Functional Characterization of UVRAG and ANP32B in Immune Cells Using Gene- Deficient Mice

Samia Afzal

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Immunology University of Toronto

Functional Characterization of UVRAG and ANP32B in Immune Cells Using Gene-Deficient Mice

Samia Afzal

Doctor of Philosophy

Department of Immunology University of Toronto

2014 Abstract

This dissertation is a functional characterization of UVRAG and ANP32B in the immune system using gene deficient mice.

Focus 1: UVRAG is a putative tumour suppressor with potential roles in autophagy and endocytic trafficking. Despite a plethora of in vitro data, the in vivo role of this gene remains poorly understood. Here, we report the generation and characterization of the T cell specific UVRAG deficient mice. These mice exhibited normal T cell development but impaired homeostasis in the periphery, rendering them unable to clear LCMV infections. We demonstrate that UVRAG may be dispensable for T cell autophagy but it is required for the homeostatic survival and proliferation of naïve T cells. Taken together, our data provide insights into the control of autophagy in T cells and identify UVRAG as a new regulator of naïve T cell homeostasis.

Focus 2: ANP32B belongs to a highly conserved and broadly expressed family of

ANP32 proteins that have been implicated in many cellular functions. Based on previous analyses of mice bearing targeted mutations of Anp32a or Anp32e, there has been speculation that all ANP32 proteins play redundant roles and are dispensable for normal

ii development. Here, we report that ANP32B expression is associated with poor prognosis in human breast cancer. Using ANP32B-deficient mice, we demonstrate a hierarchy of importance for the mammalian Anp32 genes, with Anp32b being the most critical for normal development. In addition, we reveal important roles for ANP32B in differentiation, homeostasis and activation of immune cells. Together, these are novel findings implicating ANP32B in mammalian development and immune cell function.

iii

For my parents

Acknowledgments

The research presented in this dissertation could not have been possible without the help of many different individuals. I have been truly fortunate to have met and worked with many talented people through the course of these studies.

First and foremost, I must thank my esteemed supervisor Dr. Tak Mak for allowing me the opportunity to do graduate studies in his laboratory. Tak provided an amazing scientific environment and gave me the freedom to pursue projects of my interest. His rare insight, and quiet confidence have pushed me to become an independent scientist. I would also like to acknowledge my committee members, Tania Watts, JC Zúñiga-Pflücker and Stephan Girardin for their continued guidance and scientific insight throughout my graduate studies. Many thanks go to Dr. Jennifer Gommerman, Dr. Michelle Anderson and Dr. Navdeep Chandel for agreeing to be part of my examination committee.

I would like to thank the many members of the Mak Lab, both past and present, for their mentorship and camaraderie over the years. In particular those who instilled in me the passion for scientific research and took time to teach me, including: Sophie Vasseur, Margareta Wilhelm, Enrico Arpaia, Patrick Reilly, Zhenyue Hao, Chiara Gorrini, Susan McCracken, Momoe Itsumi, Yunfei Gao, Gloria Lin, Dirk Brenner, Anne Brustle and Satoshi Inoue. I would also like to thank my fellow graduate students in the Mak lab with whom I have had the pleasure of sharing the ups and downs of graduate life: Mike Tusche, Ying-Ju Jang, Carol Cheung, Amy Lin, Elize Shirdel, Stephanie Sue, Rebecca Menzies, Dave McIlwain, Christine Chio, Isaac Harris, Val Lapin, Brian Hershenfield, Dave Cescon, and Alan Tseng. I thank you all and look forward to collaborating with you in the future.

The work in this thesis would not have been possible without the important administrative contributions of various individuals, including: Irene Ng, Marissa Luchico, Rejeanne Puran, Ed Balyut, Lynne Omotto, and Anna Frey. Many thanks also go to Maureen Cox, Patrick Brauer and Mary Saunders for help with scientific writing.

I would especially like to thank my friends Shaliny Ramachandran, Thirumahal Selvanantham, Ann McPherson, Naomi Shuman, Val Lapin and Maryam Akrami. Thank you for making me smile through the rough patches and celebrating the good times!

Next, I want to thank my family. Your support and encouragement throughout the course of my academic and non-academic life has been integral to my success. I dedicate this thesis to my parents. Mom and dad, I have no words to express my gratitude for all the sacrifices you have made to allow me to fulfill my dreams. You have instilled within me a burning desire to achieve, and strive to be the best I can be. I cannot thank you enough for all that you have given me.

Lastly, I am indebted to my parents-in-law and my husband, Ali Shahzada, for their love, support and patience throughout these long graduate studies. I feel blessed to have you all in my life and I thank you for encouraging my dreams!

Table of Contents

Abstract ...... ii

Table of Contents ...... vii

List of Tables ...... xii

List of Figures ...... xiii

Publications and Attributions ...... xvii

Abbreviations ...... xix

1 Chapter 1: Introduction ...... 2 1.1 Overview ...... 2 1.1.1 Autophagy ...... 4 1.1.2 Autophagy Readouts ...... 10 1.1.3 T cell Biology ...... 14 1.1.4 Overview of Disease Models ...... 22 1.1.5 Roles of Autophagy Genes in the Immune system ...... 27 1.1.6 UVRAG ...... 32 1.1.7 ANP32B ...... 40

2 Chapter 2: Autophagy-independent functions of UVRAG are essential for Naïve T cell survival and homeostasis ...... 47 2.1 Abstract ...... 47 2.2 Introduction ...... 48 2.3 Materials and Methods ...... 50 2.3.1 Mice ...... 50 2.3.2 Immunoblotting ...... 50 2.3.3 Flow Cytometry ...... 51 2.3.4 Intracellular Flow Cytometry ...... 51 2.3.5 ELISA ...... 51 2.3.6 DNA Damage-Induced apoptosis ...... 52 2.3.7 ROS ...... 52 2.3.8 T Cell Purification and Activation ...... 52

vii

2.3.9 3H-Thymidine Incorporation ...... 52 2.3.10 BM Chimeras ...... 53 2.3.11 Autophagy ...... 53 2.3.12 CFSE dilution ...... 54 2.3.13 Lymphopenia Induced Homeostatic Proliferation ...... 54 2.3.14 In vitro CD4+ Th cell Differentiation ...... 54 2.3.15 EAE Induction ...... 55 2.3.16 Ovalbumin-Induced Asthma ...... 56 2.3.17 LCMV Infections and Vaccinia Infections ...... 56 2.3.18 Viral Titres ...... 57 2.3.19 CTL Cytotixicity ...... 57 2.3.20 Statistics...... 58 2.4 Results ...... 59 2.4.1 UVRAG is expressed in T lymphocytes ...... 59 2.4.2 UVRAG is essential for embryonic development but dispensable for thymocyte development ...... 59 2.4.3 URfl/fl; Lck-Cre mice exhibit peripheral T cell lymphopenia ...... 66 2.4.4 UVRAG is essential for maintaining NKT cell numbers ...... 69 2.4.5 UVRAG function in T cells is cell-intrinsic ...... 72 2.4.6 UVRAG deletion leads to an enhanced memory marker profile on T cells ...... 77 2.4.7 UVRAG deficiency does not render T cells more sensitive to cell death induction 81 2.4.8 UVRAG does not regulate autophagy in T cells ...... 84 2.4.9 UVRAG is required for homeostatic expansion of naïve T cells in lymphopenic host 88 2.4.10 UVRAG is a negative regulator of TCR-mediated T cell proliferation ...... 92 2.4.11 UVRAG is required for IL-7 induced survival of naïve T cells ...... 95 2.4.12 UVRAG does not regulate CD127 trafficking, STAT5α activation, or BCL-2 upregulation ...... 98 2.4.13 UVRAG deficiency does not protect mice from developing EAE ...... 103 2.4.14 UVRAG deficiency impairs Th2 responses during OVA-mediated asthma induction ...... 107 2.4.15 UVRAG is essential for CD8+ T cell expansion during acute LCMV infection . 110

viii

2.4.16 UVRAG is not required for CD8+ effector T cell function ...... 116 2.4.17 UVRAG is essential for memory CD8+ T cell generation, expansion and function 120 2.5 Discussion ...... 125

3 The Acidic Nuclear Phosphoprotein 32kDa ANP32 (B)-Deficient Mouse Reveals a Hierarchy of ANP32 Importance in Mammalian Development ...... 134 3.1 Abstract ...... 134 3.2 Introduction ...... 136 3.3 Materials and Methods ...... 138 3.3.1 Prognostic marker identification ...... 138 3.3.2 Northern blotting ...... 138 3.3.3 Real-time RT-PCR ...... 138 3.3.4 Mice ...... 138 3.3.5 Plasmids and primers ...... 139 3.3.6 Gene targeting ...... 139 3.3.7 Genotyping ...... 141 3.3.8 Immunoblotting ...... 141 3.3.9 Cell populations ...... 141 3.3.10 Apoptosis ...... 142 3.3.11 Proliferation ...... 142 3.4 Results ...... 143 3.4.1 ANP32B as a potential prognostic marker in human cancer ...... 143 3.4.2 Anp32b mRNA levels are linked to cell proliferation in mice ...... 145 3.4.3 Gene-targeting of Anp32b in mice ...... 151 3.4.4 ANP32B protects against perinatal lethality ...... 154 3.4.5 Surviving Anp32b-/- mice show reduced size and decreased lifespan ...... 158 3.4.6 Normal cellular apoptosis and proliferation in the absence of ANP32B ...... 164 3.4.7 ANP32B masks a role for ANP32A in essential development ...... 169 3.5 Discussion ...... 175

4 Investigating the Role of ANP32B in the Immune System ...... 179 4.1 Abstract ...... 179 4.2 Introduction ...... 181

4.2.1 ANP32B is a member of a conserved protein family...... 181 4.3 Materials and Methods ...... 183 4.3.1 Mice ...... 183 4.3.2 Immunoblot Analysis ...... 183 4.3.3 Flow Cytometry ...... 183 4.3.4 T Cell Purification and Activation ...... 183 4.3.5 Fetal Liver Chimeras ...... 184 4.3.6 Generation of murine BMDCs ...... 184 4.3.7 Statistical Analysis ...... 184 4.4 Results ...... 185 4.4.1 ANP32B protein levels are linked to proliferative tissues in mice ...... 185 4.4.2 ANP32B protein appears to be cleaved in splenocytes ...... 187 4.4.3 Anp32b-/- mice exhibit reduced splenic cellularity ...... 189 4.4.4 Mixed bred Anp32b-/- mice display a normal immune system ...... 191 4.4.5 Backcrossed ANP32B deficient mice show defects in B/T cell development and homeostasis ...... 197 4.4.6 ANP32B is indispensable for B/T cell development in fetal liver chimeras ..... 199 4.4.7 B cell differentiation is dysregulated in Anp32b-/- mice ...... 201 4.4.8 Anp32b regulates rpS6 expression in resting and activated T cells...... 203 4.4.9 ANP32A and ANP32E do not regulate the expression of phospho-rpS6 in T cells 206 4.4.10 Reduced activation of Anp32b null dendritic cells ex vivo...... 208 4.5 Discussion ...... 210

5 Chapter 5: Final Perspectives ...... 215 5.1 Summary ...... 215 5.1.1 UVRAG ...... 215 5.1.2 ANP32B ...... 215 5.2 Future Directions ...... 216 5.2.1 UVRAG ...... 216 5.2.2 ANP32B ...... 220 5.3 Concluding Remarks ...... 221

6 References ...... 223

List of Tables

Table 3.1. Primer sequences used in this analysis ...... 140

Table 3.2. Reduced survival rate of Anp32b-deficient mice...... 155

Table 3.3. The survival rate of Anp32b-deficient mice depends on genetic background...... 155

Table 3.4. Normal survival rate of Anp32b-deficient embryos at E17.5...... 157

Table 3.5. Loss of Anp32b reveals a moderate viability defect attributable to Anp32a. 170

Table 3.6. Loss of Anp32b does not reveal any defect attributable to Anp32e...... 172

Table 3.7: A single functional Anp32b allele is sufficient for mouse survival...... 173

Table 4.1. Analysis of immune system cell populations in wildtype and Anp32b-/- littermate mice...... 195

xii

List of Figures

Figure 1.1. Overview of the mammalian autophagy pathways...... 5

Figure 1.2. Class III PI3K complexes in mammalian cells...... 8

Figure 1.3. UVRAG-HOPS complex involved in autophagosomal maturation...... 10

Figure 1.4. Schematic of the stages of T cell development and maturation in the thymus...... 16

Figure 1.5. Signaling pathways involved in naïve T cell homeostasis...... 17

Figure 1.6. Schematic of IL-7 receptor signaling...... 20

Figure 1.7. The different T helper cell lineages, and their signature cytokines...... 22

Figure 1.8. Immunological processes involving autophagy...... 28

Figure 1.9. Schematic structure of UVRAG and its functions...... 33

Figure 1.10. Dual roles of UVRAG in autophagy and endocytic trafficking...... 37

Figure 1.11 Schematic of the structure of ANP32 proteins ...... 40

Figure 2.1. Generation and validation of conditional Uvrag-deficient mice ...... 61

Figure 2.2. Loss of UVRAG does not impair T cell development or maturation...... 64

Figure 2.3. Impaired T cell homeostasis in the periphery of URfl/fl; Lck-Cre mice...... 67

Figure 2.4. Impaired homeostasis of iNKT cells in URfl/fl; Lck-Cre mice...... 70

Figure 2.5. UVRAG-deficient T cells show a defect in bone marrow reconstitution...... 75

Figure 2.6. Altered marker profile of residual UVRAG-deficient T cells...... 80

xiii

Figure 2.7. Normal T cell apoptosis in secondary lymphoid organs of URfl/fl;Lck-Cre mice...... 82

Figure 2.8. UVRAG-deficient T cells undergo normal autophagy...... 86

Figure 2.9. UVRAG-deficient T cells show defects in lymphopenia-induced expansion in vivo...... 90

Figure 2.10. Loss of UVRAG leads to T cell hyper-proliferation...... 93

Figure 2.11. UVRAG-deficient T cells are refractory to IL-7-mediated survival signalling...... 96

Figure 2.12. UVRAG does not influence CD127 trafficking, Stat5α activation, or Bcl-2 expression...... 101

Figure 2.13. UVRAG deficiency does not prevent mice from developing EAE...... 105

Figure 2.14. Reduced inflammation in URfl/fl; Lck-Cre mice in response to induction of OVA-mediated asthma...... 108

Figure 2.15. URfl/fl; Lck-Cre mice mount a weak response to LCMV infection...... 112

Figure 2.16. Altered time course of T cell response to LCMV infection in URfl/fl;Lck-Cre mice...... 114

Figure 2.17. UVRAG is not required for effector CD8 T cell function...... 118

Figure 2.18 UVRAG is required for generating long term MPECs post acute LCMV infection...... 121

Figure 2.19. UVRAG deficiency impairs memory responses to LCMV-GP-expressing Vaccinia virus...... 123

Figure 2.20. UVRAG is important for many aspects of T cell biology...... 126

xiv

Figure 2.21. UVRAG is required for optimal primary and secondary immune response to LCMV infection...... 127

Figure 2.22. Proposed mechanism of action of UVRAG downstream of IL-7 signalling in naïve and memory T cells...... 129

Figure 3.1. ANP32B mRNA expression is a marker for aggressive breast cancer...... 144

Figure 3.2. Elevated Anp32b mRNA expression correlates with increased cell proliferation...... 147

Figure 3.3. Correlation of elevated Anp32b mRNA expression with cell proliferation. 149

Figure 3.4. Generation and validation of Anp32b-deficient mice...... 152

Figure 3.5. Phenotypes of Anp32b-deficient mice...... 160

Figure 3.6. Weights of Anp32b-deficient male mice...... 162

Figure 3.7. No aberrant apoptotic response in Anp32b-deficient cells...... 165

Figure 3.8. No proliferation defects in Anp32b-deficient cells...... 167

Figure 4.1. ANP32B protein expression correlates with proliferative tissues...... 186

Figure 4.2. Immunoblot analysis of ANP32B expression in thymus and spleen of WT and Anp32b-/- deficient mice...... 188

Figure 4.3. Gross analysis of Anp32b-/- immune system...... 190

Figure 4.4. Anp32b-/- mice display a largely normal immune system...... 193

Figure 4.5. Dysregulated B/T cell development and homeostasis in backcrossed Anp32b-/- mice...... 198

Figure 4.6. Flow cytometric analysis of B cell development in the bone marrow...... 199

Figure 4.7. Dysregulated B/T cell homeostasis in fetal liver chimeras...... 200

Figure 4.8. Mixed bred and backcrossed Anp32b-/- mice exhibit increased MZ and decreased FO B cells populations...... 202

Figure 4.9. rpS6 expression is dysregulated in resting and activated T-lymphocytes. ... 204

Figure 4.10. rpS6 is induced with a delayed kinetics in Anp32b -/- T cells after activation...... 205

Figure 4.11. Phospho-rpS6 expression in ANP32A and ANP32E deficient T cells...... 207

Figure 4.12. Defective Stimulation of Anp32b-/- dendritic cells in response to LPS...... 209

xvi

Publications and Attributions

Chapter 2

Afzal S, Hao Z, Itsumi M, Brenner D, Gao Y, Brustle A, Abouelkheer Y, Mak TW.

Autophagy-independent functions of UVRAG are essential for naïve T cell survival and homeostasis. (Manuscript in Preparation)

UVRAG deficient mice generated by Z. Hao, EAE experiments performed by D.

Brenner, Th differentiation experiments performed alongside A. Brüstle, LCMV experiments conducted with the help of M. Itsumi, Asthma experiments conducted with

Y. Gao, phospho-STAT5 experiment performed with help of G. Lin. All other in vivo and ex vivo work was carried out by myself.

Chapter 3

Reilly PT, Afzal S, Lui K, Mak TW. The generation and Characterization of ANP32E Mice. PLoS One. 2010. 5(10): e13597.

Reilly PT, Afzal S, Gorrini C, Lui K, Bukhman Y, Wakeham A, Haight J, Ling TW,

Cheong CC, Elia A, Turner PT, Mak TW. The Acidic Nuclear Phosphoprotein 32kDa

(ANP32)B-deficient mouse reveals a hierarchy of ANP32 importance in mammalian development. ProcNatlAcadSci U SA. 2012. 108(25):10243-10248. (A version of this manuscript appears in Chapter 3.)

xvii

Chapter 4

Afzal S, Gorrini C, Lechmann M, Reilly PT, Mak TW. Investigations into the Role of

ANP32B in the Immune System. (Manuscript in Preparation)

All experiments were conducted by myself with some help from C.Gorrini and Lechmann

xviii

Abbreviations

3-MA Methyladenine

Actb β-actin

ANP32 Acidic Nuclear Phosphoprotein 32kDa

Anp32a−/− Anp32a-deﬁcient

Anp32a−/− Anp32b-deﬁcient

ANP32B Acidic Nuclear Phosphoprotein 32B

Anp32e−/− Anp32e-deﬁcient

Armstrong Acute LCMV variant

ATG14L Barkor

Autophagy Macroautophagy

BAL Bronchoalveolar Lavage

Bcl-2 B-cell lymphoma 2

Bif-1 Endophilin-1

BM Bone Marrow

C/Vps Class C Vacuolar Protein Sorting

C/Vps Class C/Vps

CCD Coiled-Coil Domain

CD62L L-selectin

CFSE Carboxyfluorescein Succinimidyl Ester xix

Class III PI3K Class III Phosphatidylinositol 3-kinase

Clone-13 Chronic LCMV variant

CMA Chaperone-Mediated Autophagy

CNS Central Nervous System

DCF 2',7'-dichlorodihydrofluorescein diacetate (H2DCFDA)

DCs Dendritic Cells

DN Double Negative

DP Double Positive

DSB Double Strand Break

EAE Experimental Autoimmune Encephalitis

EGFR Epidermal Growth Factor Receptor

EM Electron Microscopy

ER Endoplasmic Reticulum

ERGs Early Response Genes

FADD Fas Associated protein with Death domain

FasL Anti-Fas ligand

FBS Fetal Bovine Serum

FO Follicular Zone

GFP Green Fluorescent Protein

GP33 Glycoprotein 33

HIV Human Immunodeficiency Virus

IL-2 Interleukin 2

IL-7 Interleukin 7

IL-7R Interleukin 7 receptor

IL-7α IL-7 receptor alpha or CD127 iNKTs Invariant Natural Killer T cells

IR Gamma Irradiation iTregs Induced regulatory T cells

Jak1 Janus kinase 1

Jak3 Janus kinase 3

KLF-5 Kruppel-Like Factor 5

KO Knockout

LC3 ATG-8

LCMV Lymphocyte Choriomeningitis Virus

LN Lymph Node

LPS Lipopolysaccharide

LRRs Leucine-Rich Repeats

MAPK Mitogen Activated Protein Kinase

MCHC Mean Corpuscular Hemoglobin Concentration

MCV Mean Corpuscular Volume

xxi

MEFs Mouse Embryo Fibroblasts

MOG Myelin Oligodendrocyte Glycoprotein

MS Multiple Sclerosis

MSI Microsatellite Instability

MZ Marginal Zone

NEJM New England Medical Journal

NFκB Nuclear Factor Kappa B

NP396 Nucleoprotein 396

OVA Ovalbumin

P1 Postnatal day 1 p62 SQSTM1/sequestosome 1

PBL Peripheral Blood

PBS-/- Phosphate Buffered Saline without Calcium or Magnesium

PCR Polymerase Chain Reaction

PE Phycoerythrin

PI3K Phosphoinositide 3-kinases

PI3P Phosphatidylinositol 3-phosphate

PMA Phorbol 12-Myristate 13-Acetate

PP2A Protein Phosphatase 2A

PR Proline Rich

xxii

qPCR Quantitative real-time PCR

RBCs Red Blood Cells

RFP Red Fluorescent Protein

RINT-1 RAD50-Interacting protein 1

ROS Reactive Oxygen Species rpS6 Ribosomal Protein S6 rRNAs Ribosomal RNAs

Rubicon RUN domain and cysteine-rich domain containing Beclin-1 interacting protein

S6K1 rpS6 kinase 1

S6K2 rpS6 kinase 2

SP Single Positive sp-MHC Self peptide major histocompatibility complex

STAT5 Signal Transducer and Activator of Transcription 5

TBP TATA Binding Protein

TCR T cell receptor

Tfh Follicular helper T cells

Th T helper

Tregs Regulatory T cells

ULK UNC-51-like kinase

xxiii

UV UltraViolet

UVRAG UltraViolet irradiation Resistance-Associated Gene

WBCs White Blood Cells

WT Wild type

XP Xeroderma Pigmentosum

xxiv

Chapter 1

Introduction

1 Chapter 1: Introduction 1.1 Overview

The two genes chosen for this dissertation, UVRAG and ANP32B, are involved in a broad array of biological processes through diverse molecular mechanisms. However, most of the studies investigating the functions of these genes have been based on in vitro work and thus, the in vivo role of these genes remains elusive to date. Therefore, I sought to investigate the cellular functions of these genes by using a more powerful approach of reverse genetics or gene deficient mice.

Gene deficient mice are useful for elucidating the physiological function of a gene. The simultaneous deletion of the gene in the entire organism at gestation makes it possible to identify the tissues and stages of development that are most affected by the loss of the gene. Research can then be focused on these specific areas to determine the true function of the gene in vivo. The other benefit of using this technique is that it allows for study into the function of a gene in both normal cells and diseased cells. One can induce pathologies in gene deficient mice using infection models or breeding with other disease prone mice to study the role of the gene in disease. In the case that conventional deletion leads to embryonic lethality, it is now possible to generate conditional gene deficient mice and inducible gene deficient mice. As the names suggest, these allow for the deletion of the gene in specific cell types and at a certain time of development, respectively. Such advances have rendered gene deficient mice powerful tools for evaluating the physiological functions of novel genes.

