR signaling at individual microclusters was not affected by the loss of HS1 but that fewer microclusters developed. This could reflect the decreased formation of BCR microclusters or the inability to maintain coherent microclusters, which depend on the actin cytoskeleton [53]. The defective encirclement of BCR microclusters by F-actin could also result in the dissociation of BCRs from the microcluster.   In T cells, HS1 regulates the spatial dynamics of PLCγ at the IS. HS1 is recruited to the IS by ITK where it interacts directly with phosphorylated PLCγ and stabilizes the recruitment of PLCγ to the IS by linking PLCγ to the actin cytoskeleton [379]. Consistent with the role of ITK and PLCγ in TCR-induced Ca2+ mobilization, loss of HS1 leads to defects in Ca2+ release from 237 intracellular stores [142, 379]. Whether HS1 and PLCγ interact to promote Ca2+ signaling in B cells, particularly when B cells are stimulated with polarized arrays of antigen, remains to be elucidated.   5.3.6 Perspectives The control of B cell activation is critical for regulating normal immune responses to pathogens and for reducing aberrant B cell activation to avoid incidences of autoimmunity and B cell malignancies. I’ve shown that HS1 is important for IS formation in B cells, which is critical for supporting the BCR signaling that leads to B cell activation. Aberrant BCR signaling is highly correlated with the development and survival of autoimmune and cancerous B cells [380], which may be the result of actin cytoskeleton dysregulation. In Chronic Lymphocytic Leukemia (CLL), the dysregulation and hyperphosphorylation of HS1, which interacts with the actin cytoskeleton in these cells, is highly associated with poor clinical outcome [381]. Treating CLL cells with the tyrosine kinase inhibitor Dasatinib abrogates Lyn-mediated HS1 phosphorylation. Dasatinib-mediated inhibition of HS1 reduces cytoskeletal dynamics as well as decreases BCR signaling and cell survival [382]. The use of Dasatinib in mouse models further demonstrated that inhibiting the Lyn/HS1 axis could delay CLL progression [382]. Therefore, HS1-mediated cytoskeletal dynamics can be an important clinical target in B cell-mediated diseases. Although these initial studies were conducted with CLL cells, it is possible that targeting HS1 in other B cell malignancies and autoimmune diseases could reduce the BCR signaling that drives cell survival and disease progression. Thus, targeting HS1 using tyrosine kinase inhibitors such as Dasatinib may be a therapeutic approach for treating B cell-mediated diseases.   238 Chapter 6: The Rap1-cofilin pathway and the B cell cytoskeleton are important for antigen acquisition from APCs 6.1 Introduction   In vivo, the differentiation of B-lymphocytes into antibody-producing cells is often initiated by antigen-presenting cells (APCs) such as follicular dendritic cells (FDCs) and subcapsular sinus (SCS) macrophages. These APCs capture antigens via a variety of receptors including Fc receptors, complement receptors and scavenging receptors and then display them on their surface in an intact form that can be recognized by the BCR [8, 28, 38]. For such membrane-associated antigens, BCR-induced reorganization of the actin and microtubule cytoskeletons, and the resulting formation of an IS, is critical not only for optimizing BCR signaling but also for promoting antigen internalization [29, 65, 300]. The IS is particularly important for naïve B cells, which acquire antigens from APCs only after the IS has formed [66, 87]. BCR-bound antigens are internalized and delivered to antigen processing compartments via mechanisms that involve both the actin and microtubule networks [29, 300]. The resulting peptide-MHC II complexes are presented to T helper cells, which then provide additional signals for B cell activation [29, 95]. Although complex changes in cytoskeletal organization play a key role in IS formation and function, the mechanisms that coordinate these processes are not fully understood.  