I chose to focus on the immune system, and particularly T lymphocytes, as the

ideal system with which to study the normal function of a gene in vivo. Immune system is well suited for this analysis because one can examine many cellular processes, including differentiation, survival, apoptosis and proliferation, both in vivo and ex vivo using well established experimental assays. Our laboratory has previously taken advantage of this approach to elucidate the functional significance of such broadly expressed genes as

Chk1, Brca1 and Survivin [1-3].

This chapter will review the existing literature on autophagy, T cell biology,

UVRAG, and ANP32 family of proteins.

1.1.1 Autophagy

1.1.1.1 Forms of Autophagy

Autophagy is an evolutionarily conserved and highly regulated degradation process that plays important roles in multiple cellular processes. Mammalian cells can undergo three types of autophagy (Fig. 1.1): microautophagy, in which cytosolic components are directly engulfed by the lysosome; chaperone-mediated autophagy (CMA), in which chaperone proteins selectively bind target proteins and deliver them to the lysosome; and finally, macroautophagy, in which double membrane autophagic vesicles sequester materials within the cell and deliver them to lysosomes [4]. Of these, the latter (hereafter referred to as autophagy) is the most thoroughly researched and best understood.

Under steady state conditions, autophagy is active at a basal level. It can be used to degrade defective cell organelles, long-lived proteins, and endocytosed material [5, 6].

Under conditions of metabolic stress, such as starvation, or growth factor deprivation, cells undergo higher levels of autophagy [7-10]. There, it is used to break down cytoplasmic constituents to recycle metabolic intermediates and generate energy [11].

Although autophagy is primarily a cell survival mechanism, it has been shown to lead to cell death as well [12, 13]. How autophagy may be regulating these paradoxical functions remains poorly understood. Nevertheless, dysregulation of this process results in multiple pathophysiologies, including neurodegeneration, liver and heart disease, myopathies and cancer [14-17].

mTOR(

ULK1/2(

1( 2# 3(

Figure 1.1. Overview of the mammalian autophagy pathways.

Autophagy, as it is exists in mammalian cells, can be classified into three subtypes: microautophagy, CMA and macroautophagy. Macroautophagy pathway can be further categorized into 4 main stages: 1) Vesicle nucleation to form an isolation membrane or phagophore. This requires the activity of a class III PI3K complex. 2) Elongation and closure of the isolation membrane around cytoplasm to form a double membrane bound autophagosome. This requires two ubiquitin-like conjugation systems that function to conjugate LC3 to phosphatidylethanolamine (PE). 3) Maturation of the autophagosome- Fusion of the autophagosome with the endo-lysosomal pathway. 4) Breakdown of autophagosomal contents in the lysosome. Figure modified from [18] with permission from Annual Reviews of Immunology.

Autophagy is characterized by de novo formation of the double membrane structure, called the autophagosome. Most mammalian autophagy genes can be categorized into three functional groups that are involved in the induction, the elongation and closure, and the maturation of the autophagosome (Fig. 1.1). The process begins with the nucleation of the isolation membrane around the material that is destined for degradation. This is followed by the elongation and closure of the membrane to form the autophagosome. The autophagosome then undergoes subsequent maturation and fusion with the endo-lysosomal pathway, leading to the degradation of the autophagosomal cargo. The macromolecules generated from degradation are transported back to the cytosol for recycling and reuse in biosynthetic pathways. Genes involved in each of these functional groups and their regulation will be reviewed in some detail below.

1.1.1.2 Induction of the autophagosome

Induction of autophagy is under the control of the nutrient sensor mammalian target of rapamycin, mTOR. When nutrients are abundant, mTOR inhibits autophagy through phosphorylation of UNK51-like kinases (ULK1/2) (Figure 1.1). Conversely, when nutrients become scarce, mTOR is inactivated, leading to the dephosphorylation of

ULK1/2 and induction of autophagy [19, 20].

Induction of autophagy also requires the formation of the class III phosphatidyl inositol-3 kinase (PI3K) complex. Mammalian PI3K complex is composed of three core components: human vacuolar protein sorting (hVps34), p150 and Beclin-1. The complex forms when hVps34 and its adaptor protein p150 bind to Beclin-1 [21]. hVps34 is the main component of this complex and is a lipid kinase that can convert phosphatidylinositol to phosphatidylinositol 3-phosphate (PI3P). This activity is thought 6

to be necessary for recruiting autophagy components to the isolation membrane, thereby promoting the induction of autophagy [22]. Given the essential role of this complex in the induction of autophagy, its inhibition with pharmacological agents, such as 3- methyladenine (3-MA), and wortmannin, is commonly used to block autophagy [23].

1.1.1.2.1 Regulation of the Class III PI3K complex

Several proteins have been found to positively regulate the activity of the PI3K in mammalian cells (Fig. 1.2). For instance, the interaction of the core complex with

UltraViolet (UV) irradiation Resistance Associated Gene (UVRAG) was shown to positively regulate the induction of autophagy [24]. Similarly, the interaction of the PI3K complex with ATG14L (or Barkor) is thought to enhance the activity of the complex and promote the induction of autophagy [25, 26]. Interestingly, UVRAG is not found in the

ATG14L bound PI3K complex and this is thought to be due to the overlap of binding sites of these proteins on Beclin-1 [27]. Another positive mediator of autophagy induction is Bif-1 (also known as Endophilin-1), whose interaction with Beclin-1 requires the presence of UVRAG [28].

Under steady state conditions, cells have evolved multiple mechanisms to negatively regulate the formation and activity of the PI3K complex. One such regulator is the RUN domain and cysteine-rich domain containing Beclin-1 interacting protein (also known as Rubicon). Rubicon’s interaction with hVps34 has been shown to reduce PI3K activity and hence the induction of autophagy. Interestingly, Rubicon binding to Vps34 was dependent on the presence of UVRAG in the complex (Fig. 1.2) [27]. Another well- defined negative regulator of the PI3K complex is the anti-apoptotic protein B-cell lymphoma-2 (Bcl-2). Under steady state conditions, Bcl-2 can sequester Beclin-1 and 7

prevent its association with the PI3K complex of autophagy [29]. Conversely, stress conditions cause cellular Bcl-2 levels to decrease, thus freeing up Beclin-1 to mediate the induction of autophagy.

Ambra( Rubicon( BifH1( BeclinH1( BeclinH1( BeclinH1( BeclinH1( Atg14L( UVRAG( UVRAG( UVRAG( hVps34( hVps34( hVps34( hVps34(

p150( p150( p150( p150(

Autophagosome(( Inhibitory( Autophagosome( PI3K(complex( Biogenesis( Matura#on(

Figure 1.2. Class III PI3K complexes in mammalian cells. The core components of the mammalian Class III PI3K complex are hVps34, p150 and Beclin-1. This core complex can interact with UVRAG or ATG14L to mediate autophagosome formation and induction of autophagy. The core complex and UVRAG can also interact with Rubicon but this interaction negatively regulates the induction of autophagy.

1.1.1.3 Elongation/Closure of the autophagosome: Ubiquitin like conjugation systems

Autophagosome formation relies on two ubiquitin-like conjugation systems that are necessary for membrane elongation and closure (Fig. 1.1) [30]. In the first system,

ATG12 is initially activated by E1-like enzyme ATG7, transferred to E2-like enzyme

ATG10 and then covalently linked to ATG5 [31-36]. This ATG5-12 complex then non- covalently binds ATG16L, which promotes its homo-oligomerization and targets the 8

complex to the elongating autophagosome [37]. Importantly, this complex provides the macromolecules necessary for autophagosome elongation and remains on the outer membrane of the autophagosome until completion. It also acts as an E2 ubiquitin ligase for the second ubiquitin-like conjugation system (Fig. 1.1): LC3 (known as Atg8 in yeast)

[38]. In this system, LC3 is first cleaved by ATG4B to produce LC3-I [39, 40], activated by E1-like enzyme ATG7, transferred to E2-like enzyme ATG3, and finally covalently linked to phosphatidylethanoalmine (PE) [31-33, 40, 41]. This conjugated product is known as LC3-II and is present on both sides of the autophagosomal membrane. It is necessary for completion of the autophagosome [42, 43].

1.1.1.4 Maturation of the autophagosome

Unlike the other steps of the autophagic process, the molecular mechanisms behind the maturation and fusion of the autophagosome with the endo-lysosomal pathway are poorly understood [44, 45]. Liang and colleagues (2008) recently implicated UVRAG in mediating this process [46]. In their model, UVRAG recruits the HOPS/Class C Vacuolar protein sorting (C/Vps) tethering proteins to the autophagosomal membrane, leading to the activation of Rab7 GTPase and fusion with late endosomes and/or lysosomes (Fig.

1.3) [46]. Upon fusion with lysosomes, hydrolytic enzymes come in contact with autophagosomal contents, leading to their degradation [47]. This ultimately results in the recycling of the contents and their subsequent transport to the cytosol where they can be reused in biosynthetic pathways. A defect in autophagosomal maturation can result in accumulation of autophagosomes in the cell leading to serious disorders [14-17].

Rab7( GTP(

Vps41( Vps39( Class#C# Vps# Vps18( Vps11( HOPS# complex# Vps16( Vps33(

UVRAG(

Autophagosome( Matura#on(

Figure 1.3. UVRAG-HOPS complex involved in autophagosomal maturation.

UVRAG recruits the Class C Vps/HOPS tethering complex to the autophagosmal membrane. This leads to the activation of Rab7 GTPase and fusion with late endosomes and lysosomes.

1.1.2 Autophagy Readouts

The increasing significance of autophagy in disease mandates the use of reliable and quantitative assays to measure autophagy. Traditionally, autophagy was observed by electron microscopy (EM) [23, 48]. However, the qualitative nature of this technique has limited its use as a reliable readout of autophagy. Researchers have since exploited unique features of autophagosome dynamics to develop more robust assays for measuring

autophagy (reviewed in [49]). The following section will review some of the commonly used readouts of autophagy: namely, LC3 turnover, GFP-LC3 fusion proteins, p62 expression and mitophagy.

1.1.2.1 LC3 Turnover

LC3 cleavage has become the most widely used readout of autophagy. In this method, the differential electrophoretic mobility of LC3-I and II forms is exploited to determine changes in the extent of autophagy. However, given the increased affinity of the antibody for the LC3-I protein, and the varied expression of LC3-I and LC3-II forms in cell lines and tissues, this method has led to many erroneous results [50]. Consequently, a new consensus has been reached to simply normalize overall LC3-II levels to a loading control to get a better estimate of autophagy [51]. Notably, this approach only allows for the measurement of static autophagy in the cells. As a result, an increase in the LC3-II form could be indicative of increased synthesis of LC3-II or a block in autophagosomal maturation, leading to the accumulation of LC3-II. To differentiate between these two possibilities, the field now requires the performance of autophagic flux assays [49]. For measuring flux, one must include treatments with autophagy inhibitors, such as chloroquine or bafilomycin A1, to show that LC3-II, which is normally degraded when autophagy is completed, in fact accumulates in the presence of inhibitors. Failure to accumulate upon inhibitor treatment would indicate a block in autophagy in cells.

1.1.2.2 GFP Fusion Proteins

LC3 fused with Green Fluorescent Protein (GFP)-LC3 has also been widely used to examine localization and quantification of autophagosomes by fluorescent microscopy

[52]. In this method, GFP-LC3 transfected cells undergoing autophagy show the 11

characteristic green puncta staining. Among the limitations of this method are: its laborious and subjective nature, formation of GFP-LC3 protein aggregates appearing as punctate structures independently of autophagy, and the inability to examine late stage autophagy due to the sensitivity of the GFP-LC3 to lysosomal pH [49, 53, 54]. To overcome the latter problem, Red Fluorescent Protein (RFP)-GFP-LC3 fusion proteins have been developed. When this fusion protein is over-expressed in cells undergoing autophagy, red and green staining merges to produce yellow puncta [54]. When autophagosomes fuse with the lysosomes, GFP is degraded but RFP remains resistant to degradation. As a result, red puncta can still be observed and are indicative of cells undergoing late stage autophagy [54].

1.1.2.3 p62 Expression p62 (SQSTM1/sequestosome 1) expression is specifically used as a readout of selective autophagy. Selective autophagy is a subclass of autophagy that involves the binding of target cargo by adaptor proteins that then mediate their transfer to the forming autophagosome. p62 is specifically involved in turnover of protein aggregates by autophagy [55]. Through binding to ubiquitinated protein aggregates as well as the LC3 protein, p62 promotes the degradation of protein aggregates via autophagy [56].

Therefore, its degradation, alongside protein aggregates, is indicative of autophagy. It appears to be a reliable marker of selective autophagy as p62-associated protein aggregates were shown to accumulate in ATG7 deficient mouse hepatocytes and deletion of p62 prevented the accumulation of these protein aggregates [57, 58]. However, p62 expression data should be interpreted carefully because p62 has been shown be transcriptionally regulated by oxidative stress, and the Ras oncogene [59]. Its

involvement in autophagy independent functions has led to the recommendation that it too be combined with other autophagy readouts, such as LC3 turnover, for a more reliable measure of protein aggregate turnover by autophagy [49].

1.1.2.4 Mitophagy

Mitophagy, or selective clearance of the mitochondria by autophagy, is increasingly being used as a readout of autophagy in immune cells [60]. Reticulocytes use mitophagy to clear their mitochondria and defects in this process can results in anemia [60].

Similarly, T cells down regulate mitochondrial content upon thymic exit [61]. This is thought to be a coping mechanism for the increased oxygen tension present in the blood

[61]. Cells that are unable to reduce mitochondria can have increased levels of mitochondrial reactive oxygen species (ROS) leading to cell death. This down regulation is at least partly attributed to mitophagy or the process of mitochondrial autophagy. As such, ATG3, ATG5, ATG7, and Vps34 knockout T cells are unable to effectively reduce their mitochondria, leading to increased ROS production and cell death in the periphery

[61-64]. Interestingly, however, a recent T cell specific Beclin-1 knockout failed to show a defect in mitochondrial reduction, raising two possibilities. Firstly, autophagy independent functions of ATG3, ATG5, ATG7 and Vps34 are responsible for mitochondrial downregulation. Secondly and more likely, Beclin-1 may not be required for T cell specific mitophagy. Regardless, examining mitochondrial content has become a standard test for characterizing T cell specific autophagy related gene knockouts.

1.1.3 T cell Biology

1.1.3.1 T Cell Development

The thymus is the primary site of T cell development. Hematopoetic progenitors seed the thymus as early as embryonic development and undergo a series of well-defined and highly ordered stages of development that can be identified by expression of specific cell surface markers (Fig. 1.4). The process is dependent on a host of environmental cues and signaling pathways. Double negative (DN) cells- lacking the expression of CD4 and CD8 co-receptors as well as T cell receptor (TCR)- can be subdivided into four developmental stages (DN1-4), based on the expression of CD44, CD25 and CD117 [65-67]. DN1 cells are the most primitive and multipotent of the group [68-71]. They are defined as

CD44+CD25-CD117+ and are capable of proliferating and differentiating into both TCR

α/β and TCR γ/δ T cells [72]. Co-expression of CD25 marks their entry into the DN2 stage (CD44+CD25+CD117+) [73, 74]. At this time, TCRβ, TCRγ, and TCRδ locus begin to undergo rearrangement, mediated by the expression of RAG1/2 proteins [67,

75]. Recombination continues into the DN3 stage of development (CD44-CD25+CD117-

) where cells are irreversibly committed to the T cell lineage and undergo β selection [67,

76]. Cells with a productively rearranged TCRβ chain associate with pre-TCRα chain to form the pre-TCR complex. Signaling through the pre-TCR and CD3 upregulates key signaling pathways that are necessary for mediating allelic exclusion, proliferation, differentiation and survival of these cells past this checkpoint [77-79]. Successful signaling of the pre-TCR complex selects cells for the α/β lineage differentiation. It also leads to the loss of CD25 expression on DN3 cells, marking their entry into DN4 differentiation stage (CD44-CD25-CD117-). DN3 cells with successfully rearranged

TCRγ, and TCRδ are also signaled to survive but their further development is beyond the scope of this dissertation.

1.1.3.2 T cell maturation

DN4 cells begin to express the CD4 and CD8 co-receptors and become double positive

(DP) cells (Fig. 2.4). It is also at this stage that cells begin to rearrange their TCRα locus.

While in the DP stage, cells with a successfully rearranged TCR loci can undergo another selection process, namely the process of positive selection. Positive selection ensures that cells have a complete and functional TCR (i.e. they have both TCR α and β chains and can recognize antigen in the context of MHC). Only cells that elicit a weak signal upon interaction with self peptide, presented in the context of MHC class I or II by cortical epithelial cells, are positively selected and receive a survival signal. Interleukin-7

(IL-7) receptor and Bcl-2 are thought to be important for this process [80-83]. Cells that respond too strongly to self peptide-MHC (sp-MHC) complex undergo negative selection and cell death occurs in a Bim dependent manner [84, 85]. Similarly, cells that do not receive a positive signal die by neglect. Negative selection and death by neglect remove

95% of thymocytes from the repertoire and only a small percentage of DP cells produce a functional TCR to become single positive CD4 or CD8 T cells [86]. There are two models that explain how cells commit to CD4 or CD8 lineages but the details of these models are not relevant for the purposes of this study. It is sufficient to point out that this commitment restricts the class of MHC molecules used to recognize peptides. CD8 T cells recognize antigen in the context of MHC Class I molecules and CD4 cells recognize antigen in the antigen in the context of MHC Class II molecules. These mature, single positive (SP) cells then leave the thymus to form the naïve pool of T cells in the

periphery.

DN#Cells# Posi?ve#&## Nega?ve# CD4( Selec?on# βCselec?on# CD4+# CD8C# DN1( DN2( DN3( DN4( DP(

CD8( CD44+# CD44+# CD44C# CD44C# CD4+# CD25C# CD25+# CD25+# CD25C# CD8+# C# CD117+# CD117+# CD117C# CD117C# TCRB+# CD4 +# # # # CD8 TCRβ,γ,δ# T#cell## TCRα# rearrangement# commitment# rearrangement#

Figure 1.4. Schematic of the stages of T cell development and maturation in the thymus.

1.1.3.3 T cell homeostasis

The peripheral pool of T cells is composed of antigen-inexperienced (naïve cells) and antigen-experienced (effector and memory) T cells. These populations can be distinguished on the basis of L-selectin (CD62L) and CD44 expression, whereby the naïve T cells are CD62LhiCD44lo, central memory T cells are CD62LhiCD44hi and effector T cells are CD62LloCD44hi. Homeostasis of each of these cell types is tightly regulated to prevent autoimmunity while maintaining strong protective immunity and will be reviewed here.

1.1.3.4 Naïve T cell homeostasis

Mounting an effective immune response requires a diverse and large pool of naïve T cells. Naive T cell numbers are balanced by thymic output, homeostatic survival and proliferation in the periphery, and cell death. Their homeostatic survival and proliferation is dependent on low-level signal from sp-MHC-TCR interaction as well as survival signal from homeostatic cytokines (Fig. 1.5). One cytokine in particular, IL-7, has been shown to be key for naïve T cell survival in vivo [87-89].

Thymus# Periphery#

Peripheral#Pool#

Death( DN# CD4+CD8+#DP# SP# Naïve#Pool# β#selec?on# Survival(

Homeosta/c( Prolifera/on( Lineage# Commitment# Memory#

DN1# DN2# DN3# DN4# TCR# ILC7# Ac#va#on(and( ILC15# Prolifera#on(

Figure 1.5. Signaling pathways involved in naïve T cell homeostasis.

TCR and homeostatic cytokine signaling are crucial for naïve T cell development and homeostasis. Figure adapted from: http://www.nimr.mrc.ac.uk/research/benedict-seddon/.

1.1.3.5 Homeostasis of effector cells

T cell survival is also tightly regulated during an active immune response. A typical T cell immune response involves priming of T cells to the specific antigen, expansion of antigen-specific T cells, contraction of those cells once the challenge is over, and finally, maintenance of a small percentage of memory T cells. Antigen presenting cells prime T cells in secondary lymphoid organs. They do so by presenting foreign antigen to T cells in the context of MHC peptide. This leads to T cell activation and differentiation into effector cells. Whereas naïve and memory T cells do not rely on interleukin 2 (IL-2) signals for survival, the importance of IL-2 in effector T cell survival and proliferation remains a controversial issue [90, 91]. Effector cells expand at an enormous rate and it is estimated that antigen-specific T cell population can increase by more than 1000-fold during the expansion phase of infection [92, 93]. Once the infection is resolved, a great majority of these effector T cells are eliminated. A number of factors, including Bim and

Fas receptor, are necessary for mediating this contraction [94-97].

1.1.3.6 Homeostasis of memory T cells

Once an infection is resolved, only 5-10% of effector T cells survive. These are the memory T cells and they persist for the lifetime of the organism. These cells are maintained by homeostatic proliferation, dividing once every three to six weeks. It was recently shown that, in contrast to naïve T cells, memory T cell homeostasis does not require interactions with foreign antigen or sp-MHC [93, 98, 99]. Rather, it depends on

IL-7 and IL-15 signaling (Fig. 1.5) [87-89, 100]. It is thought that under steady state conditions, IL-7 is crucial for homeostatic survival of antigen specific CD8 T cells, while

IL-15 is more important for homeostatic proliferation of CD8 memory T cells [100, 101].

1.1.3.7 IL-7 receptor signaling

IL-7 is a pleiotropic cytokine with crucial roles in many aspects of T cell biology. Its receptor (IL-7R) is composed of two subunits (Fig. 1.6): the unique IL-7 receptor alpha

(IL-7α or CD127) chain and the common gamma chain (CD132), which is also utilized by other common gamma chain family of cytokines (including IL-2, IL-4, IL-15, and IL-

21) [100]. The two subunits are associated with Janus kinase1 and 3 (Jak1/3), respectively. IL-7 binding to the receptor on cell surface triggers Jak1/3 downstream signal transduction pathways leading to the activation of Signal Transducer and Activator of Transcription 5 (STAT5) and Class I PI3K pathway. Signaling downstream of STAT5 and PI3K pathways is necessary for promotion of cell survival, cell cycle regulation, metabolism and migration (Fig. 1.6).

IL@7(

T(cell( ILC7Rα# γ#chain# Cytoplasm( Jak1# Jak3#

P( PI3K# P(STAT5(

Akt#

P(STAT5( ILC7# FOXO# !BCL@2,(MCL@1( "BAX,(BIM,(BAD( p27#

Cell(Survival( Prolifera#on(

Figure 1.6. Schematic of IL-7 receptor signaling.

IL-7 binding to the receptor triggers Jak/STAT5 and Class I PI3K signalling cascades. These signaling pathways in turn mediate cell survival, migration and metabolism of the cells.

1.1.3.8 Role of IL-7 in T cell development and homeostasis

IL-7 signaling is crucial for T cell development and homeostasis. It is required for differentiation of TCR α/β cells and production of TCRγ/δ cells in the thymus. IL-7R- deficient mice lack TCRγ/δ cells due to the requirement for IL-7 in inducing TCRγ locus rearrangement [102]. In addition to its roles in T cell development, IL-7 is important for

T cell homeostasis. Its expression is high on naïve and memory T cells while activated T

cells have low expression of this cytokine [103]. Consistently, it is essential for survival of naïve and memory T cells in vivo [101]. IL-7 and IL-7R deficient mice have shown that IL-7 is dispensable in T cell activation and memory T cell generation but essential in memory T cell survival [103-106]. Conversely, exogenous IL-7 led to enhanced survival of activated and memory T cells [107-109]. Taken together, these findings demonstrate that IL-7 is a key regulator of T cell development and homeostasis.

1.1.3.9 T helper Differentiation

CD4+ T helper (Th) cells can be categorized into at least 5 distinct subsets (Fig. 1.7):

Th1, Th2, Th17, regulatory T cells (Tregs) and follicular helper T cells (Tfh). This division is based on the cytokines they secrete in response to antigen receptor stimulation and the effector functions they are involved in. Depending on the type of pathogen and route of infection, CD4+ T cells differentiate into specific T helper lineages that will be suitable for fighting the infection. Accordingly, Th1 cells are characterized by production of the signature cytokine IFN-γ and are important for cellular immunity. Th2 cells produce IL-4, IL-5 and IL-13 and are important for humoral immunity. Th17 cells produce IL-17A, IL-17F, IL-21, IL-22, TNF, and GM-CSF and are important for driving inflammation. Induced regulatory T cells (iTregs) produce TGFβ and IL-10 and are important for immunosuppression. Finally, Tfh cells produce IL-21 and they provide cognate help to antigen specific B cells. In vitro culture systems have provided researchers the ability to define the signaling molecules involved in T helper lineage differentiation.

Signature( Cytokines( Th1( IFNCγ# (T@bet)(

ILC4# Th2( ILC5# (Gata3)( IL@4( ILC13#

Naïve( ILC17a# + IL@6,(TGFβ( CD4 (T( Ac#vated( Th17( ILC17f# cells( (( (Rorγt)( TGFβ(

iTreg( ILC10# (Foxp3)( TGFβ#

T_( ILC21# (Bcl@6)(

Figure 1.7. The different T helper cell lineages, and their signature cytokines.

CD4+ T helper (Th) cells can be categorized into at least 5 distinct subsets. Th1 cells, important for cellular immunity, produce signature cytokine IFN-γ. Th2 cells, important for humoral immunity, produce IL-4, IL-5 and IL-13. Th17 cells, important for driving inflammation, produce IL-17A, IL-17F, IL-21, IL-22, TNF, and GM-CSF. Induced Regulatory T cells (or Tregs important for immunosuppression) produce TGFβ and IL- 10. Tfh cells, necessary for providing cognate help to B cells, produce IL-21.

1.1.4 Overview of Disease Models

Many experimental disease models have been developed to examine dynamics of T cell responses in vivo and ex vivo. For CD4+ T cells, these include experimental autoimmune

encephalitis (EAE), and asthma (Th2). For CD8+ T cells, lymphocytic choriomeningitis virus (LCMV) is a commonly used infection model. Each of these models will be discussed in detail below.

1.1.4.1 EAE model: Th1 and Th17

Experimental autoimmune encephalitis (EAE) is utilized as a mouse model of the human autoimmune disease multiple sclerosis (MS). EAE can be induced by active or passive induction. In active induction, mice are immunized with myelin antigens to induce EAE whereas the passive induction requires the adoptive transfer of primed myelin-specific T cells into naïve mice (reviewed in [110]). The active method is relevant to this study and will be discussed further. In this method, mice are immunized with myelin antigen(s), such as myelin oligodendrocyte glycoprotein (MOG) peptide. APCs present MOG peptide-MHC complexes to T cells in secondary lymphoid organs, leading to the breakdown of peripheral tolerance and the activation of MOG specific T cells. These T cells begin to proliferate and differentiate into effector cells. Effector cells have the ability to exit the secondary lymphoid organs and go into the periphery. Some effector cells cross the blood brain barrier and enter the central nervous system (CNS), where they are reactivated by CNS-resident APCs presenting MOG antigens [111, 112]. Reactivated effector T cells produce pro-inflammatory cytokines, IFN-γ, IL-17, TNF-α and GM-CSF, causing direct injury to the CNS [110]. Effector T cell can also produce chemokines, which induce the recruitment of nonspecific cellular effectors such as γδ T cells, monocytes, macrophages, and neutrophils into the CNS [113, 114]. Activation of these inflammatory cells and the bystander damage caused by them ultimately leads to destruction of the myelin-sheathed axonal tracts and the formation of lesions [110]. Both

Th1 and Th17 cells have been implicated in contributing to the pathogenesis of EAE

[115-118].

Following immunization with MOG peptide, mice are monitored daily for the development of disease. Disease typically begins around the second week following priming and is characterized by an ascending hind limb paralysis that begins in the tail and spreads to involve the hind limbs and forelimbs. The disease is graded on a 0–5 scale.

Thus, because EAE requires both Th1 and Th17 cells for pathogenesis [115-118], it is a prototypical model for investigating the potential role of novel genes in Th1 and Th17 mediated autoimmunity in vivo.

1.1.4.2 Ovalbumin Induced Asthma model: Th2

Ovalbumin-induced asthma (OVA-induced asthma) is a mouse model of allergic asthma.

It is widely used by the scientific community to recapitulate the airway eosinophilia, pulmonary inflammation and elevated IgE levels found during asthma [119]. In this model, mice are given repeated intraperitoneal injections of an allergen, ovalbumin.

APCs can come in contact with the allergen and present it to naïve T cells in the context of MHC class II. This leads to the differentiation of naïve CD4+ T cells into Th2 cells that are capable of secreting cytokines, such as IL-4, IL-5 and IL-13 [119]. These cytokines are necessary for B cell activation, leading to the production of ovalbumin specific IgE antibodies. These antibodies can circulate in the body or attach to the high affinity receptors on the surface of mast cells and basophils. When mice are subsequently exposed to the same allergen intra-nasally, this induces cross-linking of the antigen- specific IgE leading to degranulation of the mast cells and basophils, and release of inflammatory mediators, such as histamine and prostaglandins [119]. Inflammatory 24

reaction is further enhanced by the recruitment of eosinophils, collectively leading to enhanced vascular permeability, smooth-muscle contraction and mucus production. This ultimately results in the symptoms of asthma including airway constriction, shortness of breath, coughing, and wheezing.

Endpoint analyses in mice can include evaluation of the airway hyper- responsiveness, the quantification of analytes in the serum or bronchoalveolar lavage

(BAL) and quantification of eosinophilia in the BAL. Thus, because a robust Th2 immune response is critical to triggering the full-blown asthma in this model, this is an ideal system for exploring the role of new genes in Th2 mediated immunity in vivo.

1.1.4.3 LCMV: CD8 T cells

Lymphocytic choriomeningitis virus (LCMV) is a powerful and well-defined infection model for studying T cell responses in vivo. It is a RNA-based natural rodent virus belonging to the Arenaviridae family. The virus consists of two single-stranded RNA strands that encode the viral polymerase and glycoproteins necessary for viral fusion and replication. The virus exists as multiple strains that are each associated with a particular pattern of progression in the host. It can exist as an acute variant (Armstrong), which is resolved 1-2 weeks following infection and a chronic variant (Clone-13), which can persist for months. Of relevance to this thesis is the acute Armstrong strain, which is cleared 8 days post-infection. During acute LCMV (Armstrong strain) infection, a robust antigen-specific CD8 T cell response is elicited in spleen and lymph nodes of mice. These cells are thought to produce large amounts of IFNγ and TNFα, and exhibit cytotoxic capabilities in vivo [120]. CD4 T cells and B cells are also generated in the response but their requirement in clearance of virus remains controversial. What is not controversial is 25

the absolute requirement for CD8 T cells in clearance of the virus. Any manipulation of

CD8 T cell biology, which results in defective T cell development, signaling or proliferation will likely render these mice immuno-compromised and susceptible to viral infection [121].

Original studies on antigen specific T cells were limited by bulk assays of T cell function such as cytokine secretion, proliferation and cytotoxicity studies. In these assays, the effect of antigen-specific T cells (which comprise about 50% of cells after expansion phase) is ‘diluted’ by non-specific T cells. Thus, it was important to develop reagents that allowed for the detection of antigen-specific T cells and the measurement of cytokine secretion on a per cell basis. The development of MHC tetramer technology has filled the former gap in quantifying the antigen-specific immune response in vivo [122]. Tetramers are a tetravalent complex of biotinylated MHC molecules bound to peptides and fluorescently labeled streptavidin complexes. These MHC-peptide complexes can bind to

T cell clones with the appropriate TCR specificity. For instance, glycoprotein 33 (GP33) and nucleoprotein 396 (NP396) specific tetramers are commonly used LCMV specific tetramers that can be used to measure antigen-specific CD8 T cell response in vivo.

Intracellular flow cytometry has filled the second gap of measuring cytokine secretion at a cellular level. Collectively, LCMV is a powerful infection model for studying the role of new genes in CD8 mediated T cell response in vivo.

1.1.5 Roles of Autophagy Genes in the Immune system

Autophagy has been shown to play an essential role in both the innate and adaptive immune system (Fig. 1.8). It is required for innate restriction and clearance of intracellular pathogens, such as parasites, bacteria and viruses [123-126]. In addition, antigen presenting cells use autophagy for detection of pathogens via cross-presentation of endogenous antigens [127, 128]. Recently, the role of autophagy has also been extended to the adaptive immune system, where it is critical for the maintenance of B and

T cells. The following section will examine the current state of knowledge on the role of autophagy in T cells.

Figure 1.8. Immunological processes involving autophagy.

Innate immune cells use autophagy to detect microbes, to present pathogen-associated antigens on cell surface, and to degrade pathogens. Adaptive immune cells use autophagy for cell survival and homeostasis. Figure from [129] with permission from Nature Immunology.

1.1.5.1 Autophagy induction in T cells

There has been a great deal of research in the last decade on the role of autophagy in the adaptive immune system, and particularly in T lymphocytes. It is now appreciated that genes encoding autophagy machinery, including ATG3, ATG5, ATG7, Beclin-1, LC3,

Vps34, and p62, are expressed in T lymphocytes [130-132]. Moreover, it is evident that autophagy is constitutively active in resting T cells [8, 130, 133-135] and is upregulated upon T cell starvation, cytokine stimulation and activation [125, 130, 131, 134]. This upregulation has been validated by increased conversion of LC3I to II and by visualization of LC3-II puncta by fluorescence microscopy [130, 131]. Overall, autophagy appears to play a cytoprotective role in T cells.

Paradoxically, autophagy has also been observed to lead to cell death in T cells.

Autophagy observed in T cells lacking growth factors [130], caspase-8, FADD[136], or

Irgm-1[137], have been shown to contribute to cell death. Consistently, cell death of T cells exposed to the Env protein of Human Immunodeficiency Virus (HIV) protein can be prevented when autophagy is blocked [138]. This data would suggest that autophagy promotes death in T cells. However, T cell-specific autophagy gene knockouts, including

Atg3, 5, 7 and Beclin-1, have shown otherwise. All of these knockouts have strong defects in cell survival [61, 63, 131, 139]. Thus, it appears that autophagy related genes might be important for cell survival in normal T cells, but excessive autophagy or autophagy induced by certain genetic deletions can lead to cell death [139].

1.1.5.2 Autophagy genes in T cell development and homeostasis

Mice with T-cell specific inactivation of core autophagy proteins, ATG3, ATG5, and

ATG7, exhibited mildly reduced thymic cellularity without a specific developmental block or impairment in cell survival ex vivo [61, 63, 131]. In contrast, peripheral CD4+ and CD8+ T cell numbers were greatly reduced in the absence of these autophagy genes.

Moreover, peripheral T cells demonstrated impaired survival and TCR-induced proliferation ex vivo [61, 63, 131]. In addition, they displayed alterations in 29

naïve/activated/memory profile. For the case of ATG5, and ATG7, these defects were largely attributed to impaired mitochondrial clearance, leading to ROS buildup and cell death [61, 62, 131]. In the case of ATG3 deficient T cells, in addition to defects in mitochondrial clearance, T cells also displayed defects in ER clearance suggesting the importance of autophagy in maintaining organelle homeostasis in naïve T cells [63].

T cell specific gene deficient mice have been used to investigate the importance of components of Class III PI3K complex in T cells. Beclin-1 deficient T cells showed similar reductions in peripheral T cells as was seen for ATG5 and ATG7 [139]. Cells exhibited impaired survival and Th polarization leading to resistance to the induction of

EAE. Intriguingly, however, mitochondrial turnover was normal in these cells and enhanced sensitivity to cell death was attributed to the inability to eliminate the proapoptotic molecules. As autophagy was never directly measured in this study, the results are open to interpretation. Another group also examined the role of Beclin-1 in T cells using bone marrow chimeras [140]. In contrast to the first group, they failed to show a role for Beclin-1 in peripheral T cell maintenance or in T cell autophagy [140].

Similar to Beclin-1, the role of Vps34 in T cells remains controversial. Using Lck-

Cre to conditionally deleted Vps34 in T cells, McLeod et al showed that although T cell development and peripheral maintenance is greatly affected, Vps34 is dispensable for the induction of autophagy in T cells [135]. Impaired survival in these cells could not be attributed to defects in mitochondrial clearance and Vps34 deficient T cells showed normal proliferation in response to TCR stimulation ex vivo. The authors discovered a novel role for Vps34 in vesicular trafficking and T cell survival. They showed that Vps34 was indispensable for the intracellular recycling of IL-7Rα to the surface of naïve T cells 30

through an autophagy independent mechanism. In stark contrast, a recent publication by

Parekh et al., in which they use CD4-Cre to delete Vps34 in developing T cells, showed impaired autophagy in the absence of Vps34 [141]. Therefore, the role of Class III PI3K components, Vps34 and Beclin-1, in T cell autophagy remains controversial.

Collectively, these findings have revealed a unique role for autophagy in T cells.

However, they have also raised important questions about its regulation in this cell lineage. As UVRAG is seen as an essential regulator of autophagy in mammalian cells, we set out to investigate its role in T cell biology. The following section will review the existing literature on UVRAG.

1.1.6 UVRAG

1.1.6.1 Discovery of UVRAG

UVRAG was originally discovered in a genetic screen for providing partial resistance to

UltraViolet (UV) irradiation in Xeroderma Pigmentosum (XP) cells [142]. In fact, its name was chosen to reflect its ability to complement resistance in that system. UVRAG maps to the tumor susceptibility locus on human chromosome 11q13, which is frequently implicated in many human cancers [143-146]. Structural and biochemical studies suggest that UVRAG protein contains 5 domains (Fig. 1.9): an amino-terminal proline rich (PR) region (residues 1-41), a Ca2+ dependent phospholipid binding C2 domain (residues 42-

147), a coiled-coil domain (CCD, residues 200-269), a CEP-63 binding domain (residues

270-442) and a carboxy-terminal domain (residues 443-669). Importantly, each of these regions has been implicated in binding to distinct proteins that are involved in distinct cellular pathways, thus suggesting that UVRAG may be important for a wide array of biological functions (Fig. 1.9).

Bax CEP63 DNA-PK

PR C2 CCD

Bif-1 C/Vps Beclin-1

Figure 1.9. Schematic structure of UVRAG and its functions.

UVRAG contains 5 domains: an amino-terminal proline rich (PR) region (residues 1-41), a Ca2+ dependent phospholipid binding C2 domain (residues 42-147), a coiled-coil domain (CCD, residues 200-269), a CEP-63 binding domain (residues 270-442) and a carboxy-terminal domain (residues 443-669). Each of those domains allows UVRAG to interact with different proteins (shown above and below) and thus mediate such processes as promotion of autophagy, endocytic trafficking, inhibition of cell death, centrosomal stability and DNA damage repair.

1.1.6.2 Functions of UVRAG

It was originally accepted that UVRAG primarily functions to promote autophagy.

However, recent literature has raised the possibility that UVRAG may also have autophagy independent functions in mammalian cells. These include: involvement in endocytic trafficking, in DNA damage repair and in prevention of apoptosis (Fig. 1.9).

We will examine some of these functions in detail below.

1.1.6.3 UVRAG positively regulates autophagy

Since its initial discovery, UVRAG has attracted much attention for its role in autophagy.

In vitro work by several laboratories has shown that UVRAG promotes autophagosome formation by associating with Beclin-1 and positively regulating the activity of the Class

III PI3K in autophagy [24, 28, 46]. This binding to Beclin-1 has been mapped to CCD of

UVRAG. UVRAG has also been shown to bind to Bif-1, through the PR domain and this interaction allows Bif-1 to bind to the Class III PI3K complex, thereby inducing autophagy and membrane curvature. Together, these findings suggest that UVRAG is crucial for autophagosome formation in mammalian cells.

Recently, the role of UVRAG in autophagy was extended to the maturation of the autophagosome. Not only does UVRAG induce autophagosome formation by interaction with the Beclin-1/Vps34 complex [24] but it also promotes autophagosome maturation

[46, 147]. The involvement of Rab proteins and the HOPS complex had been previously demonstrated in autophagosome maturation, but how this machinery is recruited to the autophagosomal membrane was unclear [44, 45]. Liang et al proposed that UVRAG recruits the Class C tethering proteins to the autophagosomal membrane, leading to the activation of the GTPase Rab7 and fusion with late endosomes and/or lysosomes [46].

Overexpression of wild type but not mutant UVRAG, which lacks C/Vps binding, was able to stimulate Rab7 GTP hydrolysis, a step necessary for fusion. Importantly, the 34

authors showed that UVRAG-C/Vps interaction in autophagosome maturation was physically and functionally independent of UVRAG-Beclin-1 interaction in autophagosome formation [46]. Whereas the binding to Beclin-1 occurs through the CCD domain, binding to C/Vps complex occurs through the N terminal C2 domain. These results imply that UVRAG has dual roles in the regulation of autophagy: firstly in the formation of the autophagosome via Beclin-1 and secondly, in the maturation of the autophagosome via C/Vps proteins.

Consistently, UVRAG gain-of-function experiments led to enhanced activation of autophagy and inhibition of cancer cell proliferation, suggesting that UVRAG may control cell growth at least in part through its regulation of autophagy [24, 28]. Recently,

Song et al addressed the in vivo role of UVRAG, by generating mice with the transposon- induced deletion of UVRAG and reported defective autophagy in mouse embryonic fibroblasts (MEFs) and cardiomyocytes [148]. In summary, UVRAG appears to be an essential positive regulator of autophagosome formation and maturation in mammalian cells.

1.1.6.4 UVRAG promotes endocytic trafficking

Recent literature has also raised the possibility that UVRAG may have autophagy independent functions in mammalian cells. One such function is its involvement in endocytic trafficking. This originated from the observation that UVRAG is not exclusively localized to autophagosomes [46]. It can also be found on endosomes, specifically tethered to the C/Vps complex on early endosomes [46]. There, it is thought to promote late endocytic trafficking independently of its roles in autophagy [46]. Several lines of evidence support this claim. To begin, UVRAG can physically associate with 35

C/Vps complex at a region that is distinct from its binding to Beclin-1. Secondly,

UVRAG gain of function experiments showed enhanced endocytic trafficking and epidermal growth factor receptor (EGFR) degradation. Consistently, loss of function studies demonstrated sustained EGFR expression at cell surface and continued signaling

[27, 46]. Thirdly, it was through regulation of Notch receptor endocytosis, rather than autophagy, that UVRAG was able to mediate proper organ rotation in Drosophila [149].

Finally, UVRAG promotes endosomal fusion both in vitro and in vivo, as measured by endosome fusion assays [42]. Collectively, these studies have helped identify a distinct role for UVRAG in promoting endocytic trafficking independently of autophagy.

The topological similarity of autophagosome and endosome delivery to lysosomes suggests that maturation of these processes may rely on similar machinery to promote lysosomal delivery (Fig. 1.10). In light of this knowledge, this novel role of UVRAG should not come as a complete surprise. Several others proteins, including HOPS and Rab proteins, have been shown to play a role in both processes. However, the possibility remains that this role may be an artifact of the overlap between the two pathways and thus more precise genetic studies are needed to delineate the effects. A recent report by

Liang et al seems to minimize these concerns [150]. Authors identified a role for

UVRAG in ER-Golgi membrane transport. They showed that UVRAG is able to bind

PI3P, leading to its localization to ER, where it interacts with RAD50-Interacting Protein

1 (RINT-1) to coordinate Golgi-ER retrograde [150]. Therefore, UVRAG generally appears to be essential for trafficking pathways within the cell.

hVps34# p150#

Bif@1(

Figure 1.10. Dual roles of UVRAG in autophagy and endocytic trafficking.

UVRAG is involved in autophagosome formation and maturation. In addition to its dual roles in autophagy, it mediates trafficking of endosomes and lysosomes. Figure adapted from [46] with permission from Natrue Cell Biology.

1.1.6.5 UVRAG is a putative tumour suppressor

Several studies have implicated UVRAG in cancer suppression. First of all, UVRAG maps to the tumor susceptibility locus on human chromosome 11q13, which is frequently implicated in such human cancers as breast, colorectal, and gastric cancers [143-146].

Secondly, tumor suppressors PTEN and p53 can induce UVRAG expression while oncogenes Bcl-2 and Akt/PKB can inhibit UVRAG expression [24, 151]. Thirdly,

UVRAG gain-of-function experiments led to the induction of autophagy activation and suppression of colon cancer growth suggesting that UVRAG inhibits cancer cell growth

[24]. Beclin-1 binding mutant of UVRAG failed to inhibit cell growth while dominant negative UVRAG was found to hinder autophagy and promote tumor cell growth in culture and xenograft mouse models. Finally, UVRAG appears to be mono-allelically deleted in approximately 20% of microsatellite instability (MSI) colorectal cancers and

7% of MSI gastric cancers in humans [144, 152]. Therefore, there appears to be a strong correlation between loss of UVRAG and tumor susceptibility.

1.1.6.6 UVRAG mediates DNA double strand break repair

UVRAG’s role was recently extended to double strand break (DSB) repair [153].

Deletion of UVRAG led to increased gamma H2AX levels in somatic and embryonic cells. Moreover, cells were more sensitive to DNA damage induced death. Zhao et al convincingly showed that UVRAG deficient cells were unable to repair DNA damage independently of autophagy. Furthermore, UVRAG could be localized to nuclear sites of

DNA damage where it was shown to interact with DNA-PK, a key player in DSB repair.

This interaction was necessary for DNA-PK activation.

1.1.6.7 UVRAG inhibits cell death

UVRAG’s involvement in cell death originated with the observations that tumor therapies

(chemotherapy and irradiation) induce UVRAG expression in tumour cells and that

UVRAG interacts with Bax [154]. This binding to Bax, via the C3 domain of UVRAG, led to inhibition of Bax translocation to mitochondria, reducing cytochrome c release and caspase activation and inhibiting cell death [154]. Notably, suppression of UVRAG increased Bax-induced apoptosis in tumor cells, whereas overexpression of UVRAG by 38

gene transfection inhibited Bax-induced apoptosis. Thus, these findings suggest a potential role for UVRAG in prevention of cell death.

1.1.6.8 Summary

UVRAG encodes a protein considered to be an essential regulator of multiple cellular pathways, including autophagy, endocytic trafficking, DNA damage repair and cell death. As the in vivo role of this gene remains poorly understood, in Chapter 2 of this dissertation, we use the gene-deficient mouse model approach to study its function in T cell biology in vivo.

1.1.7 ANP32B

1.1.7.1 ANP32B is a member of a conserved protein family

ANP32B (a.k.a APRIL, PAL31, PHAP1b) is a mammalian member of the highly conserved “acidic nuclear phosphoprotein 32kDa” (ANP32) family of gene products

[155]. These metazoan-specific factors are characterized by the presence of an amino- terminal leucine-rich repeat (LRR) domain and a carboxy-terminal (C-terminal) region that is highly enriched in acidic amino acid residues (Fig. 1.11) [155]. These features are found in ANP32 proteins from mapmodulin, the single representative in Drosophila, and the three vertebrate family members identified, ANP32A, ANP32B, and ANP32E [156,

157].