B cells use their BCRs to extract antigens that are bound by receptors on the surface of APCs, and then using endocytic mechanisms to internalize the extracted antigen. BCR-mediated antigen extraction involves actomyosin-dependent mechanisms that provide sufficient force to overcome the binding affinity of the antigen for the capture receptor on the APC [37, 107, 115]. Reorientation of the MTOC also contributes to antigen extraction by recruiting lysosomes, the 239 contents of which can be secreted at the IS [85]. The actions of lysosomal proteases can release APC-bound into the synaptic space so that they can be internalized by the BCR [85]. This may be particularly important for the extraction of high affinity antigens from stiffer APC membranes or substrates, where actin-based mechanical forces alone are insufficient [109]. The internalization of extracted antigens often requires clathrin-mediated endocytosis (CME), which depends on Arp2/3-mediated actin polymerization to provide force to invaginate the membrane around BCR-antigen complexes [37, 107, 115]. Upon internalization, these BCR-antigen complexes are trafficked to lysosomes and MIICs that are recruited to the site of MTOC reorientation. Peptides derived from the antigen are then loaded onto MHC II proteins and trafficked to the cell surface for presentation to T cells. Therefore, both actin and microtubule-based processes are required for the extraction, internalization, and processing of BCR-bound antigens, which are essential for B cells to gain T cell help.  Although initial antigen-mediated activation of naïve B cells can induce their differentiation into plasmablasts that secrete low affinity IgM antibodies, for T-dependent antigens B cells require T cell help from primed T cells as well as follicular T helper cells in order to undergo a GC reaction and produce high affinity, class-switched antibodies [15, 29]. Therefore, the APC function of B cells is critical for mounting a robust humoral response. T cells provide help to B cells in the form of cytokines (e.g. BAFF, IL-4, IL-21, IL-17) as well as co-stimulatory molecules (e.g. CD40L, ICOS, PD-1) that promote B cell survival, somatic hypermutation, Ig class switching, and differentiation into long-lived plasma cells and memory B cells [15, 29]. The amount of antigen that a B cell internalizes determines its antigen presentation capacity and ability to elicit T cell help, a limiting step in B cell activation [64, 327]. FDCs within GCs present limited amounts of antigen that GC B cells compete for in order to acquire 240 and present antigen to TFH cells. Hence optimizing the antigen acquisition process, especially when the amount of available antigen is low, may be essential for B cell activation, and for generating a robust affinity-matured antibody response.  The Rap1-cofilin pathway controls actin cytoskeleton reorganization as well as MTOC repositioning to the site of antigen contact, both of which are important for IS formation (see Chapter 3) [55]. In order to extract and internalize antigen, naive B cells require IS formation, actin-dependent forces, and secretory lysosomes that are recruited to the site of antigen contact via the reorientation of the MTOC. Given the critical role of the Rap1-cofilin pathway in reorganizing the cytoskeleton at the IS, I tested the hypothesis that Rap1 activation is important for B cells to acquire antigens from APCs. I report here that Rap1 and cofilin are important for the acquisition of antigen from APC membranes. In this chapter I show that Rap1 expression and activation are required for B cells to acquire and internalize antigens from APCs. This reveals an important new role for Rap in mediating a key function of the B cell IS.  6.2 Results 6.2.1 Antigen acquisition from APCs requires MTs and actin To study how B cells acquire antigens from APCs, we used Cos-7 cells expressing a transmembrane form anti-Igκ light chain antibody that binds to the BCR [297]. This ‘surrogate antigen’ is mobile within the APC plasma membrane, stimulates BCR signaling, and induces the formation of BCR microclusters that coalesce into a cSMAC [55]. I found that primary B cells extracted this surrogate antigen from the APC membrane, which could be detected within the B cell as puncta that co-localized with