Figure 1.11 Schematic of the structure of ANP32 proteins

The family is characterized by the presence of an amino-terminal leucine-rich repeat (LRR) domain and a carboxy-terminal (C-terminal) region that is highly enriched in acidic amino acid residues. Figure adapted from Patrick Reilly.

1.1.7.2 ANP32 proteins are involved in multiple cellular processes

ANP32 proteins have been implicated in a broad array of physiological processes, including cell differentiation [158-161], apoptotic cell death [162-168], and cell proliferation [169-171]. Diverse mechanisms have been postulated for how these proteins perform their function(s). Some studies indicate that ANP32 proteins may directly control enzymatic activities, such as via inhibition of protein phosphatase 2A (PP2A) [172-174] or activation of caspases [164, 165, 167, 168, 175]. Others suggest that they may regulate intracellular transport at nuclear pores or microtubules [176-178]. Several studies present evidence that nuclear ANP32 proteins may influence transcription either through the

“inhibitor of acetyl transferases” (INHAT) complex [179-182] or by direct effects on transcription factors [183, 184]. Most of the above reports have focused on ANP32A, the founding member of the ANP32 family, but none of these studies has specifically excluded a particular ANP32 protein as contributing to the activities examined.

1.1.7.3 Activities associated exclusively with ANP32B

More recent work has demonstrated functions that are exclusive to the ANP32B protein, at least in humans. Firstly, ANP32B, but not ANP32A, controls the expression of the dendritic cell (DC) maturation factor CD83 by regulating the transport of its mRNA to the cytoplasm [176]. Secondly, ANP32B modulates the activity of the transcription factor Kruppel-like factor 5 (KLF5), whereas ANP32A cannot [185]. Finally, ANP32B is a caspase substrate, whereas ANP32A is not [186]. These findings suggest that ANP32B likely plays an important role in mammalian development. Some of these roles will be examined in more detail below.

1.1.7.3.1 ANP32B can influence mRNA dynamics.

All ANP32 family members have been found to influence post-transcriptional gene expression, particularly through the regulation of mRNA transport and stability. They act as adaptor proteins for HuR, an RNA binding protein also known as Human Antigen R and ELAV, and mediate its shuttling between nucleus and cytosol [176, 187-189]. HuR can bind and stabilize mRNAs containing AU-rich elements (AREs) in their

3’untranslated regions (UTRs), which would otherwise predispose them to quick degradation and deadenylation. As protein ligands for HuR, ANP32 family members may emerge as important mediators of post-transcriptional gene regulation in cellular processes such as cell growth, differentiation, metabolism, migration and cellular senescence. This function of ANP32 family becomes even more interesting in light of the fact that AREs are commonly present in mRNAs of early response genes (ERGs) such as growth factors, cytokines, lymphokines and proto-oncogenes. It was recently reported that ANP32B, in particular, participates in the regulation of CD83 expression [176], an activation marker expressed on dendritic cells (DCs) and T and B cells. ANP32B promotes the transport of CD83 mRNA from the nucleus to the cytoplasm [176, 189].

Others have shown that ANP32B is also important for the shuttling of c-fos mRNA via the export receptor chromosome region maintenance 1 (CRM1) [187, 190].

1.1.7.3.2 ANP32B regulates histone modifications.

Reversible modification of histone tails is a primary method of controlling gene transcription, and this process is altered during many important cellular processes. In vitro, ANP32 proteins can bind to histone H3 tails and alter their modification patterns, both independently and as part of the Inhibitor of Histone Acetyltransferase (INHAT)

complex [179-181, 191, 192]. Recently, ANP32B was identified as a novel histone chaperone that interacts with the transcription factor (KLF5)[185]. This interaction leads to transcriptional repression of a KLF5-downstream gene through the inhibition of histone acetylation at the KLF5 promoter region [185]. This is important because KLF5 is an important mediator for the pro-inflammatory response elicited by LPS in intestinal epithelial cells[193]. These new findings imply that ANP32B may also play a critical role in regulating innate immune responses.

1.1.7.3.3 ANP32B is implicated in the regulation of ribosomal biogenesis.

ANP32B has been found to interact with RNA polymerase I, the polymerase responsible for the transcription of all ribosomal RNAs (rRNAs) [194]. This particular interaction implicates ANP32B in the potential regulation of ribosomal biogenesis. Of all the ribosomal proteins, ribosomal protein S6 (rpS6) has attracted the most attention since it is the first, and was for many years the only one, shown to undergo inducible phosphorylation. This phosphorylation can be seen in response to a wide variety of stimuli that activate either the phosphoinositide 3-kinases (PI3K) or the mitogen activated protein kinase (MAPK) pathway.

In recent studies targeting the genes encoding rpS6, its phosphorylation residues, or its respective kinases, the unique role of rpS6 and its post-translational modification have started to be elucidated. The current model suggests that rpS6 is an indispensable ribosomal protein, and upon phosphorylation, it plays a critical regulatory role in multiple cellular and organismal processes. This importance of rpS6 is further underscored by the fact that it is the only ribosomal protein, in which heterozygosity has been shown to lead

to early embryonic lethality [195]. Hepatocytes from the liver of conditional rpS6 knockout mice have an abnormal 40S ribosomal subunit and consequently, fail to proliferate following partial hepatectomy [196]. In a different mouse model, rpS6 heterozygosity significantly reduced the number of mature T cells in peripheral lymphoid organs [30]. Even though global protein synthesis was relatively unaffected, rpS6+/- T cells failed to proliferate upon mitogenic stimulation, as a result of a block at the G1/S checkpoint of the cell cycle, and partially due to increased p53-dependent apoptosis

[197].

1.1.7.4 ANP32B can potentially influence immune related functions.

Although ANP32 family members have many redundant functions, one that has specifically been ascribed to ANP32B is an association with proliferation. Knockdown of

ANP32B inhibited cell cycle progression from G1 to S phase in T and B cell lines from rats and mice, respectively [171]. Moreover, several groups have recently reported that

ANP32B participates in the regulation of CD83 and c-fos gene expression[176, 189, 190]

. These findings, in concert with the observation that Anp32b mRNA is highly expressed in proliferative tissues [198], suggest that ANP32B likely plays an important role in the immune system. Overall, these observations provide us with a clear rationale for investigating the biological functions of ANP32 proteins in the immune system.

1.1.7.5 Published knockouts of ANP32 proteins

It is important to point out that although many biochemical activities have been attributed to this family, elucidation of the precise, non-redundant biological functions of each family member can best be elucidated by use of a reverse genetics approach. Reported

loss-of-function mutants for ANP32 family members include two independently targeted

ANP32A-deficient mice [199, 200], an ANP32E-deficient mouse [200], and a presumptive null mutant of mapmodulin in Drosophila [201]. All of these mutants were viable and fertile with no obvious abnormalities.

In Chapters 3 and 4 of this dissertation, we describe the generation and characterization of the ANP32B-deficeint mice.

Chapter 2

UVRAG

2 Chapter 2: Autophagy-independent functions of UVRAG are essential for Naïve T cell survival and homeostasis 2.1 Abstract

Ultraviolet resistance-associated gene (UVRAG) encodes a tumour suppressor with putative roles in autophagy, endocytic trafficking, and DNA damage repair. Despite a plethora of in vitro data, the in vivo role of this gene, especially in T cells, remains elusive. Attempts by our laboratory to generate the homozygous null mutants of UVRAG in mice resulted in early embryonic lethality. Therefore, we generated conditional

UVRAG deficient Lck-Cre mice that specifically lacked UVRAG expression in T cells.

Loss of UVRAG led to defects in the development of invariant Natural Killer T cells

(iNKTs) and the homeostasis of peripheral T cells. The homeostatic defect could not be explained by enhanced sensitivity to cell death or impairment of proliferation in vitro, as has been observed for other autophagy gene knockout mice. Instead, our data strongly suggest that UVRAG has non-autophagic roles that are critical for naïve T cell survival and homeostatic proliferation. Indeed, UVRAG-deficient T cells exhibit normal mitochondrial clearance and activation-induced autophagy, suggesting that UVRAG may not be essential for autophagy in this cell lineage at all. In vivo, T cell-specific loss of

UVRAG in mice dampened their immune responses to LCMV and rendered them less able to clear the viral infection. Taken together, our data provide novel insights into the control of autophagy in T cells and identify UVRAG as a new regulator of naïve T cell homeostasis.

2.2 Introduction

The previous decade has witnessed a surge in research on the role of autophagy in the adaptive immune system, and particularly in T lymphocytes. It is now appreciated that genes encoding elements of the autophagy machinery are expressed in this lineage and that autophagy can be observed in both resting and activated T cells [130, 131]. Studies of knockout mice bearing T cell-specific deletions of autophagy genes, including ATG3,

ATG5, ATG7 and Beclin-1, have revealed an indispensable role for autophagy in T cell homeostasis [61, 63, 131, 139], but have also raised important questions about regulation of this process in these cells.

UVRAG encodes a protein considered to be an essential regulator of autophagy in mammalian cells. Initially identified as a protein that complements UV sensitivity of

Xeroderma Pigmentosum group C cells [202], UVRAG has since attracted attention for its dual roles in autophagy. In vitro work by several laboratories has shown that UVRAG promotes autophagosome formation by associating with Beclin-1 and positively regulating the activity of the class III phosphatidylinositol 3-kinase (class III PI3K) in this process [24, 28, 46]. Subsequently, UVRAG promotes autophagosome maturation by binding to the C/Vps HOPS complex [147, 203]. In mouse embryonic fibroblasts (MEFs) and cardiomyocytes derived from mice bearing transposon-induced deletion of UVRAG, autophagy is defective [148]. In cancer cells, UVRAG overexpression enhances autophagy and reduces proliferation, suggesting that UVRAG may control cell growth at least in part through its regulation of autophagy [24, 28].

Several lines of evidence indicate that UVRAG also has autophagy-independent functions, at least in vitro. 1) UVRAG is not exclusively localized in autophagosomes 48

and can be found in endosomes, where it promotes endocytic trafficking by interacting with the C/Vps complex [46]; 2) In Drosophila, UVRAG mediates proper organ rotation through regulation of Notch receptor endocytosis rather than through autophagy [149]; 3)

UVRAG helps to maintain chromosomal stability by mediating DNA damage repair

[153], and 4) UVRAG suppresses apoptosis by regulating the subcellular localization of

Bax [204]. These findings collectively imply that UVRAG is a pleiotropic gene with autophagic and non-autophagic roles. Thus, UVRAG is best studied in vivo to determine precisely how it functions in any given context.

T cells offer a physiologically relevant and easily manipulated system to study the gene function and regulation in vivo. In this study, we generated and characterized conditional UVRAG knockout Lck-Cre mice to define the precise role(s) of UVRAG in T cell biology in vivo. We show that UVRAG is indispensable for T cell homeostasis in the periphery and invariant natural killer T cell (iNKT) development in the thymus. Using mixed bone (BM) chimeras and adoptive transfer mouse models, we demonstrate that

UVRAG is essential for naïve T cell survival and proliferation. Furthermore, we report that UVRAG is not required for the induction of experimental autoimmune encephalitis

(EAE) in mice, but is essential for an effective immune response to lymphocytic choriomeningitis virus (LCMV) infection. Surprisingly, T cells from our mutants showed normal autophagy. Our data therefore also strongly suggest that the observed effects of

UVRAG on T cell biology are independent of this process.

2.3 Materials and Methods

2.3.1 Mice

UVRAG floxed (URfl/fl) mice were generated using Cre/LoxP recombination and the gene targeting and screening strategies outlined in Fig. 1C. The experimental procedures used for the culture, transfection, and selection of E14K ES cells (129/Ola) have been previously described [205]. To generate mice in which UVRAG was deleted specifically in T cells, URfl/fl mice were bred with Lck-Cre transgenic mice (C57BL/6). The resulting

URfl/fl; Lck-Cre mice were backcrossed for 6–10 generations to C57BL/6 animals. Mice used in experiments were 4-16 months old unless otherwise specified. Mice were maintained under specific pathogen-free conditions in individually ventilated cages and fed 5% irradiated meal. All animal experiments were approved by the University Health

Network Animal Care Committee.

2.3.2 Immunoblotting

T cells were lysed in 1x RIPA buffer (50 mM Tris pH8, 1% Nonidet P-40, 150 mM

NaCl, 1% 0.1M PMSF, protease inhibitors; Roche). Protein extracts were fractionated by

SDS-PAGE, transferred to nitrocellulose membranes, and probed with anti-UVRAG

(MBL), anti-LC3 (MBL), anti-mouse p62 (Cell Signalling), or anti-Bcl-2 (Cell

Signalling) antibodies. Loading controls were anti-β-tubulin (Sigma-Aldrich) or anti- actin (Sigma-Aldrich) antibodies. The appropriate Alexa Fluor-conjugated secondary antibodies (Molecular Probes) were used to bind to primary antibodies and visualized using the Odyssey Infrared Imaging System (LI-COR Biosciences).

2.3.3 Flow Cytometry

Single cell suspensions were prepared from thymus, spleen or LN. Spleen and LN suspensions were treated with red blood cell (RBC) lysis buffer to remove erythrocytes.

Cells (1–2×106) were pre-incubated with Fc block for 15 min at 4°C and stained with antibodies recognizing the following antigens: CD16/CD32 (2.4G2), CD25 (PC61),

CD45R/B220 (RA3-6B2), CD69 (H1.2F3), CD4, CD8a, CD44, CD62L, TCRβ, CD5,

CD3, NK1.1, CD11c, CD11b, GR-1, CD45.1, CD45.2, PD-1 or CD95 (all from BD or eBioscience unless otherwise specified). CD1d PBS-57 Tetramer was a kind gift of Dr. T.

Mallevaey. Flow cytometry data were acquired using either a FACSCalibur (BD) or

FACSCanto (BD) flow cytometer, and analyzed with either the CellQuest software (BD) or the FlowJo analysis program (Tree Star).

2.3.4 Intracellular Flow Cytometry

Permeabilization and intracellular staining to detect STAT5, Foxp3, IL-17A or IFNγ were performed using the appropriate Cytofix/Cytoperm kits (BD Biosciences) according to the manufacturer’s instructions. Flow cytometry was performed on a FacsCalibur instrument (BD Biosciences) and analyzed using FlowJo 7.5 software (Tree Star).

2.3.5 ELISA

For quantitation of cytokines in culture supernatants, ELISAs were performed using the following kits from R & D and eBioscience according to the manufacturers’ instructions:

IL-13, IL-5, IL-6, IL-10, IFNγ and TNFα. Serum IgE was also detected by ELISA using a kit (eBioscience).

2.3.6 DNA Damage-Induced apoptosis

Cells were exposed to apoptotic stimuli as detailed in Figure 7 and cultured overnight in

RPMI complete medium. Apoptosis was evaluated by Annexin V/7-AAD or Annexin/PI staining. Briefly, cells bearing surface-bound antibodies were stained with fluorochrome- conjugated Annexin V and/or 7-AAD/PI for 15 min at room temperature in 10X Annexin

V binding buffer (all from BD Biosciences). Cells were analyzed by a FACSCalibur flow cytometer immediately after staining.

2.3.7 ROS

For measurement of ROS levels in T cells, surface stained T cells were incubated with

300nM CM-H2-DCFDA (DCF) for 5 min at 37°C in the dark. DCF (FITC channel) fluorescence was analyzed by flow cytometry.

2.3.8 T Cell Purification and Activation

T cells were purified using the IMag cell separation system (BD Biosciences) from single cell suspensions of spleen or LN prepared as described above. Briefly, total leukocytes were incubated with mouse anti-CD16/32 blocking antibodies, after which T cells were negatively selected using biotinylated anti-Ter119/B220/CD19/CD11b/Nk1.1/CD11c antibodies. For naïve T cell isolation, biotinylated anti-CD44 antibody was added to the above list. Purified T cells were stimulated with plate-bound anti-mouse CD3 (clone

2C11; BD Biosciences) plus anti-mouse CD28 (clone 37.51; BD Biosciences) antibodies at the concentrations indicated in the figures.

2.3.9 3H-Thymidine Incorporation

Purified T cells (1x105) in 200 µl complete RPMI were seeded into each well of U-

bottom 96 well plates that were pre-coated with indicated concentrations of anti-CD3 and anti-CD28 antibodies. 3H-thymidine (1 µCi) was added to each well and cells were cultured for another 8 hr, after which proliferation was determined at 48 hr or 72 hr. 3H- thymidine uptake was measured using a liquid scintillation β-counter (TopCount reader).

2.3.10 BM Chimeras

Donor mice (at least 6 weeks old) underwent two rounds of CD4+/CD8+ T cell depletion using anti-CD4 plus anti-CD8 antibodies. Bone marrow was harvested from femurs and tibias of CD45.1+ WT or CD45.2+ URfl/fl; Lck-Cre donor mice. Recipient mice (Rag-2-/- or ptprc CD45.1 C57BL/6 mice) were irradiated with 6Gy or 10Gy, respectively. WT and

URfl/fl; Lck-Cre BM samples were mixed 1:1, and either 3-5x106 mixed BM cells were transferred intravenously (i.v.) into Rag-2-/- hosts, or 7-10x106 mixed BM cells were i.v. transferred into ptprc C57BL/6 hosts. Reconstitution was monitored by examination of peripheral blood of chimeric mice at 6-16 weeks post-injection.

2.3.11 Autophagy

For Mitotracker staining, cells were stained for 30 min with 100 nM Mitotracker Green in

RPMI-1640 complete medium before surface antibody staining and analysis by flow cytometry. For the LC3I-II conversion assay, cells were treated with either medium alone, or plate-bound 2µg/ml anti-CD3 antibody plus 0.2 µg/ml anti-CD28 antibody, or anti-CD3/CD28 antibodies plus chloroquine (25µM). Cells were cultured overnight and lysates subjected to immunoblot analysis using mouse monoclonal LC3 (MBL) and mouse clonal p62 antibodies (MBL).

2.3.12 CFSE dilution

For evaluation of T cell proliferation by CFSE dilution, purified T cells were washed twice with phosphate buffered saline without calcium or magnesium (PBS-/-) and incubated with 1 µM CFSE for 15 min at 37°C. The labeling reaction was stopped with cold fetal bovine serum (FBS) and excess CFSE was removed by two additional washes in PBS-/-. CFSE dilution (and thus cell proliferation) was assessed by flow cytometry as previously described above.

For evaluation of reconstitution in lymphopenic hosts, WT (CD45.1+) and URfl/fl; Lck-

Cre (CD45.2+) cells were mixed 1:1 and labelled with CSFE as described above. Labelled mixed cells (5x106) were i.v. injected into CD45.1/2 recipient C57BL/6 mice and CFSE dilution was assessed as above.

2.3.13 Lymphopenia Induced Homeostatic Proliferation

Purified, naive cells were washed twice with PBS-/- and labeled with 1uM CFSE for 15 min at 37°C. Reaction was stopped with cold FBS and excess CFSE was removed with two additional washes in PBS-/-. CFSE dilution was analyzed by flow cytometry. WT

(CD45.1) and KO (CD45.2) labeled cells were mixed 1:1 and 5x106 cells were injected into CD45.1/2 C57BL/6 mice intravenously.

2.3.14 In vitro CD4+ Th cell Differentiation

In vitro differentiation of Th cell subsets was induced as described previously [206].

Briefly, naive CD4+CD62L+ Th cells were isolated from spleen and LN and sorted using a magnetic bead cell purification kit according to the manufacturer’s instructions

(Miltenyi). For priming, enriched naïve Th0 cells were stimulated for 72 hr with 1 µg/ml

plate-bound anti-CD3 antibody plus 1 µg/ml anti-CD28 antibody (both from BD

Biosciences). Naive cells were induced to differentiate into Tregs by addition of 3 ng/ml recombinant human TGF-β (rhTGF-β; R&D), 50 U rhIL-2 (Peprotech), and 5 µg/ml anti-

IFNγ antibody (BD Biosciences); into Th1 cells by addition of 4 ng/ml IL-12 (Peprotech) plus 50 U rhIL-2; or into Th2 cells by addition of 10 ng/ml IL-4 (Peprotech), 5 µg/ml anti-IFNγ antibody, and 50 U rhIL-2. After 72 hr of differentiation, Th2 cells were transferred from anti-CD3-coated plates to non-coated plates and treated with an additional 50 U rhIL-2. Th2 cells were rested for 2 days prior to use in experiments. For differentiation of Th17 cells, naive Th cells received 5 µg/ml anti-IFNγ antibody, 30 ng/ml rmIL-6 (Peprotech), and 2 ng/ml rhTGF-β.

2.3.15 EAE Induction

EAE was induced in mice as described previously [206]. Briefly, mice were subcutaneously immunized with 115 µg MOG35–55 peptide (Washington Biotech) emulsified in CFA (Difco) supplemented with 400 µg/ml Mycobacterium tuberculosis

(Difco). On days 0 and 2 after immunization, mice received intraperitoneal (i.p.) injection of 300 ng pertussis toxin (List Biological). Clinical signs of EAE were monitored daily according to the following criteria: 0, no disease; 1, decreased tail tone; 2, hind limb weakness or partial paralysis; 3, complete hind limb paralysis; 4, front and hind limb paralysis; 5, moribund state. Mice were sacrificed when they reached an EAE disease score of 4; these animals were assigned a score of 5 for the rest of the observation period for the purpose of calculating mean EAE disease score.

For T cell proliferation assays, mice subjected to EAE induction were sacrificed on day

14 post-immunization. Splenocytes were isolated and restimulated in vitro with indicated 55

3 concentrations of MOG33-55 peptide. Proliferation was assessed by [ H] thymidine incorporation as described above.

2.3.16 Ovalbumin-Induced Asthma

Mice were sensitized by i.p. injection of 200 µl OVA/alum solution containing 100 µg

OVA on days 1, 8 and 15. Mice were challenged with intranasal OVA/PBS on days 22,

23 and 24 (5 mg OVA administered via a nebulizer). Mice were sacrificed on day 25 and samples of BAL recovered. Serum cytokines were determined by ELISA as described above.

For determination of eosinophilia in BAL, cells were centrifuged onto slides at 1400 rpm for 5 min, and slides were dried in air for 30-120 min. Slides were fixed 5 times for 1 second each and stained in Solution I and II 3-5 times for 1 second each. Slides were washed in distilled water, dried, and observed under a light microscope to detect eosinophils.

2.3.17 LCMV Infections and Vaccinia Infections Mice (8-12 weeks old) were i.v. injected with 105 pfu Armstrong LCMV strain. PBL samples were obtained over the course of the infection and numbers of CD8+ T cells,

GP33 tetramer-specific CD8+ T cells, NP396 tretramer-specific CD8+ T cells, and CD4+

T cells were determined by flow cytometric analysis of the appropriate markers and tetramers. The following MHC tetramers were used to evaluate antigen specific T cells:

MHC I: GP33-41 H2-Db and NP396–404H2-Db Tetramers were provided by the Dr.

Ohashi. Some mice were sacrificed on day 8 post-infection and splenocytes were subjected to parallel immunophenotyping, GP33 restimulation and tetramer analyses.

For examination of memory T cell responses to LCMV, at 1 month after the primary LCMV infection, mice were i.v. infected with 105 pfu of a recombinant vaccinia virus expressing the LCMV-GP protein. At 5 days post infection, T cells were isolated from spleen and subjected to immunophenotyping, GP33 restimulation, and CTL cytotoxicity assays (see below). For GP33 restimulation, cells were incubated with no peptide or 1 µM GP33 peptide for 5 hr in the presence of brefledin A or golgi-stop. IFNγ production was measured by intracellular flow cytometry as described above.