the BCR and the early endosome marker EEA-1 (Figure 6.1A). Of the B cells that were in contact with APCs, ~45% had extracted and internalized antigen within 5 min and ~80% had done so by 20 min (see Figure 6.5C control siRNA 241 samples). BCR-mediated antigen acquisition was also visualized by real-time imaging using Cos-7 cells expressing a transmembrane form of HEL fused to GFP [66]. As early as 5 min, A20 B-lymphoma cells expressing the HEL-specific D1.3 BCR [30] had acquired HEL-GFP from the APC (Movie 21 and Figure 6.1B). The GFP signal was present at the B cell:APC interface and in puncta that flowed upwards into the B cell or along the B cell surface (Movie 21). Thus, B cells rapidly extract antigens from APCs and deliver them to endocytic compartments.    242   Figure 6.1. Antigen internalized from APCs is trafficked to early endosomes  (A) LPS-activated primary B cells were mixed in suspension with Cos-7 APCs expressing the anti-Igκ surrogate antigen for 20 min. B cell:APC conjugates were stained for antigen, the BCR, and the early endosome marker EEA-1. Internalized antigen that co-localized with the BCR and EEA-1 is indicated by the white merge signal. (B) A20/D1.3 cells expressing a HEL-specific BCR were mixed with HEL-GFP-expressing Cos-7 APCs for 30 min. Arrows indicate HEL-GFP internalized by A20 cells. Scale bars: 5 μm.        243 Movie 21. B cells rapidly acquire antigen from APCs  Real-time 4D imaging of A20/D1.3 B cells (red) acquiring antigen from adherent Cos-7 APCs expressing HEL-GFP (green). A20 cells were stained with CellMask Deep Red and allowed to settle onto the APCs for 5 min. Time-lapse spinning disk confocal microscopy was then used to acquire images every 30 s for 15 min (i.e. between 5 min and 20 min after initiating the B cell:APC interaction). The video is played back at 4 frames per second (120X real time).   https://blogs.ubc.ca/jiawangthesis/2016/05/16/chapter-3-movie-1/ Password: JWthesis   The actin and microtubule cytoskeletons have a critical role in BCR-mediated antigen extraction and antigen internalization. Myosin IIA exerts actin-dependent mechanical forces required for antigen extraction whereas Arp2/3-dependent actin polymerization exerts forces on the plasma membrane, which is required for forming an endosome that encloses BCR-antigen complexes [37, 106, 107, 112, 115]. BCR-induced actin reorganization is also required for the reorientation of the MTOC towards the IS. This is important for the coalescence of BCR microclusters into a cSMAC [9], which is required for antigen extraction by naïve B cells [87], and for recruiting lysosomes to the IS in order to facilitate proteolytic cleavage of APC-bound antigens [29]. Although Tolar and colleagues have investigated the role of actin and cSMAC formation in BCR-mediated antigen extraction from APC membranes, the role of MTOC reorientation in this process has not been assessed. Therefore, based on the findings in Chapters 3 and 4, I tested the hypothesis that BCR-induced MTOC polarization is required for B cells to acquire antigens from APCs. Moreover, because BCR-induced MTOC reorientation towards the IS requires actin reorganization and proteins that link the actin and microtubule cytoskeletons, I tested whether this was also required for antigen extraction. 244 The BCR-mediated acquisition of bead-bound antigens requires PKCζ-dependent reorientation of the MTOC towards the antigen contact site [85]. I found that treating primary B cells with a cell-permeable inhibitor of PKCζ, which blocked MTOC reorientation (Figure 3.4), significantly reduced the ability of B cells to extract and internalize APC-bound antigens (Figure 6.2; see Figure 2.2 for quantification method). Similarly, nocodazole-induced microtubule disassembly, which prevented antigen-bound BCRs from coalescing into a cSMAC (see Chapter 3), abrogated antigen uptake from APCs (Figure 6.2). In Chapter 4, I showed that BCR-induced reorganization of the actin cytoskeleton is required for MTOC polarization towards the antigen contact site and that the IQGAP1 scaffolding protein and the CLIP-170 plus-end binding protein are required for this process. Consistent with these results, I found that depolymerizing F-actin with latrunculin A greatly reduced antigen uptake even though some antigen clusters accumulated at the B cell:APC contact site (Figure 6.2). Depleting the microtubule plus-end binding protein CLIP-170 also reduced antigen internalization (Figure 6.3). Thus, both the microtubule and actin cytoskeletons play a role in the ability of B cells to extract and internalize antigens from APCs.  245    Figure 6.2. Internalization of antigens extracted from APCs requires the microtubule and actin networks as well as PKCζ activity   LPS-activated primary B cells were pre-treated with DMSO (control), with 20 μM PKCζ pseudosubstrate inhibitor (PKCζ-PS) for 1 hr, or with 5 μM nocodazole (Noc) or 2 μM latrunculin A (Lat A) for 5 min. The B cells were then mixed with anti-Igκ-expressing Cos-7 APCs for 20 min while maintaining the same concentration of DMSO or inhibitors. B cell:APC conjugates were stained for antigen, F-actin, and the BCR. The arrow in (A) indicates antigen internalized by the B cell. The amount of internalized antigen was quantified (B) as described in Figure 2.2 where only the antigen within the ring of BCR fluorescence was considered (n > 35 cells from 3 independent experiments). Red lines indicate means. ****P <0.0001, two-tailed unpaired t-test. Representative xy confocal slices are shown. Scale bar: 10 μm.    246   Figure 6.3. CLIP-170 is required for antigen acquisition from APCs  Control A20 cells transduced with the empty pGipz vector, as well as A20 cells expressing CLIP-170 shRNA were mixed with anti-Igκ-expressing APCs for 30 min and then stained for the antigen and the BCR. Representative xy confocal slices are shown. The arrows indicate antigen (Ag) and BCR that have been internalized. The graph shows the amount of internalized antigen, normalized to the antigen density on the APC for n >57 cells from three experiments. ****P <0.0001, two-tailed unpaired t-test. Scale bar: 5 μm.   247 6.2.2 Lysosome accumulation near the antigen contact site depends on Rap1  The polarization of lysosomes towards the antigen contact site is an important mechanism for extracting antigens from beads and stiff membranes [85, 109]. Lysosomes are often associated with the microtubule network [383]. Consistent with previous reports [85], I found that lysosomes were clustered around the reoriented MTOC at the site of antigen contact, placing these processing compartments in close proximity to the site of antigen acquisition (Figure 6.4). In Chapter 3 I showed that the Rap1 GTPase is required for BCR-induced polarization of the MTOC towards the antigen contact site. Consistent with this finding, I found that depleting Rap1 also prevented the repositioning of lysosomes to the bead contact site (Figure 6.4).   Figure 6.4. Lysosome accumulation near the antigen contact site depends on Rap1  LPS-activated splenic B cells that had been transduced with control or Rap1 siRNAs were mixed with fluorescent anti-Ig-coated amino beads for 30 min and then immunostained for the lysosomal marker LAMP1 and for α-tubulin. Representative 3D reconstruction images are shown. Scale bar: 5 μm.   248 6.2.3 B cells acquire antigens from APCs in a Rap1-dependent manner Because Rap1 regulates cell polarity, actin organization, IS formation, and the polarization of lysosomes towards the IS, we asked if Rap1 is important for B cells to acquire antigens from APCs. To do this, we used siRNA to knock down the expression of both Rap1 isoforms, Rap1a and Rap1b in primary B cells (Figure 6.5A). Rap1a/b knockdown (KD) cells could form conjugates with APCs but acquired less antigen than control B cells (Figure 6.5B,C). After being mixed with APCs for 5 min, 45% of the control B cells that were in contact with APCs had internalized APC-derived antigen, whereas only 20% of the Rap1a/b KD cells had done so (Figure 6.5C). By 20 min ~80% of the control B cells in contact with APCs had acquired antigen from the APCs, compared to 40-45% of the Rap1a/b KD cells (Figure 6.5C). Importantly, Rap1a/b KD B cells internalized significantly less antigen per cell than control cells (Figure 6.5D, E; and Movie 22). Thus, Rap1 is important for B cells to extract and internalize APC-bound antigens.   