2.3.18 Viral Titres

Viral titres were determined using a plaque forming assay as described in [207]. Briefly, tissues of LCMV-infected mice (spleen, liver, lung, brain and kidney) were homogenized and centrifuged to produce virus-containing supernatants. Supernatant dilutions were used to infect monolayers of MC57 cells, which were grown under an overlay of methylcellulose in DMEM for 2 days. Cells were subsequently fixed with PFA, stained with VL-4 rat anti-LCMV monoclonal antibody, and incubated with ortho- phenylenediamine to identify viral plaques by colour reaction. Titres were expressed as plaque-forming units (pfu/organ).

2.3.19 CTL Cytotixicity

WT CD45.1+ splenocytes were used as target cells. Half of these cells were pulsed with

GP33 and labelled with CFSE, and half were left unpulsed. Pulsed and unpulsed cells were then mixed 50:50. For CTLs, CD8+GP33+ T cells were isolated from CD45.2+ WT and URfl/fl; Lck-Cre mice previously infected with LCMV and GP33-expressing vaccinia.

CTLs and target cells were co-incubated for 5 hr at 37°C, after which CFSE dilution was measured by flow cytometry to determine target cell killing. 57

2.3.20 Statistics.

Where appropriate, all differences were evaluated using the unpaired 2-tailed Student’s t test, as calculated using GraphPad Prism software. Data are presented as mean ± SEM unless otherwise indicated. Statistically significant differences are indicated as: *, p<0.05; **, p<0.01; or ***, p<0.001.

2.4 Results

2.4.1 UVRAG is expressed in T lymphocytes

As a first step towards characterizing the physiological functions of UVRAG in mouse T cells, we used quantitative PCR (qPCR) to examine UVRAG mRNA expression in sorted primary T cell subsets. Although UVRAG mRNA was easily detected in all T lineage cells, its expression was much higher in thymocytes, particularly double positive (DP) thymocytes, than in subsets of mature peripheral T cells (Fig. 2.1A). Immunoblotting confirmed that UVRAG protein levels were higher in thymocytes than in peripheral

CD4+ and CD8+ T cells (Fig. 2.1B). Intriguingly, peripheral CD8+ T cells showed relatively higher levels of UVRAG protein than CD4+ T cells (Fig. 2.1B). These observations suggest a potential role for UVRAG in T cell biology, especially during T cell development and in mature CD8+ T cells.

2.4.2 UVRAG is essential for embryonic development but dispensable for thymocyte development

To decipher the functions of UVRAG in T cells in vivo, we first generated homozygous null mutant UVRAG-deficient mice but found that complete loss of UVRAG resulted in embryonic lethality by day E7.5 (data not shown). To circumvent this lethality, we used the cre-loxP system to create conditional UVRAG-deficient mice (URfl/fl mice) (Fig.

2.1C). These mutants were bred with Lck-Cre transgenic mice to delete UVRAG specifically in T cells (URfl/fl;Lck-Cre mice). These mice were born at the expected

Mendelian ratio and appeared phenotypically normal (data not shown). High efficiency of

UVRAG deletion in peripheral T cells of these animals was confirmed by immunoblot analysis (Fig. 2.1D).

We next subjected URfl/fl;Lck-Cre mice to comprehensive analyses of T cell development and maturation in the thymus. There was no difference in thymic cellularity between URfl/fl; Lck-Cre mice and wild type (WT) control littermates (Fig. 2.2A). Early thymocyte development [as marked by transitions through the double negative 1 (DN1) to DN4 stages] was also intact in the absence of UVRAG (Fig. 2.2B). Although the maturation of DN thymocytes to single positive (SP) CD4+ and CD8+ T cells appeared to be modestly reduced in URfl/fl;Lck-Cre mice compared to WT mice (Fig. 2.2C), this difference failed to reach statistical significance. Similarly, the expression levels of numerous developmental markers, including TCRβ, CD44, CD62L, CD69 and CD25, were equivalent on T lineage cells from URfl/fl; Lck-Cre and WT littermates (Fig. 2.2D).

Thus, UVRAG appears to be dispensable for thymic T cell development and maturation.

SPL/LN&

A& B& & 10 Thy CD4+& CD8+& 8 98) Non"speciﬁc) 6 UVRAG& 4

2 50) β"ac,n) 0

UVRAG mRNA (relative to 18S) to mRNA(relative UVRAG DN DP ! C& SP THY CD4 SPL/LNCD8 SPL/LN

UV-RAG Targeting Scheme 12.5 kb

8.65 kb 1. WT B B H H

LA II SA

B B 2. TV frt neo DTA loxp II 10 kb B

3. Targeted locus 6.3 kb B H B frt B H B ! neo II 4. Locus after Flpe- frt recombination 8.65 kb

6.3 kb B B H B H B

5. Locus after Cre-loxp recombination 7.65 kb

6.3 kb B H H D& UR&fl/fl;&& URfl/fl& Lck-Cre& 98) Non"specific) UVRAG& 50) β"tubulin)

Figure 2.1. Generation and validation of conditional Uvrag-deficient mice

(A) Quantitation of relative UVRAG mRNA expression in WT T cell subsets. The indicated thymocyte and peripheral T cell subsets were sorted from WT mice and their expression of UVRAG mRNA was determined by qPCR. DN, double negative (CD4- CD8-); DP, double positive (CD4+CD8+); SP Thy, single positive CD4+ or CD8+ thymocytes; CD4+ SPL/LN or CD8+ SPL/LN, mature CD4+ or CD8+ T cells from spleen (SPL) or lymph nodes (LN), respectively. Data are expressed relative to 18S RNA and are the mean ± SEM of triplicates. Results are representative of a single experiment involving 2 WT mice. (B) Immunoblot to detect UVRAG protein in lysates of thymocytes and purified CD4+ and CD8+ T peripheral cells from a WT mouse. Upper band, non-specific protein. β-actin, loading control. Results are representative of a single WT mice examined. (C) Schematic diagram of conditional targeting of the mouse Uvrag gene. 1. WT murine genomic Uvrag locus indicating the long arm (LA) and short arm (SA) of homology as well as UVRAG exon II, which is essential for UVRAG function. Probes used to verify targeting events are marked by short horizontal lines, and the expected sizes of restriction fragments are shown. H, HindIII; B, BamHI. 2. Targeting vector (TV) design. LoxP sites (triangles) were inserted to flank UVRAG exon II and the neomycin resistance gene (neo), which was used as a selection marker during ES cell culture. DTA, Diptheria toxin A. 3. Targeted Uvrag allele. 4. Locus after Flpe-frt- mediated removal of neo. 5. Locus after Cre recombinase-mediated removal of exon II. (D) Immunoblot confirming loss of UVRAG protein in peripheral T cells purified from URfl/fl; Lck-Cre mice. URfl/fl, littermate WT control. β-tubulin, loading control. Results are representative of at least 3 trials involving 2-4 mice per genotype.

A& 1000 UR fl/fl URfl/fl; Lck-Cre ) 6

100 Cells (x10 Cells

10 THYMUS

UR&fl/fl& UR&fl/fl;& B& 50 UR fl/fl Lck-Cre& fl/fl 40 UR ; Lck-Cre

% population % 10

CD44) DN1 DN2 DN3 DN4 CD25) C& UR&fl/fl& UR&fl/fl;& 100 UR fl/fl fl/fl Lck-Cre& ) UR ; Lck-Cre 6

10 # cells (10

CD8) DN DP CD4) CD8 SP CD4 SP

DN& DP& CD8+& CD4+& D& TCR)β)

CD44)

URfl/fl CD62L) URfl/fl; Lck-Cre

CD69)

CD25)

Fluorescence)

Figure 2.2. Loss of UVRAG does not impair T cell development or maturation.

(A) Quantitation of absolute cellularity of the thymus (THY) in control and URflfl; Lck- Cre mice. Each data point represents an individual mouse. Horizontal lines are the cumulative geometric mean ± SEM of 13 independent experiments involving 1-4 mice per genotype. (B) Left: Flow cytometric analysis of DN thymocyte stages in thymus of littermate control and URflfl; Lck-Cre mice. Numbers in quadrants are percentages of lineage negative (CD4-CD8-CD3-) cells that were DN1 (CD44+CD25-), DN2 (CD44+CD25+), DN3 (CD44-CD25+) or DN4 (CD44-CD25-) cells. Results are representative of 3 mice/group. Right: Quantitation of mean percentages ± SEM of the indicated DN thymocyte subsets from 3 WT control and 3 URfl/fl; Lck-Cre mice. (C) Left: Flow cytometric analysis of DN, DP and SP thymocytes from littermate control and URfl/fl; Lck-Cre mice. Numbers are percentages of total live thymocytes and are representative of 4 mice/group. Right: Quantitation of the mean absolute numbers ± SEM

of the indicated thymocyte subsets in the thymus of WT and URflfl; Lck-Cre mice. Results are derived from 13 independent experiments involving 1-4 mice per genotype. (D) Flow cytometric profiling of the indicated markers in the indicated thymocyte subsets from WT and URfl/fl; Lck-Cre mice. Results are representative of at least 3 independent experiments involving 1-4 mice/group.

2.4.3 URfl/fl; Lck-Cre mice exhibit peripheral T cell lymphopenia

To determine whether UVRAG deficiency in T cells affects peripheral T cell maintenance, we compared the secondary lymphoid organs of URfl/fl; Lck-Cre mice with those of their WT littermates. Cellularity was significantly reduced in spleen and lymph nodes (LN) of the mutants (Fig. 2.3A). To investigate the effect of UVRAG loss specifically on peripheral T cell homeostasis, we measured the proportions and cellularity of various T cell subsets in spleen, LN and peripheral blood (PBL). In contrast to the thymus, the secondary lymphoid organs and blood showed dramatic reductions in proportions and numbers of CD4+ and CD8+ T cells (Fig. 2.3B-D). In the mutant spleen, the total CD4+ T cell number (4.69x106 cells) was less than 50% of that in WT spleen

(10.9x106 cells; p value<0.0003), and this difference was even more prominent for CD8+

T cells (WT, 5.83x106 versus URfl/fl; Lck-Cre, 1.76x106; p value <0.00007) (Fig. 2.3B-

D). As a result, the CD4:CD8 ratio was altered in URfl/fl; Lck-Cre mice. The absence of

UVRAG caused a similar decline in CD4+ and CD8+ T cells in the LN and PBL (Fig.

2.3B-D). This general reduction in peripheral CD4+ and CD8+ T cells was maintained in aged URfl/fl;Lck-Cre mice (data not shown), indicating that the deficit was not restricted to the initial maturation of the immune system in young mice. Interestingly, the homeostasis of CD4+CD25+Foxp3+ regulatory T cells was not affected in any lymphoid tissue by loss of UVRAG (Fig. 2.3E). These data suggest that UVRAG is largely dispensable for thymocyte differentiation but may play a critical role in regulating peripheral T cell homeostasis.

A& B& SPL& LN& PBL&

1000 UR fl/fl **** fl/fl ﬂ/ﬂ& UR ; Lck-Cre UR&

) * 6 100 UR&ﬂ/ﬂ&;& Cells (x10 Cells 10 Lck-Cre&

LN SPL CD8) CD4) C& Propor,on) D& Cellularity) 100 50 fl/fl URfl/fl; Lck-Cre UR fl/fl 40 fl/fl ) UR ; Lck-Cre UR 6 *** * 30 * ** **** 10 * 20

Cells (%) Cells * Cells (x10 Cells 10

0 1

CD4 CD8 CD4 CD8 CD4+ CD8+ CD4+ CD8 + SPL LN SPL LN

E& THY& SPL& LN&

UR&ﬂ/ﬂ&

UR&ﬂ/ﬂ&;& Lck-Cre& FOXP3)

CD25)

Figure 2.3. Impaired T cell homeostasis in the periphery of URfl/fl; Lck-Cre mice.

(A) Quantitation of absolute cellularity of spleens and LN in control and URflfl; Lck-Cre mice (n=13-16 mice/group). Each data point represents an individual mouse and horizontal bars are cumulative geometric mean values. Results are derived from 13 trials. ****p<0.00005. *p<0.05. (B) Flow cytometric analysis of total CD4+ and CD8+ T cells isolated from spleen, LN and peripheral blood (PBL) of littermate control and URfl/fl; Lck-Cre mice (n=13-16/group). Numbers are percentages of total live lymphocytes. Results are representative of 13 trials. (C, D) Quantitation of percentages (C) and absolute numbers (D) of CD4+ and CD8+ T cells in the spleen and LN of WT and URflfl; Lck-Cre mice (n=13-16 mice/group). Results are the mean ± SEM. (E) Flow cytometric analysis of CD25+FoxP3+ regulatory T cells (Tregs) in thymus, spleen and LN of WT and URfl/fl; Lck-Cre mice. Numbers are percentages of total CD4+ T cells. Results are representative of 3 independent experiments.

2.4.4 UVRAG is essential for maintaining NKT cell numbers

Although the deletion of UVRAG is specific to T cells in our URfl/fl; Lck-Cre mice, it is well known that perturbations in one cell type can have dramatic effects on other populations [208]. We therefore examined the development and homeostasis of other leukocyte populations in our mutant mice. As expected, the significant reduction in T cell numbers in the peripheral compartments of URfl/fl; Lck-Cre mice led to increased proportions of B (B220+) and NK (NK1.1+CD3-) cells in LN and spleen (Fig. 2.4A, B).

However, total numbers of B and NK cells were not altered by T cell-specific UVRAG deficiency. Similarly, the proportions of B cells within the splenic marginal zone (MZ) and follicular zone (FO) were normal in our mutants (Fig. 2.4C). Dendritic cells

(CD11c+) and macrophages (CD11b+ and/or GR-1+) were also present in normal proportions in spleen and bone marrow of URfl/fl; Lck-Cre mice (data not shown). In stark contrast, the proportion of iNKT cells (TCRβ+CD1d tetramer+) was greatly reduced in mutant thymus, spleen and liver (Fig. 2.4D). This result was not unexpected since T cells and iNKT cells are derived from the same thymic progenitor cells and the Lck is expressed in the developing and mature iNKT cell population. Thus, UVRAG is important for the peripheral maintenance of T cells and development and peripheral maintenance of iNKT cells.

A& B&cells& NK&cells& SPL& LN& SPL& LN&

URﬂ/ﬂ&

URfl/fl;& Lck-Cre& NK1.1) B220) CD3) CD3) B& C& URfl/fl& URfl/fl;&Lck-Cre& 100 UR fl/fl fl/fl ) UR ; Lck-Cre 6 10 FO)

1 Cells (x 10 (x Cells

0.1 MZ) B220+ NK1.1+ B220+ NK1.1+

SPL LN CD23) CD21) D& THY& SPL& LN& LIV&

UR&ﬂ/ﬂ&

UR&ﬂ/ﬂ&;& Lck-Cre& CD1d/PBS"47)Tetramer)

TCRB)

Figure 2.4. Impaired homeostasis of iNKT cells in URfl/fl; Lck-Cre mice.

A) Flow cytometric analysis of B cell (B220+) and NK cell (NK1.1+) populations from spleen and LN of littermate control and URfl/fl; Lck-Cre mice (n=6-10 mice/group). Numbers are percentages of total live lymphocytes. Results are representative of 6 independent trials. (B) Quantitation of absolute numbers of B (B220+) and NK (NK1.1+) cells inspleen and LN of WT and URfl/fl; Lck-Cre mice. Results are mean ± SEM (n=1-4 mice/group) and are representative of 10 and 6 independent experiments (B and NK cells, respectively). (C) Flow cytometric analysis of B cell subsets in spleens of WT and URfl/fl; Lck-Cre mice (n=1-2 mice/group). Numbers are percentages of B220+-gated cells. FO, follicular cells (CD23+CD21+); MZ, marginal zone cells (CD23-CD21+). Results are representative of 3 independent trials. (D) Flow cytometric analysis of iNKT cell populations (circled) in thymus, spleen, LN and liver (LIV) of littermate control and URfl/fl; Lck-Cre mice (n=1-2 mice/group). Numbers are percentages of total live lymphocytes. Data are representative of 3 independent experiments.

2.4.5 UVRAG function in T cells is cell-intrinsic

To formally test whether the UVRAG deficiency altering the peripheral maintenance of T and NKT cells was cell-intrinsic, we generated bone marrow (BM) chimeras in Rag-2 knockout mice, which lack both B and T cells. To assess whether UVRAG was essential for the reconstitution of the T cell compartment in these chimeras, we compared the relative contribution of CD45.1 vs CD45.2 donor BM cells to reconstituted lymphoid organs at 1.5-4 months post-reconstitution. A ratio of 1:1 indicates that both donor cell populations have equal potential to repopulate the irradiated host, whereas a skewing of the ratio indicates an advantage for one donor. In control animals, where the CD45.1 and

CD45.2 donor cell populations both expressed UVRAG, B and T cells were found to originate equally from CD45.1 and CD45.2 donor cells (data not shown). Similarly, in experimental chimeras where the CD45.2 donor cells were derived from the BM of

URfl/fl; Lck-Cre mice, the B cell compartment showed a 1:1 ratio (Fig. 2.5A), consistent with our evaluations of intact URfl/fl;Lck-Cre mice. In contrast, T cells in these experimental chimeras showed a striking bias toward derivation from WT donor BM cells, resulting in a ratio of 10:1 in spleen and LN, and 6:1 in PBL (Fig. 2.5A). We observed an even greater skewing towards the WT donor population when the chimeric mice were examined at 4 months of age (Fig. 2.5B). Similar results were observed in lethally irradiated C57BL/6 reconstitution system. Interestingly, the competitive disadvantage faced by the mutant T cells differed between the CD4+ and CD8+ subsets, with CD8+ T cells showing a greater dependence on UVRAG (Fig. 2.5C). All of these findings are consistent with the phenotype of our URfl/fl; Lck-Cre mice at steady-state, and suggest that UVRAG plays a crucial and cell-autonomous role in maintaining

peripheral T cell numbers. Moreover, the differential expression of UVRAG in CD4+ and

CD8+ T cells (Fig. 2.1B), coupled with the altered CD4+ and CD8+ T cell numbers in the secondary lymphoid organs of both URfl/fl; Lck-Cre mice and our BM chimeras, indicate that UVRAG may play different roles in different T cell subsets, at least in mice.

When we examined thymic reconstitution by UVRAG-deficient CD45.2 T cells, we observed a 2-4 fold reduction in thymocytes derived from mutant BM compared to thymocytes derived from WT BM (Fig. 2.5D). This result contrasted with our analysis of

URfl/fl; Lck-Cre mice at steady-state, in which T cell development appeared to be normal.

However, BM competition experiments can sometimes bring to light subtle differences between two cell populations, and as such are ideal for teasing apart mild phenotypes.

Our thymic reconstitution data suggest that UVRAG does play a role in T cell development but that the effect of its loss is only observed when UVRAG-deficient thymocytes are in competition with WT thymocytes. These results implicate UVRAG in sustaining T cell survival when resources are limited.

THY& SPL& LN& PBL&

B220+&

CD3+& CD45.1# CD45.2#

B& *

25 WT:WT 20 WT:KO

15 ** 10

CD45.1: CD45.2 CD45.1: 5

B220 CD3 B220 CD3 1.5 months 4 months

C& CD4&+& CD8&+&

SPL&

LN& CD45.1) CD45.2)

D& THY&

DN& DP& CD4&SP& CD8&SP& CD45.1) CD45.2)

DN1& DN2& DN3& DN4& CD45.1) CD45.2)

Figure 2.5. UVRAG-deficient T cells show a defect in bone marrow reconstitution.

(A) Flow cytometric analysis of B and T cell populations in secondary lymphoid organs of BM chimeras. BM from CD45.1+ WT and CD45.2+ URfl/fl; Lck-Cre mice was mixed 1:1 and intravenously (i.v.) injected into sub-lethally irradiated Rag-2-/- mice (n=3-4 mice/group). After 2 months, lymphoid organ reconstitution was assessed by measuring the relative contribution of WT (CD45.1+) and UVRAG-deficient (CD45.2+) BM to generating B (B220+) and T (CD3+) cell populations in thymus, spleen, LN and PBL. Numbers are percentages of live B220+ or CD3+ cells. Results are representative of 3 trials. (B) Quantitation of ratio of WT (CD45.1+) vs. UVRAG-deficient (CD45.2+) cells contributing to the reconstitution of B and T cell populations in PBL of the sub-lethally irradiated Rag-2-/- mice in (A) at 1.5 months or 4 months after reconstitution. Each data point represents a single mouse and horizontal bars are geometric mean values. WT:WT, mice injected with a control BM mixture. Data are the cumulative mean ± SEM of 2 independent experiments involving 3-4 mice per genotype. (C) Flow cytometric analysis of reconstitution of CD4+ and CD8+ populations in spleen and LN of the mice in (A). Numbers are percentages of live CD4+ and CD8+ populations in spleen and LN. Results are representative of a single experiment involving 3 mice per genotype. (D) Flow cytometric analysis of reconstitution of the indicated thymocyte subsets in the thymus of the mice in (A). Numbers are percentages of live, lineage negative DN cells. Results are representative of 1 experiment involving 3 mice.

2.4.6 UVRAG deletion leads to an enhanced memory marker profile on T cells

The residual T cells in mice harbouring lymphopenia, or reduced peripheral T cell numbers, often express a marker profile characteristic of activated or central memory T cells [209, 210]. Populations of naïve, activated and memory T cells can easily be distinguished on the basis of CD62L and CD44 expression, in that naïve T cells are

CD62LhiCD44lo, central memory T cells are CD62LhiCD44hi, and effector T cells are

CD62LloCD44hi. We found that, consistent with their lymphopenic phenotype, URfl/fl;

Lck-Cre mice possessed T cells in spleen and LN that exhibited increased expression of

CD44 (compared to WT T cells) with no concomitant decrease in CD62L (Fig. 2.6A-C).

These data indicate that a loss of UVRAG causes a reduction in naïve T cells and a preferential retention of central memory T cells. Furthermore, the upregulation of CD44 on the residual T cells in URfl/fl; Lck-Cre mice was not accompanied by an increase in other T cell activation markers such as CD25 and CD69, indicating that these UVRAG- deficient T cells were not activated. These findings support a role for UVRAG in maintaining homeostatic T cell proliferation.

To determine whether the increased memory T cell population in steady-state URfl/fl;

Lck-Cre mice was due to their lymphopenic environment or to an intrinsic role of

UVRAG, we compared CD44 and CD62L expression by T cells in WT and URfl/fl; Lck-

Cre mice in the non-lymphopenic setting of mixed BM chimeras. Indeed, the effector

(CD44+CD62L-) and memory (CD44+CD62L+) populations derived from mutant

(CD45.2) donor cells were comparable to those derived from WT CD45.1 donor cells

(Fig. 2.6D). Thus, the increase in memory cell markers on UVRAG-deficient T cells is due to the lymphopenic environment and not to loss of an intrinsic function of UVRAG.

+& A& CD8+& CD4 SPL& LN& SPL& LN& UR&ﬂ/ﬂ&

UR&ﬂ/ﬂ&;& Lck-Cre& CD44) B& CD62L) 100 UR fl/fl URfl/fl; Lck-Cre ** 80 ** ** ** 60 ** * 40 ** * % population % 20 * * 0

Naive Naive Naive Naive MemoryEffector MemoryEffector MemoryEffector MemoryEffector Spl CD8+ Spl CD4+ LN CD8+ LN CD4+

C& CD8+& CD4+& SPL&

UR&fl/fl&& & LN UR&fl/fl;& Lck-Cre& Cell)Count)

CD69) CD25) CD69) CD25)

D& CD8+& CD4+& SPL& LN& SPL& LN&

CD45.1+&

CD45.2+& CD44) CD62L)

Figure 2.6. Altered marker profile of residual UVRAG-deficient T cells.