249    250 Figure 6.5. Rap1 is important for primary B cells to internalize antigens that are acquired from APCs   (A) Anti-Rap1a/b immunoblot of B cells transfected with a control siRNA or with siRNAs that target both Rap1a and Rap1b. Actin is a loading control. (B,C) LPS-activated primary B cells that had been transfected with control siRNA or with Rap1a/Rap1b siRNAs were mixed with anti-Igκ-expressing APCs for 5 min or 20 min and then stained for the surrogate antigen, the BCR, and α-tubulin. Representative xy confocal slices are shown. Arrows indicate antigen internalized by B cells. For the B cells that were in contact with an APC, the percent with detectable amounts of internalized antigen is graphed (C; mean ± SEM for three experiments; n = 22-60 cells per condition in each experiment.). *P <0.05, two-tailed paired t-test. (D,E) LPS-activated primary B cells that had been transfected with control siRNA or with Rap1a/Rap1b siRNAs were mixed with APCs for 20 min and stained as in B. Arrows indicate antigen that was internalized by B cells (D). Representative xy confocal slices (upper panels) as well as maximum projections of the same field (lower panels) are shown. For each B cell that was in contact with an APC, the fluorescence intensity of internalized antigen was quantified and normalized to the antigen density on the APC (E; n > 44 cells per condition from 3 independent experiments). ****P <0.0001, two-tailed unpaired t-test. Scale bars: 5 μm.        Movie 22. Rap1 is important for B cells to internalize antigens that are acquired from APCs.  A20/D1.3 cells that had been transfected with control siRNA or Rap1a/Rap1b siRNAs were stained with CellMask Deep Red and allowed to settle onto APCs expressing HEL-GFP (green). Time-lapse spinning disk confocal microscopy was used to acquire images every 30 s for 15 min (i.e. between 5 min and 20 min after initiating the B cell:APC interaction). The video is played back at 5 frames per second (150X real time). Note that the antigen (green) moves upwards from the bottom of the B cell (contact site with APC) in the control siRNA cells but not in the Rap1 siRNA cells. White boxes indicate individual representative B cells that are subsequently shown in the xz plane of view.   https://blogs.ubc.ca/jiawangthesis/2016/05/16/chapter-3-movie-2/ Password: JWthesis  251 6.2.4 Rap1 is dispensable for the internalization of soluble antigen In response to localized antigen binding and BCR signaling initiated by anti-Ig-coated beads [247] or APC-bound antigens, the active GTP-bound form of Rap1 accumulates at the site of antigen engagement and establishes polarized cytoskeletal reorganization at this site [55, 82, 247]. However, soluble antigens, which are readily internalized by receptor-mediated endocytosis [384], also induce robust Rap activation [252]. This raised the question of whether Rap activation has a general role in antigen internalization or whether this is a special requirement for antigen acquisition from polarized sources, which may require Rap-dependent cytoskeletal reorganization. To test this, control and RapGAPII-expressing A20 cells were incubated on ice with intact goat anti-IgG antibodies, which mimic soluble antigens and induce BCR signaling. The cells were then warmed to 37oC so that internalization of the bound goat anti-IgG could occur. The cells were fixed at different time points and the amount of goat anti-IgG remaining on the surface of the B cells was quantified by flow cytometry. This showed that the kinetics of BCR internalization were similar in control and RapGAPII-expressing cells. By 5 min, as much as 60% of the goat anti-IgG that had initially bound to the cells had been internalized in both control and RapGAPII-expressing cells, with maximum internalization at 15 min (Figure 6.6A,B). To directly image the BCR-mediated internalization of soluble multivalent antigens, control and RapGAPII-expressing A20 cells were incubated with Alexa Fluor 488-conjugated goat anti-IgG on ice and then warmed to 37oC. Puncta of Alexa Fluor 488