(A) Flow cytometric analysis of T cells from spleen and LN of littermate control and URfl/fl; Lck-Cre mice (n=1-4/group) that were immunostained in vitro to detect the activation/memory markers CD44 and CD62L. Numbers are percentages of total CD8+ T cells (left) or total CD4+ T cells (right) and are representative of 8 trials. (B) Quantitation of flow cytometric data for the mice in (A) distinguishing among naïve (CD62L+CD44-), memory (CD62L+CD44+), and effector (CD62L-CD44+) T cells in spleen and LN. Results are the mean ± SEM and are representative of 8 independent experiments involving 1-4 mice/genotype. **p<0.005; *p<0.05. (C) Flow cytometric plot of CD69 and CD25 expression by T cells from spleen or LN of littermate control and URfl/fl; Lck- Cre mice. Data are representative of 6 independent experiments involving 1-4 mice/genotype. (D) Flow cytometric analysis of expression of the activation/memory markers CD44 and CD62L by T cells from spleen and LN of mixed BM chimeras containing WT (CD45.1+) and UVRAG-deficient (CD45.2+) donor cells (n=3 mice). Data were analyzed as in (A). Numbers are percentages of total CD8+ or total CD4+ T cells and are representative of 1 trial.

2.4.7 UVRAG deficiency does not render T cells more sensitive to cell death induction

The lymphopenia observed in ATG5-, ATG7- and Beclin-1deficient animals has been largely explained by an increase in apoptosis in the periphery of resting mice [61, 63,

131, 139]. Having confirmed a cell-intrinsic role for UVRAG in maintaining peripheral T cell numbers (if not their marker profile), we explored the potential role of apoptosis in mediating this phenotype. We first compared base levels of apoptosis in various T cell subsets isolated from WT and URfl/fl; Lck-Cre littermate mice. In contrast to the results obtained using animals deficient in other autophagy-related genes, apoptosis was not significantly enhanced in cultures of T cells isolated from steady-state URfl/fl;Lck-Cre mice (Fig. 2.7A). We then investigated whether loss of UVRAG sensitizes T cells to apoptotic stimuli. WT and mutant peripheral T cells and thymocytes were subjected in vitro to a panel of apoptotic stimuli that included anti-CD3 stimulation, γ-irradiation (IR),

UV irradiation, and anti-Fas ligand (Fas L) treatment. Surprisingly, no significant differences in apoptotic sensitivity were detected between WT and mutant T cells or thymocytes (Fig. 2.7B, C). These data suggest that the reduced T cell numbers in URfl/fl;

Lck-Cre mice are not due to increased apoptosis. Importantly, these findings differentiate the UVRAG-deficient phenotype from that of other autophagy-related genes [61, 63, 131,

139], and imply an autophagy-independent role for UVRAG in T cells.

A& CD4+& CD8+& B&

60 SPL& URfl/fl; Lck-Cre CD4

cells fl/fl

+ UR CD4 40 URfl/fl; Lck-Cre CD8 UR fl/flCD8 20

LN& /PI V Annexin % 0 0 1 2 3 γ Irradiation (Gy)

Annexin)V+) C& 100 55 URfl/fl; Lck-Cre URfl/fl; Lck-Cre 50 UR fl/fl UR fl/fl 80 45 60 40 40 35 30 20 25 0 % Annexin V /7AAD+ cells /7AAD+ V Annexin % % Annexin V /7AAD+ cells /7AAD+ V Annexin % 0 1 0 2 4 10 50 0.01 IR (Gy) anti-CD3 (µg/ml) 100 100 URfl/fl; Lck-Cre URfl/fl; Lck-Cre 80 fl/fl 80 UR fl/fl UR 60 60

40 40

20 20

0 0 % Annexin V/7AAD + cells + V/7AAD Annexin % % Annexin V 7AAD+ cells 7AAD+ V Annexin % 0 0 1 40 80 0.1 0.01 2) UV (mJ/cm JO-2 (µg/ml)

Figure 2.7. Normal T cell apoptosis in secondary lymphoid organs of URfl/fl;Lck-Cre mice.

(A) Flow cytometric analysis of resting CD4+ or CD8+ T cells that were isolated from spleen or LN of WT (blue line) or URfl/fl; Lck-Cre (red line) mice (n=1-4/group) and stained with Annexin V to determine apoptosis. Data are representative of at least three independent experiments. (B) Quantitation of apoptosis in cultures of CD4+ or CD8+ T cells that were isolated from spleen of WT or URfl/fl; Lck-Cre mice (n=1/group), treated with indicated amounts of IR, and cultured overnight. Apoptosis was measured by Annexin V/PI staining and flow cytometry. Data were obtained from a single experiment involving 1 WT and 1 mutant mouse. No significant differences in DNA damage-induced apoptosis were observed. (C) Quantitation of apoptosis in thymocytes that were isolated from WT or URfl/fl; Lck-Cre mice (n=3/group) and treated with indicated levels of plate- coated anti-CD3, UV radiation, IR (Gy), or anti-Fas L antibody (JO-2). Cells were cultured overnight after treatment and apoptosis was measured by Annexin V/7AAD staining and flow cytometry as above. Data shown are the mean ± SEM (n=3) and are representative of 2 trials.

2.4.8 UVRAG does not regulate autophagy in T cells

The well-described role of UVRAG in autophagy in vitro prompted us to determine if the phenotypes observed in our URfl/fl; Lck-Cre mice were due to impaired autophagy. We first examined the mitochondrial content of UVRAG-deficient T cells. T cells are known to decrease their mitochondrial mass upon thymic exit [61], a finding we confirmed (Fig.

2.8A). This reduction is thought to be a mechanism that allows newly produced T cells to cope with the increased oxygen tension present in the blood compared to the thymus [61].

Exiting T cells that are unable to reduce their mitochondria suffer an increase in mitochondrial ROS that leads to cell death [61]. Mitochondrial downregulation is at least partly attributable to mitophagy (autophagic destruction of mitochondria). T cells in mice lacking ATG3, ATG5 or ATG7 are unable to reduce their mitochondria sufficiently, triggering ROS-induced death [61-63]. To examine mitochondrial content in T cells of our URfl/fl; Lck-Cre mice, we stained thymic, splenic and LN populations of T cells with

Mitotracker green dye and measured mitochondrial mass by flow cytometry. As shown in

Figure 2.8B, there were no appreciable differences between WT and mutant mice in T cell mitochondrial content in any organs examined, suggesting that UVRAG-deficient T cells can reduce their mitochondria normally upon thymic exit. Although these findings imply that UVRAG is not required for mitophagy, it is still possible that UVRAG may be required for other types of autophagy in T cells.

T cell activation is known to induce robust autophagosome formation [131, 134].

To measure autophagic activity in UVRAG-deficient T cells, we turned to the LC3 cleavage assay in which LC3-I is converted to PE-conjugated LC3-II upon induction of autophagy. PE-labelled LC3-II is then incorporated into the autophagosomal membrane

[31-33, 40, 41]. Thus, to measure autophagosomal flux, cells are treated with or without an autophagy inhibitor and accumulation of LC3-II is monitored. We cultured peripheral

T cells from WT and URfl/fl; Lck-Cre mice in complete medium alone, or in medium containing anti-CD3 plus anti-CD28 antibodies, with or without the autophagy inhibitor chloroquine. We found that anti-CD3/CD28 treatment increased LC3-I and LC3-II levels in WT T cells, consistent with published data [134], and that this elevation was even greater in UVRAG-deficient T cells (Fig. 2.8C, D). Moreover, whereas only a faint LC3-

II band was present in WT T cells treated with chloroquine (as expected), large amounts of LC3-II were present in mutant T cells (Fig. 2.8C, D). In addition, p62 levels were generally lower in UVRAG-deficient T cells (Fig. 2.8C, D), another reflection of higher autophagic activity in these cells. Collectively, these results suggest that, in contrast to other autophagy related genes, UVRAG may not be essential for autophagy in T cells.

Thus, UVRAG likely regulates peripheral T cell numbers by an autophagy-independent mechanism.

A& B& THY SPL LN THY WT SPL KO CD8+ LN KO Cell)Count) Mitotracker) Mitotracker)

UR&fl/fl;& C& UR&fl/fl& Lck-Cre& 2)CD3/CD28)+)CQ) Medium) Medium) 2)CD3/CD28)+)CQ) 2)CD3/CD28) 2)CD3/CD28) LC3"I) LC3"II) UVRAG)

p62)

β"ac,n) D& 3) 3) 2.5) p62) 2.5) LC3"I) 2.5) LC3"II) 2) 2) 2) 1.5) 1.5) 1.5) 1) 1) 1)

Normalized&p62&levels& 0.5) Normalizd&LC3-II&levels& 0.5) Normalized&LC3-I&levels& 0.5)

0) 0) 0) Media) CD3/28)CD3/28)+) Media) CD3/28) CD3/28)+) Media) CD3/28) CD3/28)+) CQ) CQ) CQ)

Figure 2.8. UVRAG-deficient T cells undergo normal autophagy.

(A) Flow cytometric plot of mitochondrial content changes in WT T cells upon migration to the periphery. T cells from the thymus, spleen and LN of resting WT mice (n=1-3) were stained with FITC-Mitotracker to measure mitochondrial content. Plot shows results gated on CD8+ SP T cells and is representative of 3 trials. (B) Flow cytometric plot of mitochondrial content in CD8+ SP T cells from the indicated lymphoid organs of one WT and two URfl/fl; Lck-Cre mice. Results were obtained as in (A) and are representative of two independent experiments involving 1-2 mice/genotype. (C) Autophagic flux in activated T cells as measured by LC3I-II cleavage. Purified T cells from spleen and LN of WT and URfl/fl; Lck-Cre mice (n=2/group) were pooled and activated in vitro overnight with 2µg/ml anti-CD3 plus 0.2 µg/ml anti-CD28 antibodies, with or without 25nM chloroquine (CQ). Lysates were immunoblotted to detect LC3I/II, UVRAG and p62. β-actin, loading control. Results shown are representative of two independent experiments. (D) Quantitation of the immunoblot results from (C). Band densities were determined by Odyssey Licor and normalized to the β-actin levels in medium-treated WT samples. Data are from 2 mouse/genotype.

2.4.9 UVRAG is required for homeostatic expansion of naïve T cells in lymphopenic host

We speculated that UVRAG might be essential for the homeostatic survival and proliferation of naïve T cells needed to maintain normal peripheral T cell numbers. To formally test this hypothesis, we compared the expansion of URfl/fl; Lck-Cre and WT naïve T cells transplanted into a lymphopenic host. We purified naïve T cells from WT

(CD45.1) and URfl/fl; Lck-Cre (CD45.2) mice, labeled them separately with carboxyfluorescein succinimidyl ester (CFSE), and adoptively transferred a mixture of these cells into lethally irradiated C57BL/6 host (CD45.1/2) mice. An equal ratio of donor cells was confirmed by flow cytometry before injection (data not shown). Six days later, donor T cells were recovered from spleen and LN, and cell numbers and marker profiles were assessed by flow cytometry. We observed a bias towards WT (CD45.1) T cells in the spleen and LN in recipients that had received mixed donor cell populations

(Fig. 2.9A). Moreover, UVRAG-deficient T cells showed decreased CFSE dilution

(proliferation) compared with WT T cells (Fig. 2.9B). These data strongly suggest that

UVRAG plays an essential role in maintaining peripheral T cell numbers, and that the lymphopenia observed in URfl/fl;Lck-Cre mice may be due to impaired T cell survival and homeostatic proliferation.

To rule out the possibility that altered trafficking or homing to secondary lymphoid organs was responsible for the reduced peripheral T cell numbers in the mutant, we measured the recovery of WT and URfl/fl; Lck-Cre T cells at an earlier time point following adoptive transfer. At 4 days post-transfer, there was no difference between the

WT and mutant in the number of T cells in PBL (Fig. 2.9C). Moreover, our earlier data

demonstrated a deficit in T cells in all secondary lymphoid organs of steady-state URfl/fl;

Lck-Cre mice and mixed BM chimeras, arguing against a trafficking defect. Thus, while

UVRAG does not appear to be required for homing and trafficking of T cells to secondary lymphoid organs, it may be an important regulator of naïve T cell homeostasis in vivo.

A& Day&6:&CD3+&& C& Day&4:&CD3+&

Spleen& LN& PBL& CD45.1) CD45.1) CD45.2) CD45.2) B& Spleen& LN&

URfl/fl&CD45.1+& URfl/fl;Lck-Cre& CD45.2+& Count) CFSE)

Figure 2.9. UVRAG-deficient T cells show defects in lymphopenia-induced expansion in vivo.

(A) Flow cytometric analysis of recovery of UVRAG-deficient T cells from spleen and LN of a lymphopenic host. CFSE-labelled naïve WT (CD45.1+) or UVRAG-deficient (CD45.2+) T cells were i.v. transferred into irradiated C57BL/6 mice (CD45.1/CD45.2) and the indicated T cell populations were recovered and analyzed 6 days later. Numbers are percentages of live CD3+ T cells that were CD45.1+ WT or CD45.2+ UVRAG- deficient cells. Results are representative of 2 trials involving 3-4 mice/group. (B) Flow cytometric plot of the proliferation of T cells in a lymphopenic host. Histogram depicts

CFSE dilution in WT (CD45.1+) and UVRAG-deficient (CD45.2+) T cells among live CD3+ cells in spleen (left panel) and LN (right panel) of one mouse from (A). Data are representative of 2 independent experiments involving 3-4 mice/group. (C) The experiment in (A) was repeated and the mice were analyzed by flow cytometry to detect T cell reconstitution in PBL on day 4 after transfer. Data were analyzed as in (A) and are representative of 1 trials involving 3 mice/group.

2.4.10 UVRAG is a negative regulator of TCR-mediated T cell proliferation

The survival of naïve T cells depends in large part on sp-MHC interactions with

TCRs. To test the role of UVRAG in TCR signalling, we evaluated the proliferation of naïve T cells in response to TCR engagement in vitro. T cells from mice deficient in other autophagy-related genes, including ATG3, ATG5, ATG7 and Beclin-1, exhibit impaired proliferation in response to anti-CD3/CD28 stimulation in vitro [61, 63, 131, 139].

Strikingly, we found that naïve URfl/fl; Lck-Cre T cells stimulated with anti-CD3/CD28 underwent hyper-proliferation compared to stimulated WT cells (Fig. 2.10). This hyper- proliferation was observed for both CD4+ and CD8+ T cells when measured by either 3H- thymidine incorporation or CFSE dilution, and at all anti-CD3 antibody concentrations tested (Fig. 2.10B). Notably, naive T cells from URfl/fl; Lck-Cre mice also blasted at lower concentrations of anti-CD3 antibody than did naive WT T cells, indicating that the mutant T cells had a lower activation threshold than WT T cells. Proliferation in response to stimulation with a combination of phorbol 12-myristate 13-acetate (PMA) and ionomycin, which activates T cells but bypasses TCR signalling, did not differ between

WT and mutant T cells (Fig. 2.10A). Thus, in contrast to other autophagy genes, UVRAG may be a negative regulator of T cell proliferation induced by TCR engagement.

40000 UR fl/fl *** ** URfl/fl; Lck-Cre 30000 ***

20000

10000 *** ** *** 0 0 0.1 1.0 10 0.01 3H- Thymidine Incorporation (CPM)

1.0 + CD28 PMA + Iono Anti-CD3 [µg/ml] B& CD8+&

URfl/fl;& URfl/fl;& UR&fl/fl& Lck-Cre& UR&fl/fl& Lck-Cre& αCD3[μg/ml]) 1/CD28& 5& 1& 0.1& 0.01& 0& FSC& CFSE& 13)

Figure 2.10. Loss of UVRAG leads to T cell hyper-proliferation.

(A) 3H-thymidine incorporation assay of proliferation in vitro of naïve CD4+ T cells that were magnetically isolated from the periphery of WT or URfl/fl; Lck-Cre mice (n=2- 4/group) and stimulated for 72 hr in vitro with the indicated concentrations of plate- bound anti-CD3 plus anti-CD28 antibodies, or with PMA (10ng/ml) plus ionomycin (100ng/ml). Data are the mean ± SEM of triplicates from two independent experiments.

***p<0.0005. (B) Flow cytometric analysis of the proliferation of naïve CD8+ T cells that were labelled in vitro with CFSE (1µM) and treated with the indicated concentrations of plate-bound anti-CD3/CD28 antibodies for 72 hrs. Both FSC (left panel) and CFSE (right panel) profiles are shown. Results are representative of at least 3 independent trials involving 2-4 mice/group.

2.4.11 UVRAG is required for IL-7 induced survival of naïve T cells

Survival of naïve T cells also depends on signals from homeostatic cytokines, such as interleukin-7 (IL-7) and interleukin-15 (IL-15). These cytokines promote naïve T cell survival by inducing the expression of the anti-apoptotic protein Bcl-2. To determine whether UVRAG-deficient T cells could respond to IL-7, we cultured naïve T cells from

WT and URfl/fl; Lck-Cre mice for over 100 hours in the presence or absence of IL-7. In the absence of IL-7, both the WT and mutant T cells exhibited the expected pattern of progressive cell death, and this pattern was the same for CD4+ and CD8+ T cells (Fig.

2.11A). While the addition of IL-7 to the culture medium maintained the survival of WT

T cells at 89%, only 40% of URfl/fl; Lck-Cre T cells were viable even in the presence of

IL-7 (Fig. 2.11B). Naïve UVRAG-deficient T cells consistently exhibited a greater degree of cell death compared to WT T cells in the presence of IL-7, as determined by

7AAD positivity. These data strongly suggest that UVRAG may regulate peripheral T cell numbers by influencing the signalling of homeostatic cytokines.

100 100 URfl/fl; Lck-Cre URfl/fl; Lck-Cre fl/fl 80 UR fl/fl 80 UR

60 60

40 40

20 20

0 0 % of 7AAD- CD4+ T cells T CD4+ % 7AAD- of % of 7AAD-CD8+ T cells T % 7AAD-CD8+ of 0 0 24 48 72 96 24 48 72 96 112 112 B& Hours in culture Hours in culture

50 100 URfl/fl; Lck-Cre URfl/fl; Lck-Cre fl/fl fl/fl 80 UR 40 UR

60 30

40 20

20 10

0 0 % of 7AAD-CD4+ T cells T % 7AAD-CD4+ of % of 7AAD-CD8+ T cells T % 7AAD-CD8+ of 0 2 0 2 20 20 0.2 100 0.2 100 [IL-7 ng/ml] [IL-7 ng/ml]

Figure 2.11. UVRAG-deficient T cells are refractory to IL-7-mediated survival signalling.

(A) Quantitation of the survival of resting T cells over time in culture. Splenocytes from WT or URfl/fl; Lck-Cre mice were cultured in complete RPMI medium for >100 hrs and survival was assessed by 7AAD exclusion. Data are the mean percentage ± SEM (n=5- 7/group) of CD8+ (left) and CD4+ (right) T cells that were 7AAD-. Results are representative of 5 trials. No significant differences in the survival of resting T cells were observed. (B) Quantitation of the survival and expansion of IL-7-stimulated T cells over time in culture. Cells from the mice in (A) were cultured as in (A) with the addition of the

indicated concentrations of recombinant IL-7 (rIL-7). The survival and expansion of CD8+ (left) and CD4+ (right) T cells were measured by 7AAD exclusion as in (A). Data are the mean percentage ± SEM (n=2/group) and are representative of 2 independent experiments.

2.4.12 UVRAG does not regulate CD127 trafficking, STAT5α activation, or BCL-2 upregulation

To dissect the mechanism by which UVRAG promotes the survival of naïve T cells, we first compared expression levels of the IL-7 receptor α chain (IL-7Rα or CD127) on WT and URfl/fl; Lck-Cre T cell subsets. However, there were no appreciable differences in steady-state CD127 expression in any population examined (Fig. 2.12A). We next hypothesized that UVRAG might be regulating the trafficking of the IL-7 receptor, much as it controls the trafficking of the Notch receptor in Drosophila [149]. To study CD127 trafficking, we forced the internalization of this receptor by treating WT and mutant T cells with increasing concentrations of recombinant IL-7. Surface CD127 expression decreased to a similar extent on WT and UVRAG-deficient T cells (Fig. 2.12B), suggesting that UVRAG is not involved in CD127 trafficking.

IL-7 signaling in T cells is known to induce STAT5α phosphorylation via activation of Jak kinases 1 and 3. We therefore compared the ability of WT and mutant T cells to phosphorylate STAT5α in response to treatment with exogenous IL-7 in vitro. As shown in Figure 2.12C, phosphorylation of STAT5α in response to IL-7 was comparable in WT and mutant T cells, indicating that UVRAG maintains peripheral T cell homeostasis independently of STAT5α.

IL-7 signalling also upregulates expression of the anti-apoptotic molecule Bcl-2 in T cells. We compared expression levels of Bcl-2 in naïve WT and URfl/fl; Lck-Cre T cells cultured with or without IL-7 for 24 hours but found that a lack of UVRAG did not alter IL-7-induced Bcl-2 upregulation (Fig. 2.12D). Collectively, these data imply that

UVRAG is dispensable for IL-7-mediated STAT5α activation, or BCL-2 upregulation in

T cell homeostasis. The precise reason why UVRAG-deficient T cells are more susceptible to apoptosis than WT T cells in the presence of IL-7 remains under investigation.

A& B& + CD8)SP) DP) SPL CD8 IL-7[μg/ml]& CD8+) CD4+) WT&FMO&

KO&FMO&

DN) CD4)SP) WT&2& KO&2& CD127& WT&0.2&

CD8+) CD4+) KO&0.2& &WT&0&

&KO&0&

CD122& CD127& CD127&

UNT& IL-7& IL-15& Cell&Count&

phospho-Stat5α&

100

15)

D& URfl/fl;Lck- URfl/fl& Cre&

IL-7&[ng/ml]& 0& 100& 0& 100& kDa)

27) Bcl-2&

39) β"ac,n)

1.5 UR fl/fl URfl/fl; Lck-Cre

1.0

0.5

Bcl-2 protein expression protein Bcl-2 0.0

IL-7 IL-7 none none

Figure 2.12. UVRAG does not influence CD127 trafficking, Stat5α activation, or Bcl-2 expression.

(A) Flow cytometric analysis of steady-state CD127 expresson by the indicated thymocyte and mature peripheral CD4+ and CD8+ T cell subsets in WT (blue) and URfl/fl; Lck-Cre mice (red) mice (n=1-2/group). Data are representative of at least 3 independent experiments. No significant differences were seen. (B) Flow cytometric analysis of CD127 internalization in response to IL-7 treatment. CD4+ SP thymocytes from WT (blue shades) and URfl/fl; Lck-Cre (red shades) mice (n=1-2/group) were treated for 5 hr with the indicated concentrations of rIL-7 (ng/ml) and CD127 expression was monitored by flow cytometry. FMO, fluorescence minus one. Data are representative of 2

101

experiments. (C) Flow cytometric plot showing Stat5α phosphorylation (activation) in response to IL-7 or IL-15 treatment. Splenocytes from WT (blue) or URfl/fl; Lck-Cre (red) mice (n=1/group) were treated with 100µgml of IL-7 or IL-15 for 10 min, and phospho- STAT5α levels were detected by intracellular staining. UNT, untreated. Plots are gated on CD4+ T cells. Data are representative of at least 3 experiments. (D) Top: Immunoblot detecting Bcl-2 protein in total naïve T cells that were isolated from WT and URfl/fl; Lck- Cre mice (n=2/group) and cultured for 24 hr in the presence (or not) of 100 ng/ml rIL-7. β-actin, loading control. Bottom: Quantitation of band densities in the top panel determined by Odyssey Licor and normalized to β-actin. Data are expressed relative to the untreated WT value (set to 1) Results are representative of 1 trial.

102

2.4.13 UVRAG deficiency does not protect mice from developing EAE

We consistently observed a differential effect of UVRAG loss on CD4+ vs. CD8+ T cells in both URfl/fl; Lck-Cre mice at steady-state and in BM chimeras. The loss of UVRAG always reduced both T cell populations but the effect was less severe on CD4+ T cells. To study T cell subset-specific roles of UVRAG, we investigated T helper cell (Th) differentiation in vitro in the absence of UVRAG. To this end, we isolated naive

CD4+CD62L+ T cells from the periphery of WT and URfl/fl; Lck-Cre mice and primed these cells in vitro with anti-CD3 plus anti-CD28 antibodies for 3 days in the presence of specific cytokine cocktails to drive Th1, Th2, Treg or Th17 cell differentiation. After this polarization period, cytokine production was evaluated by intracellular cytokine staining and flow cytometry. Whereas the differentiation of Th2, Treg and Th17 cells was not altered in the absence of UVRAG, we observed almost twice as many Th1 cells in the mutant cultures (72%) as in WT cultures (41%) (Fig. 2.13A).

To test the in vivo relevance of this bias towards Th1 differentiation, we utilized the experimental autoimmune encephalomyelitis (EAE) disease model. EAE serves as a mouse model of human multiple sclerosis (MS), and both Th1 and Th17 cells are required for pathogenesis [115-118]. Mice are immunized with myelin oligodendrocyte glycoprotein (MOG) peptide plus pertussis toxin and monitored for 30 days to score disease onset and severity. WT mice typically develop severe EAE within 30 days of induction. We speculated that the T cell lymphopenia observed in URfl/fl; Lck-Cre mice might result in resistance to EAE induction due to a lower frequency of autoreactive cells.

Such is the case for the Beclin-1 knockout mice, which exhibit defects in Th1 and Th17

103

differentiation and are completely resistant to EAE induction [139]. On the other hand, the more robust Th1 cell differentiation we observed in our cultures of UVRAG-deficient

T cells might increase the frequency and severity of EAE developing in URfl/fl; Lck-Cre mice. However, we observed no difference in disease onset or severity between WT and

URfl/fl; Lck-Cre mice immunized with MOG peptide plus pertussis toxin (Fig. 2.13B).

This result suggests that the skewing towards Th1 cells in URfl/fl; Lck-Cre mice compensates for the marked overall lymphopenia of these mutants and is sufficient to drive active EAE.

To further test our hypothesis, we compared the proliferation capacities of MOG- specific T cells isolated from spleens of WT and URfl/fl; Lck-Cre mice at 14 days post-

EAE induction. WT and mutant splenocytes were treated in vitro with various concentrations of MOG peptide for 48 hours and proliferation was measured by 3H- thymidine uptake. The proliferation of UVRAG-deficient MOG-specific CD4+ T cells was similar to that of WT T cells (Fig. 2.13C), in line with our theory that that a more robust Th1 response was likely responsible for the mutant’s sensitivity to EAE induction despite its overall lymphopenia. These results support the notion that UVRAG, although essential for cytokine driven homeostatic proliferation, may be dispensable for TCR driven antigen-induced proliferation. These data are also consistent with our observation that UVRAG does not influence TCR signalling.

104

2012-05-23 EAE UVRAG.pzf:Data 1 - Mon Jul 09 13:59:02 2012

A& TH0) TH1) Treg) Th17)

UR&ﬂ/ﬂ&

UR&ﬂ/ﬂ;& Lck-Cre& IL"17) B& IFN"γ) 4 UVRAG fl/fl UVRAG -/- Lck-cre 3

2 EAE score EAE 1

0 0 5 10 15 20 25 30 C& Day

40000 UR fl/fl 30000 URfl/fl; Lck-Cre

20000

10000

0 0 1 10 0.1 100 3H-Thymidine Incorporation (CPM) MOG peptide [µg/ml]

Figure 2.13. UVRAG deficiency does not prevent mice from developing EAE.

105

Figure 2.13. UVRAG deficiency does not prevent mice from developing EAE.

(A) Flow cytometric analysis of in vitro Th differentiation by naïve CD4+ T cells that were purified from WT or URfl/fl; Lck-Cre mice (n=1/group) and cultured in cocktails of Th subset-specific cytokines in the presence of anti-CD3/CD28 plus IL-2. Th cells were restimulated in vitro with PMA/ionomycin for 6 hr, followed by intracellular staining to detect IL-17 and IFNγ. Numbers are percentages of live CD4+ T cells, and red circles indicate Th1 cells. Results are representative of 2 independent experiments. (B) Kaplan- Meier analysis of EAE scores of WT and URfl/fl; Lck-Cre mice (n=5/group) that were immunized subcutaneously (s.c.) with MOG peptide in complete Freund’s adjuvant, followed by i.p. injection of pertussis toxin (PT). Mice were scored daily for disease severity. Data are the cumulative mean ± SEM of two independent experiments each involving 5 mice/genotype. (C) 3H-thymidine incorporation assay of T cell proliferation in vitro of splenocytes isolated from WT and URfl/fl; Lck-Cre mice at 10 days after injection of MOG peptide to induce EAE induction. Cells were cultured for 48 hr in the presence of the indicated concentrations of MOG peptide. Data are the mean ± SEM of triplicates from a single experiment involving 3 mice/genotype.

106

2.4.14 UVRAG deficiency impairs Th2 responses during OVA- mediated asthma induction

Although Th2 differentiation appeared normal in URfl/fl; Lck-Cre mice, we delved deeper into the potential role of UVRAG in Th2-driven pathology using a model of ovalbumin (OVA)-mediated asthma. OVA primed mice treated with aerosolized OVA peptide typically develop the pulmonary inflammation and bronchiolar eosinophilia commonly found in human asthma. We sensitized WT and URfl/fl; Lck-Cre mice with

OVA for 21 days and measured serum cytokines, IgE production, and eosinophil infiltration into the bronchiolar lavage (BAL). Although we observed comparable production of IgE and IL-10 in OVA-treated WT and mutant mice, levels of the Th2 cytokines IL-13 and IL-5 and the inflammatory cytokines IFNγ, TNFα and IL-6 were decreased in the serum of UVRAG-deficient mice (Fig. 1.14A). Eosinophil infiltration into the BAL, a process dependent on the Th2 cytokines IL-4 and IL-5, was also diminished in the mutant mice (Fig. 2.14B). Overall, these data suggest that UVRAG normally contributes to an effective Th2 response during the induction of OVA-mediated asthma.

107

100000 g/ml)

ρ UR fl/fl 10000 URfl/fl; Lck-Cre * 1000

100

1 Serum Analyte Level ( Level Analyte Serum IL-5 IL-6 IgE IL-13 IFN-g TNF a IL-10

UR&fl/fl& UR&fl/fl;&Lck-Cre&

Figure 2.14. Reduced inflammation in URfl/fl; Lck-Cre mice in response to induction n=12)wt/ko) of OVA-mediated asthma.

(A) Quantitation of levels of the indicated cytokines and IgE in serum of WT and URfl/fl; Lck-Cre mice that were sensitized by i.p. injection of 100 µg OVA on days 1, 8 and 15. On days 22, 23, and 24 after the initial sensitization, the mice were challenged for 30 min

108

with aerosolized OVA in PBS using a nebulizer. Mice were sacrificed on day 25, and serum was collected for analysis by ELISA. Data are the cumulative mean ± SEM of 3 independent experiments involving 3-5 mice per genotype. (B) Histological analysis of eosinophils in bronchoalveolar lavage (BAL) from the mice in (A). BAL cells were stained to detect eosinophils. Results shown are representative of 3 independent trials involving 3-5 mice/genotype.

109

2.4.15 UVRAG is essential for CD8+ T cell expansion during acute LCMV infection

The differential effect of UVRAG loss on CD4+ and CD8+ T cell numbers prompted us to investigate the role of UVRAG in CD8+ T cell-driven immune responses. To examine

CD8+ T cell function in vivo, we used an LCMV infection model. In immuno-competent mice, infection with the Armstrong strain of LCMV results in acute disease, robust CD8+

T cell expansion, rapid clearance of the virus and the generation of long-lived memory

CD8+ T cells. We infected WT and URfl/fl; Lck-Cre mice with a high dose (105 pfu) of

Armstrong LCMV and analyzed T cell responses at the usual peak of the response (8 days post-infection) (Fig. 2.15A). Antigen-specific CD8+ T cell responses to the immunodominant LCMV GP33 and NP396 epitopes were measured using fluorescently- labeled tetramers. Both WT and URfl/fl; Lck-Cre mice exhibited robust proliferation of

CD8+ T cells following LCMV infection but the mutants still continued to show a persistent and significant reduction in CD4+ and CD8+ T cell numbers compared with

WT mice (Fig. 2.15B). Importantly, the mutants harbored a reduced proportion and absolute number of CD8+ T cells directed against GP33 (Fig. 2.15C). Consequently, these animals exhibited relatively high viral titres in the brain, lung, spleen and kidney at day 8 post-infection, in contrast to the much lower viral titres in organs of infected WT mice

(Fig. 2.15D).

110

UR fl/fl and URfl/fl;Lck- A& Cre mice

High dose day 8 • Immunophenotyping Armstrong • Tetramer analysis LCMV • Ex vivo restimulation 105 pfu

100 B& ** 100 ) ) 6 6

* UR&ﬂ/ﬂ& Cells (10 Cells Cells (x10 Cells + + CD4 CD8 10 10

fl/fl fl/fl ﬂ/ﬂ UR& &;& UR UR ; Lck-Cre ; Lck-Cre Lck-Cre& fl/fl fl/fl UR UR CD8a) CD4)

C& ) 6 3 * 10 * ﬂ/ﬂ& UR& cells + 2 Cells (x10 Cells CD8 + + 1 CD8

1 + UR&ﬂ/ﬂ&;&

Lck-Cre& Tet GP33 % 0 0.1 GP33 Tet fl/fl fl/fl

UR UR ; Lck-Cre ; Lck-Cre fl/fl

GP33)tetramer) fl/fl CD8a) UR UR

111

Viral titres D& 1000000 UR fl/fl 800000 URfl/fl; Lck-Cre

600000

400000 PFU/organ 200000

0 li ki sp lu br

Figure 2.15. URfl/fl; Lck-Cre mice mount a weak response to LCMV infection.

(A) Schematic diagram of initial LCMV infection protocol. (B) Flow cytometric analysis of CD4+ and CD8+ T cells from the spleens of 6-8 week old WT and URfl/fl; Lck-Cre mice (n=3/group) that were injected with LCMV (105pfu) and examined on day 8 post- injection. Left: Representative plots. Right: Quantitation. Each data point represents an individual mouse: WT, blue dots; URfl/fl; Lck-Cre, red squares. Horizontal lines are the cumulative geometric mean ± SEM. (C) Flow cytometric analysis of GP33 tetramer- specific CD8+ T cells from the mice in (B). Data were analyzed as in (B). (D) Plate count pfu determinations of viral titres in liver (Li), kidney (Ki), spleen (Sp), lung (Lu) and brain (Br) of the infected mice in (A). Results are the mean ± SEM (n=3/group) and are from a single trial.

112

To investigate if these differences in the CD8+ T cell response could be seen during the course of LCMV infection, we monitored numbers of total CD8+ T cells and antigen-specific CD8+ T cells as well as viral titres in PBL on every 3 days following

LCMV infection. Consistent with our observations at day 8 post-infection, total and antigen-specific CD8+ T cells were significantly decreased in UVRAG-deficient mice throughout the course of the infection (Fig. 2.16A-D). Consequently, viral titres appear to be enhanced in mutant PBL as compared with WT (Fig. 2.16E). These results indicate that UVRAG is required for an optimal CD8+ T cell response to LCMV infection.

113

A& B& 60 fl/fl 25 UR ; Lck-Cre URfl/fl; Lck-Cre fl/fl * UR fl/fl * * 20 UR * * 40 15 ** * 10 * 20 % CD4 T cells T CD4 %

% CD8 T cells T CD8 % 5

0 0 0 3 6 8 0 3 6 8 11 14 20 11 14 20 Day post infection Day post infection C& D&

5 5 ** ** 4 4

3 3 * * * * 2 2 *

* cells T CD8 CD8 T cells T CD8 1 1

0 0 % NP396 tetramer positive tetramer NP396 % % GP33 tetramer positive tetramer GP33 % 6 8 6 8 11 14 20 11 14 20 Day post infection Day post infection E& 5000

4000

3000

2000

pfu/ml in serum 1000

0 0 3 6 8 11 14 20 Day post infection

Figure 2.16. Altered time course of T cell response to LCMV infection in URfl/fl;Lck- Cre mice.

WT and URfl/fl; Lck-Cre mice (6-8 weeks old; n=3/group) were injected with LCMV (105pfu) and PBL samples were drawn on the indicated days post-infection. Percentages of PBL cells in these samples that were total CD8+ T cells (A), total CD4+ T cells (B),

114

GP33 tetramer-specific CD8+ T cells (C), and NP396 tetramer-specific CD8+ T cells (D) were determined by flow cytometry. Results are mean ± SEM (n=3) and are representative of 1 trial. (E) Quantitation of viral titres in serum of the mice in (A-D). Results are the mean ± SEM (n=3).

115

2.4.16 UVRAG is not required for CD8+ effector T cell function

Differences in viral clearance can result from poor function of effector T cells [211]. To investigate if loss of UVRAG impairs not only the expansion of CD8+ T cells but also their differentiation into antigen-specific effector CTLs, we examined the capacity of

LCMV-primed effector cells isolated from LCMV-infected WT and URfl/fl; Lck-Cre mice to produce IFNγ following ex vivo restimulation. Splenocytes from LCMV-infected mice were stimulated in vitro with the LCMV-derived peptides GP33 or NP396, and IFNγ production was measured by intracellular cytokine staining plus flow cytometry. The proportion of UVRAG-deficient CD8+ T cells able to produce IFNγ in response to peptide stimulation in vitro was comparable to that in the WT culture (Fig. 2.17A, B). In addition, UVRAG deficiency did not affect the ability of these CD8+ T cells to be activated, as measured by CD44 expression (Fig. 2.17C). Neither were there any significant differences in reactive oxygen species (ROS) levels, apoptosis, or expression of PD-1 or Fas receptors on the cell surface (Fig. 2.17C). Therefore, UVRAG may be essential for optimal CD8+ T cell expansion in response to acute LCMV infection, but it does not influence the ability of these residual cells to differentiate into functional effectors.

116

A& NONE& NP396& GP33&

UR&ﬂ/ﬂ&

UR&ﬂ/ﬂ&;& Lck-Cre& CD8) B& IFN"Υ) cells

+ 15 UR fl/fl URfl/fl; Lck-Cre 10

5 cells of total CD8 total of cells + γ 0 % IFN % None GP33 NP396

117

UR&ﬂ/ﬂ&

UR&ﬂ/ﬂ&;& Lck-Cre& PI) CD11c) CD44) CD62L) CD3) Annexin)V)

URfl/fl& & URfl/fl;& Lck-Cre& Cell&Count& PD1) DCF) CD95)

Figure 2.17. UVRAG is not required for effector CD8 T cell function.

(A) Flow cytometric analysis of the percentages of IFNγ-producing CD8+ T cells in splenocytes that were isolated from WT or URfl/fl; Lck-Cre mice (n=3/group) and stimulated in vitro for 5 hr with no peptide (None), or with NP396 or GP33 peptide. IFNγ production was detected by intracellular staining. Plots are gated on live CD8+ T cells and numbers in upper right quadrants represent IFNγ-producing CD8+ T cells. Data were obtained from one trial. (B) Quantitation of IFNγ-producing CD8+ T cells from the mice in (A). Results are the mean ± SEM (n=3 mice/group) and were obtained from one trial. (C) Top: Flow cytometric analysis of the expression of the indicated markers by splenocytes from the mice in (A). Results were obtained from a single experiment

118

involving 3 mice per genotype. Bottom: Flow cytometric plot of expression of the indicated markers by the mice in (A). No significant differences were seen for any marker examined.

119

2.4.17 UVRAG is essential for memory CD8+ T cell generation, expansion and function

We next assessed the role of UVRAG in generating memory CD8+ T cells. It is well known that the resolution of an acute infection is followed by the generation of long-term memory precursor effector cells (MPECs). These cells can be distinguished by the cell surface expression of IL-7Rα chain (CD127). We therefore compared the kinetics of appearance and expansion of this MPEC population in WT and URfl/fl; Lck-Cre mice following acute LCMV infection. As expected, IL-7Rα population can be seen in WT mice 8 days post infection, coinciding with the clearance of acute LCMV infection (Fig.

2.18 upper panel). However, there is no such population in the absence of UVRAG up to

20 days post infection (Fig. 2.18 middle panel). This data strongly suggests that UVRAG is required for the generation of long term MPECs post acute infection. It also implies that their absence would lead to impaired memory/recall responses in UVRAG-deficient mice.

120

UVRAG is necessary for generating memory precursors effector CD8+ T cells

LCMV Arm Day 8 Day 14 Day 20 CD8+ gate

WT KO Cell Count

CD127

Figure 2.18 UVRAG is required for generating long term MPECs post acute LCMV infection.

WT and URfl/fl; Lck-Cre mice (6-8 weeks old; n=3/group) were injected with LCMV (105pfu) and PBL samples were drawn on the indicated days post-infection. Proportions of PBL CD8+ T cells in these samples that were CD127+ were determined by flow cytometry. Results are mean ± SEM (n=3) and are representative of 1 trial.

121

We next assessed the role of UVRAG in memory CD8+ T cell responses. One month after an acute LCMV infection, WT and URfl/fl; Lck-Cre mice were infected with recombinant Vaccinia virus expressing LCMV-GP. At 5 days after this challenge, we evaluated total and antigen-specific CD8+ T cell numbers and proportions to determine if secondary immune responses were intact in URfl/fl; Lck-Cre mice. Intriguingly, compared to WT mice, URfl/fl; Lck-Cre mice displayed a reduced antigen-specific CD8+ T cell response which was associated with a dramatic drop in splenic CD8+ T cells (Fig. 2.19A).

Our earlier experiment showed that, after a primary LCMV infection, mutant CD8+ T cell levels were approximately half of WT levels. After the secondary infection, however, there were 5-fold fewer total CD8+ T cells in the mutant mice than in WT controls (Fig.

2.19A). When we examined the effector functions of these memory CD8+ T cells using ex vivo restimulation with GP33 peptide, the proportion of UVRAG-deficient CD8+ T cells producing IFNγ was much lower than in the WT (Fig. 2.19B), contrary to our findings for the primary response. These results suggest that UVRAG may be important for the effector functions of memory CD8+ T cells. When we performed an in vivo CTL assay, we observed reduced cytotoxic killing by UVRAG-deficient memory T cells as compared to the WT (Fig. 2.19C). Taken together, our data indicate that UVRAG is important for both primary and secondary antigen-specific T cell responses in vivo, and that UVRAG deficiency renders mice susceptible to both initial and subsequent infection with LCMV.

122

URfl/fl& URfl/fl;& A& 40 2.0 ** **

Lck-Cre& cells + 30 1.5 & CD8 + cells + 20 1.0 Tet

% CD8 % 10 0.5

GP33& 0.0

0 Tet GP33 of %

fl/fl fl/fl

CD8a& UR UR ; Lck-Cre ; Lck-Cre fl/fl fl/fl UR UR URfl/fl& URfl/fl;& B& Lck-Cre& 0&GP33&

40 UR fl/fl ** URfl/fl; Lck-Cre 30

1uM& 20 & GP33 10 & cells of total CD8+ cells total of cells + γ 0 IFNγ % IFN % CD8a& None GP33 C& fl/fl URfl/fl& UR ;& Lck-Cre& Count& CFSE&

Figure 2.19. UVRAG deficiency impairs memory responses to LCMV-GP- expressing Vaccinia virus.

123

Figure 2.19. UVRAG deficiency impairs memory responses to LCMV-GP- expressing Vaccinia virus.

(A) Left: Flow cytometric analysis of total and GP33 tetramer-specific CD8+ T cells among splenocytes isolated from WT or URfl/fl; Lck-Cre mice that had been infected first with high dose LCMV and then one month later with a high dose of LCMV-GP- expressing Vaccinia virus. Splenocytes were examined 5 days after the second infection to detect GP33 tetramer-specific CD8+ T cells. Numbers are percentages of live splenic lymphocytes and are representative of 1 trial. Right: Quantitation of total CD8+ T cells and GP33+CD8+ T cells from the mice in the left panel. Data points are individual mice and horizontal lines are geometric means ± SEM (n=3 mice/group). (B) Left: Flow cytometric analysis of IFNγ production by the CD8+ T cells in (A) restimulated in vitro (or not) with GP33. IFNγ production was detected by intracellular staining. Numbers are percentages of live splenic lymphocytes and are representative of 1 trial. Right: Quantitation of IFNγ-producing CD8+ T cells among splenocytes of the mice in the left panel. Data points are individual mice and horizontal lines are geometric means ± SEM (n=3 mice/group). (C) In vitro CTL cytotoxicity assay. For target cells, CD45.1+ splenocytes from control mice were labeled with CFSE and pulsed (or not) with GP33 peptide. These two target cell populations were mixed 1:1 and cultured with CD45.2+GP33+CD8+ T cells from WT or URfl/fl; Lck-Cre mice for 5 hrs at 37°C. Target cell killing was monitored by flow cytometric analysis of dilution of CFSE-labeled target cells. Numbers are percentages of CFSE+ target cells that remained after 5 hrs. Data were obtained from a single experiment involving 3 mice per genotype.

124

2.5 Discussion

The objective of this study was to determine the precise roles of UVRAG, an autophagic tumour suppressor, in T cell biology in vivo. By conditionally deleting the UVRAG gene in developing T cells, we have identified an essential and cell-intrinsic role for UVRAG in maintaining the survival and homeostasis of peripheral naïve T cells (Fig. 2.20).

Intriguingly, our data suggest that these UVRAG functions are likely independent of its role in autophagy. Indeed, we have shown that UVRAG is not required for mitophagy or activation-induced autophagy in T cells. In vivo, we have demonstrated that UVRAG is important for T cell-mediated antigen-specific immune responses to LCMV infection

(Fig. 2.21). At the mechanistic level, our data imply that UVRAG likely acts downstream of homeostatic cytokine signalling pathways to maintain naïve T cell homeostasis. Our findings provide new insights into the regulation of autophagy in T cells and identify

UVRAG as a novel regulator of naïve and memory T cell homeostasis.

125

BM Thymus Periphery

Peripheral Pool

DN DP SP Naïve Death β selection Lineage Pool Survival Commitment CD4+

CD8+ Homeostatic Proliferation

iNKT

Memory

TCR IL-7 IL-15 Activation and Proliferation

Figure 2.20. UVRAG is important for many aspects of T cell biology.

A thorough examination of the phenotype of URfl/fl:Lck-Cre mice allowed us to identify the many functions of UVRAG in T cell biology (highlighted by red arrows). These include: defects in development an homeostasis of iNKT cells, reduced peripheral compartment of conventional T cells, impaired survival and homeostatic proliferation of naïve T cells, and defects in generation/maintenance/function of memory T cells post infection.

126

UVRAG is important for 1o and 2nd response to LCMV

1o!Infec@on! 2nd!Infec@on! response+ +

UR fl/fl UR fl/fl; Lck-Cre Memory! Agspecific CD8

Time

Figure 2.21. UVRAG is required for optimal primary and secondary immune response to LCMV infection.

Model of WT and URfl/fl;Lck-Cre mice response to primary and secondary infections. UVRAG deficient mice mount a weak response to acute LCMV infection. Upon resolution of infection, these mice fail to generate long term memory precursor cells, leading to an absent secondary response to GP-Vaccinia infection.

127

Peripheral T cell homeostasis is a tightly regulated process that ensures a constant pool of naïve T cells in the secondary lymphoid organs and blood circulation. In addition to the influx into the periphery of recent thymic emigrants and the loss of T cells through normal programmed cell death, mechanisms that promote the survival and homeostatic proliferation of naïve T cells are essential for maintaining this balance. Previous work by others has established that IL-7 signalling and interactions between TCRs and self peptide-MHC complexes are crucial for naïve T cell survival and homeostatic proliferation. We found that naïve T cells lacking UVRAG were recovered in lower numbers after adoptive transfer into lymphopenic hosts, and also were refractory to the effects of in vitro pro-survival IL-7 signalling compared to WT T cells (Fig. 2.22). These results suggest that UVRAG influences peripheral naïve T cell homeostasis by regulating the survival and homeostatic proliferation of these cells. UVRAG-deficient T cells showed hyper-proliferation in TCR-mediated stimulation in vitro, implying that the interaction between TCRs and self peptide-MHC complexes was likely intact. In addition, our mechanistic data indicate that UVRAG is not required for the IL-7 receptor trafficking, STAT5α activation, or Bcl-2 upregulation known to contribute to T cell homeostasis. The exact mechanism underlying UVRAG’s effects on IL-7 signalling remains under investigation (Fig. 2.22).

128

UVRAG WT UVRAG KO

Naïve T cell Memory T cell IL-7 IL-7

Cytoplasm Cytoplasm

PI3K PI3K

Akt Akt "BCL-2, MCL-1 "BCL-2, MCL-1 !BAX, BIM, BAD !BAX, BIM, BAD FOXO FOXO Cell Survival Cell Survival Cell Proliferation Cell Proliferation

Figure 2.22. Proposed mechanism of action of UVRAG downstream of IL-7 signalling in naïve and memory T cells. UVRAG-deficient naïve T cells demonstrated impaired homeostatic proliferation responses in vivo and were refractory to IL-7 mediated survival in vitro. This led us to propose that UVRAG likely acts downstream of homeostatic IL-7 signalling pathway to maintain naïve T cell homeostasis.

129

An unexpected finding was that UVRAG was not required for activation-induced autophagy in T cells. Many in vitro studies have pointed to an indisputable role for

UVRAG in autophagy. These reports show that UVRAG is required not only for the initiation of autophagosome formation but also autophagosome maturation. Since these diverse activities depend on UVRAG’s ability to bind to the Beclin-1 and C/Vps complexes, respectively, it was surprising that we did not see a defect in autophagy in

UVRAG-deficient T cells. However, it is possible that UVRAG’s role in autophagy is cell type-specific, making it necessary to study UVRAG deletion in other cell types.

In addition to the presence of normal T cell autophagy, the phenotype of URfl/fl;

Lck-Cre mice differs significantly from that of other autophagy gene-deficient T cell- specific mouse models. In contrast to our UVRAG-deficient mutants, mice lacking

ATG5, ATG7, ATG3 or Beclin1 in T cells all show peripheral lymphopenia that is largely explained by defects in cell survival and/or impaired TCR-mediated proliferation

(at least in vitro) [61, 63, 131, 139]. Mice deficient for ATG5 or ATG7 also show defects in clearance of their mitochondria that lead to a build-up of ROS and consequently cell death [58, 60, 127]. We found that T cells from our UVRAG-deficient mice exhibited not only normal autophagy but also normal survival and antigen-induced proliferation in vitro, normal mitochondrial mass, and normal ROS levels (data not shown). The phenotype of T cell-specific UVRAG deletion is thus very different from that of loss of other autophagy-related genes.

The effect of UVRAG inactivation on T cell development is still unclear. In retrospect, the use of URfl/fl; Lck-Cre mice to study this process was not ideal, as Lck expression, and thus the consequent deletion of UVRAG, initiates relatively late during 130

the DN2 to DN3 stages of T cell development. Depending on the half-life of the UVRAG protein, it may or may not have been completely eliminated from maturing thymocytes.

To study UVRAG’s role in T cell development from its beginning, URfl/fl mice should be bred to transgenic Vav-Cre mice, which would ensure UVRAG deletion in the earliest T cell progenitor populations. Nevertheless, because we observed complete deletion of

UVRAG protein in peripheral T cells (Fig. 2.1D), we consider URfl/fl; Lck-Cre mice to be a good model for studying mature naive T cell homeostasis in the periphery. Our finding that UVRAG loss only affected T cell development when WT T cells were present as competitors is quite intriguing and implies a helpful role for UVRAG when resources are limited. Normal T cell development and peripheral naive T cell homeostasis depend on adequate access to both particular cytokines and self peptide-MHC complexes presented by professional antigen-presenting cells (APCs). Theoretically, APCs should not be affected in our URfl/fl; Lck-Cre mice, leading to our hypothesis that the major function of

UVRAG in T cells is to ensure adequate response to homeostatic cytokines.

One of our most interesting results was that UVRAG loss exerted a differential effect on CD4+ and CD8+ T cells. This phenomenon, in which UVRAG deficiency always led to a greater defect in CD8+ T cells compared with CD4+ T cells, was consistently observed in steady-state mice as well as in BM chimeras. In line with this observation, levels of UVRAG mRNA and protein were higher in purified peripheral WT

CD8+ T cells compared to WT CD4+ T cells. Importantly, this differential effect extended to pathologies involving one or the other of these T cell subsets. An absence of UVRAG did not increase the susceptibility of mice to induction of EAE, a pathology driven by

CD4+ Th1 and Th17 cells. However, UVRAG deficiency did render mice more

131

susceptible to primary and secondary LCMV infections, clearance of which requires an effective CD8+ T cell response. Further experiments designed to shed light on the differential effect of UVRAG loss on CD4+ and CD8+ T cells are under way.

In conclusion, our data have identified UVRAG as a novel regulator of peripheral

T cell homeostasis, a process crucial for maintaining effective immunity. Furthermore, our findings highlight critical differences between UVRAG and other autophagy-related genes, and indicate that this regulator has important non-autophagic functions in T cell biology.

132

Chapter 3

ANP32B

A version of this chapter is published:

Reilly PT, Afzal S, Gorrini C, Lui K, Bukhman Y, Wakeham A, Haight J, Ling TW, Cheong CC, Elia A, Turner PT, Mak TW. The Acidic Nuclear Phosphoprotein 32kDa (ANP32)B-deficient mouse reveals a hierarchy of ANP32 importance in mammalian development. ProcNatlAcadSci U SA 108(25):10243-10248. 2012.

133

3 The Acidic Nuclear Phosphoprotein 32kDa ANP32 (B)-Deficient Mouse Reveals a Hierarchy of ANP32 Importance in Mammalian Development

3.1 Abstract

The highly conserved ANP32 proteins are proposed to function in a broad array of physiological activities through molecular mechanisms as diverse as phosphatase inhibition, chromatin regulation, caspase activation, and intracellular transport. Based on previous analyses of mice bearing targeted mutations of Anp32a or Anp32e, there has been speculation that all ANP32 proteins play redundant roles and are dispensable for normal development. However, more recent work has suggested that ANP32B may in fact have functions that are not shared by other ANP32 family members. Here we report that ANP32B expression is associated with a poor prognosis in human breast cancer, consistent with the increased levels of Anp32b mRNA present in proliferating wild type

(WT) murine embryonic fibroblasts (MEFs) and stimulated WT B and T lymphocytes.

Moreover, we show that, contrary to previous assumptions, Anp32b is very important for murine embryogenesis. In a mixed genetic background, ANP32B-deficient mice displayed a partially penetrant perinatal lethality that became fully penetrant in a pure

C57BL/6 background. Surviving ANP32B-deficient mice showed reduced viability due to variable defects in various organ systems. Study of compound mutants lacking

ANP32A, ANP32B and/or ANP32E revealed previously hidden roles for ANP32A in mouse development that only became apparent in the complete absence of ANP32B. Our

134

data demonstrate a hierarchy of importance for the mammalian Anp32 genes, with

Anp32b being the most critical for normal development.

135

3.2 Introduction

ANP32B (a.k.a APRIL, PAL31, PHAP1b) is a mammalian member of the highly conserved “acidic nuclear phosphoprotein 32kDa” (ANP32) family of gene products

[157]. These metazoan-specific factors are characterized by the presence of an amino- terminal leucine-rich repeat domain and a carboxy-terminal region that is highly enriched in acidic amino acid residues [155]. These features are found in ANP32 proteins from mapmodulin, the single representative in Drosophila, and the three vertebrate family members identified, ANP32A, ANP32B, and ANP32E [156, 157].

The ANP32 proteins have been implicated in a broad array of physiological processes, including cell differentiation [158-161], apoptotic cell death [162-168], and cell proliferation [169-171]. Diverse mechanisms have been postulated for how these proteins perform their function(s). Some studies indicate that ANP32 proteins may directly control enzymatic activities, such as via inhibition of protein phosphatase 2A

(PP2A) [172-174] or activation of caspases [164, 165, 167, 168, 175]. Other studies suggest that they may regulate intracellular transport at nuclear pores or microtubules

[176-178]. Several studies present evidence that nuclear ANP32 proteins may influence transcription either through the “inhibitor of acetyl transferases” (INHAT) complex [179-

182] or by direct effects on transcription factors [183, 184]. Most of the above reports have focused on ANP32A, the founding member of the ANP32 family, but none of these studies has specifically excluded a particular ANP32 protein as contributing to the activities examined.

More recent work has demonstrated functions that are exclusive to the ANP32B protein, at least in humans. Firstly, ANP32B, but not ANP32A, controls the expression of 136

the dendritic cell (DC) maturation factor CD83 by regulating the transport of its mRNA to the cytoplasm [176]. Secondly, ANP32B modulates the activity of the transcription factor Kruppel-like factor 5 (KLF5), whereas ANP32A cannot [185]. Finally, ANP32B is a caspase substrate, whereas ANP32A is not [186].

Reported loss-of-function mutants for ANP32 family members include two independently targeted ANP32A-deficient mice [199, 200], an ANP32E-deficient mouse

[200], and a presumptive null mutant of mapmodulin in Drosophila [201]. All of these mutants were viable and fertile with no obvious abnormalities. Here, we report on the phenotype of Anp32b-deficient (Anp32b-/-) mice and show, using compound mutants, that

ANP32 family members are not completely redundant in mammals. We demonstrate that

ANP32B is the most important ANP32 family member for mammalian development and likely plays a complex role in sustaining viability that has yet to be completely defined.

In addition, we provide evidence that the level of ANP32B mRNA expression in human breast cancers may serve as a prognostic marker.

137

3.3 Materials and Methods

3.3.1 Prognostic marker identification

Data were acquired from three publicly available datasets [212-214]. The prognostic risk associated with particular gene expression was computed using application of the Cox proportional hazard regression model.

3.3.2 Northern blotting

The multiple murine tissue blot for Northern analysis was acquired from Zyagen (cat#

MN-MT-2). Anp32b mRNA was detected by hybridizing to a labeled probe consisting of the EcoRI-BamHI fragment at the 5’ end of the open reading frame. Gapdh was detected by probing with a 214 bp fragment that hybridized to exons 5, 6 and 7.

3.3.3 Real-time RT-PCR

RNA was extracted from specified cell populations using a standard protocol and the

RNeasy kit (Qiagen), quantified, and reverse-transcribed using a Superscript II first strand synthesis kit (Invitrogen). cDNA samples were then used as templates for quantitative real-time PCR (qPCR) using a ABI 7900HT detection system and SYBR

Green (Applied Biosystems). Data were normalized to either Actb (β-actin) or Tbp mRNA or Rn18S (18S rRNA). Primer sequences are provided in Supplementary Table

3.1. Statistical significance of differences in normalized values were assessed by

Student’s t-test.

3.3.4 Mice

Mice were maintained under specified pathogen-free conditions in individually ventilated cages and fed 5% irradiated meal. Anp32b-/- mice analyzed in this study were derived

138

from two separate homologous recombinant clones and analyzed in approximately equal proportion for all experiments. Unless otherwise stated, analyses were performed in the mixed-bred 129ola;C57BL/6 genetic background. Statistical analyses for weights and longevity were performed using Student’s t-test and log-rank analysis, respectively.

3.3.5 Plasmids and primers

Sequences directing the expression of diphtheria toxin A in mouse ES cells were cloned into pBluescript (pgk-neo) to give the plasmid pBSneoDTA [200]. Regions of the

Anp32b gene adjacent to the targeted exons were cloned by high fidelity PCR. Primers used to clone the upstream and downstream arms of homology into the XhoI site and

XbaI site, respectively, are shown in Table 3.1.

3.3.6 Gene targeting

Targeting constructs were linearized and transfected by electroporation into E14K mouse

ES cells as previously described [215, 216]. Southern probes used to detect Anp32b genomic sequences were amplified from genomic ES cell DNA. PCR primer sequences for Anp32b flanking probe generation are presented in Table 3.1.

139

Table S2. Primer sequences used in this analysis Name Use Sequences

32b 1 Anp32b qPCR 5′-AGCCGTTCGAGAACTTGTCTT-3′ 32b 2 Anp32b qPCR 5′-CAGGTTATTGCCACTTAGGTTCA-3′ Tbp 1 Tbp qPCR 5′-GCTCTGGAATTGTACCGCAG-3′ Tbp 2 Tbp qPCR 5′-CTGGCTCATAGCTCTTGGCTC-3′ Myc 1 Myc qPCR 5′-CTGGATTTCCTTTGGGCGTT-3′ Myc 2 Myc qPCR 5′-TGGTGAAGTTCACGTTGAGGG-3′ Actin 1 Actb (β-actin) qPCR 5′-TAGCCATCCAGGCTGTGC-3′ Actin 2 Actb (β-actin) qPCR 5′-TCAGGATCTTCATGAGGTAG-3′ 18S 1 Rn18S (18S rRNA) qPCR 5′-AGTTCCAGCACATTTTGCGAG-3′ 18S 2 Rn18S (18S rRNA) qPCR 5′-TCATCCTCCGTGAGTTCTCCA-3′ uAOH 1 Cloning upstream arm of homology 5′-CCCCTCGAGTCTTTGGACCATGTTATAAATGTGTACTAGCTGGC-3′ uAOH 2 Cloning upstream arm of homology 5′-GGGGTCGACTCACTACCATCACTCAGAGTTCCAATAGTCTTCTG-3′ dAOH 1 Cloning downstream arm of homology 5′-GGGTCTAGAAACTCAATAGTAGATCAGGCTGGC-3′ dAOH 2 Cloning downstream arm of homology 5′-GGGTCTAGACCACACAACTCAGCAGTTCTCAG-3′ FPfwd FlankingSouthernprobesynthesis 5′-CACCTGGAGGGTTCACTGAGAATAAATTG-3′ FPrev FlankingSouthernprobesynthesis 5′-CACTACCAAAATGCACAGACGTAAGGTTAAG-3′ 32b wt-rev Anp32b wt genotyping 5′-GGCACACTTACAGAGTTCGGTTCACAAGTTGAGC-3′ 32b-fwd Anp32b wt and mutant genotyping 5′-GGTCACAGTGTCTCTTCACATCAGTAAAACCCTAAGTAAG-3′ 32a wt-rev Anp32a wt genotyping 5′-GAATGAGGTGAGAGGTCAAGATTCAGCTGC-3′ 32a-fwd Anp32a wt and mutant genotyping 5′-CTAATCCCTCTTCAGAGAACTGCCCTGTTCAAG-3′ 32e wt-rev Anp32e wt genotyping 5′-GGAGTCACGAGCTAACTGCACTGTCACCAAC-3′ 32e-fwd Anp32e wt and mutant genotyping 5′-GGAGGAAGATGACGATGAGGATGAAGCTGG-3′ neo-rev1 Anp32b, Anp32a, and Anp32e mutant genotyping 5′-CTACCGGTGGATGTGGAATGTGTGCG-3′ qPCR, quantitative real-time PCR.

Table 3.1. Primer sequences used in this analysis

Reilly et al. www.pnas.org/cgi/content/short/1106211108 4of4

140

3.3.7 Genotyping

For standard genotyping, PCR was performed on proteinase-treated biopsies. Primer sequences are provided in Table 3.1. Statistical analyses of progeny derived from intercrosses were performed using Chi square analysis.

3.3.8 Immunoblotting

Mouse thymi were homogenized and lysed in buffer containing NP40 detergent. Protein extracts (5 µg) were fractionated by SDS-PAGE, transferred to PVDF membranes, and probed with anti-ANP32B antibody (10843-1-AP, Proteintech) or anti-β-tubulin (loading control; Li-Cor Bioscience). Protein bands on blots were visualized using Li-Cor

Odyssey standard protocols.

3.3.9 Cell populations

Primary MEFs were prepared from littermate Anp32b+/+, Anp32b+/- and Anp32b-/- E14.5 embryos derived from timed interbreedings of Anp32b+/- mice. For transformed MEFs, primary MEFs (5x105 cells) were infected with a retrovirus generated from pLPC

E1A/rasV12 using published techniques [217]. Single cell suspensions of total thymocytes were prepared from three sex-matched littermate pairs of Anp32b +/+ and -/- mice (4-8 weeks old). Single cell suspensions derived from spleen or lymph nodes were treated with red blood cell lysis buffer (Sigma) to remove erythrocytes, and B and T cells were purified using the IMag cell separation system (Becton Dickinson). Briefly, total leukocytes were incubated with mouse anti-CD16/32 blocking antibodies, after which T cells were negatively selected using biotinylated anti-

141

Ter119/B220/CD19/CD11b/Nk1.1/CD11c antibodies, and B cells were positively selected using biotinylated anti-CD19/B220 antibodies.

3.3.10 Apoptosis

Apoptosis assays were performed as described previously [3, 200]. Briefly, cells were exposed to apoptotic stimuli as detailed in Fig. 3.8 and cultured overnight in medium containing penicillin/streptomycin, 10% fetal bovine serum (FBS) and L-glutamine.

Viability was determined by propidium iodide exclusion and flow cytometry using a

FACSCaliber (BD Pharmingen).

3.3.11 Proliferation

The growth of primary MEFs in culture was analyzed as previously described [200].

Purified B cells were treated with 1 µg/ml LPS for 24 hrs. Purified T cells were stimulated by seeding on plates pre-coated with anti-CD3 and anti-CD28 antibodies as previously described [218]. Statistical differences in cell growth were assessed using

Student’s t-test.

142

3.4 Results

3.4.1 ANP32B as a potential prognostic marker in human cancer

Reports that ANP32B expression was related to cell proliferation [169, 171] prompted us to investigate whether ANP32B expression was altered in human tumor samples. We examined the relationship between ANP32B mRNA expression and breast cancer patient prognosis using information from three available studies, namely the NEJM [214], the

BMC [212], and the PNAS datasets [213]. When we compared the survival of these three groups of patients among all three datasets, we found that patients whose tumors showed the highest ANP32B mRNA levels had significantly shorter survival times using the Cox proportional hazard model (p<10-4). To display this effect, we stratified the patients in these datasets into three major groups: those with tumors expressing high levels of

ANP32B mRNA (highest 33.3% or tertile), tumors with medium ANP32B mRNA levels

(middle tertile), and tumors with low ANP32B mRNA levels (lowest tertile). Here, the survival effect of the highest and lowest ANP32B expressing tumors is clearly demonstrated (Fig. 3.1; middle tertile not shown). These results imply that elevated

ANP32B expression in a tumor may be a marker of poor patient prognosis.

143

Figure 3.1. ANP32B mRNA expression is a marker for aggressive breast cancer.

Survival analyses of patients with tumors expressing the highest levels of ANP32B mRNA (highest tertile; dashed line), and patients with tumors expressing the lowest levels of ANP32B mRNA (lowest tertile; solid line). Middle tertile is not shown. Values were acquired from three separate datasets: NEMJ [214], PNAS [213], and BMC [212].

144

3.4.2 Anp32b mRNA levels are linked to cell proliferation in mice

The less favorable prognosis of breast cancer patients with high ANP32B-expressing tumors suggested that the elevated ANP32B in these samples might be linked to more aggressive tumor cell proliferation. To investigate this hypothesis, we performed

Northern blot analysis of a wide range of tissues from adult wild type (WT) mice. We found that Anp32b expression was low (relative to Gapdh) in tissues where less cell proliferation usually occurs, i.e. in brain (Fig. 3.2A). In contrast, tissues that generally have high cell proliferation rates, i.e. spleen and thymus, showed higher relative levels of

Anp32b mRNA expression. We confirmed this pattern at the mRNA level by quantitating

Anp32b expression (relative to TATA binding protein or Tbp) using quantitative RT-PCR

(Fig. 3.3A). Again, proliferative tissues such as the spleen and thymus showed elevated levels of Anp32b mRNA, whereas the forebrain and hindbrain exhibited only low Anp32b expression. These data further suggest a link between Anp32b level and rate of cell proliferation.

To test whether proliferative stimuli could induce Anp32b expression, we subjected WT primary mouse embryo fibroblasts (MEFs) to serum deprivation in order to induce cell cycle arrest, and then stimulated their proliferation by restoring standard serum levels to the culture. Under conditions of serum deprivation, Anp32b mRNA was barely detectable in WT MEFs (Fig. 3.2B). However, upon the addition of serum and the resumption of proliferation, Anp32b mRNA was significantly increased (p<0.01) by more than 38-fold relative to control Actb. Under identical conditions, the induction of mRNA encoding Myc, a well-documented early proliferation marker, achieved a 4.8-fold increase (Fig. 3.3B). Similarly, WT B cells stimulated with LPS, and WT T cells

145

activated with anti-CD3 plus anti-CD28 antibodies (activators of the T cell receptor complex and T cell costimulatory complex, respectively), showed significant elevation

(p<0.01) of Anp32b mRNA levels (Fig. 3.2C). These results indicate that Anp32b is upregulated under conditions that favor cell proliferation.

146

Figure 3.2. Elevated Anp32b mRNA expression correlates with increased cell proliferation.

(A) Correlation with highly proliferative tissues. A commercial mouse tissue blot was examined by Northern blotting to detect Anp32b mRNA. Gapdh, relative expression control. Anp32b mRNA is decreased relative to Gapdh in brain and muscle. (B) Correlation with nutrient-induced proliferation. WT MEFs were deprived of serum for 72 hrs, and then re-supplied with medium containing 10% FBS for 24 hrs. Anp32b mRNA levels were determined by qPCR relative to β-actin. Data shown are the mean ± SD of two independent WT MEF cultures. **p<0.01. (C) Correlation with lymphocyte stimulation. Left: Purified B cells from WT mice were treated with 1 µg/ml LPS for 24 147

hrs. Right: Purified T cells from WT mice were stimulated with plate-bound anti-CD3 (2 µg/ml) plus anti-CD28 (0.2 µg/ml) for 36 hrs. In both cases, Anp32b mRNA levels were determined by qPCR relative to 18S rRNA. Results shown are the mean ± SD of triplicates and are expressed as fold increase over levels in untreated controls. Results are representative of two trials. **p<0.01

148

Figure 3.3. Correlation of elevated Anp32b mRNA expression with cell proliferation.

(A) Quantitation of Anp32b mRNA levels in normal tissues. Total RNA from the indicated tissues of 8-week old wild-type (WT) C57BL6 mice was harvested and subjected to quantitative RT-PCR to detect expression of Anp32b mRNA. Values were

149

normalized to TATA-box binding protein (Tbp) expression and compared to the expression level in the skin. Data shown are the mean fold induction in relative Anp32b mRNA expression ± SE. Results shown are representative of independent RNA isolations from three pair of different WT mice. (B) Upregulation of Myc mRNA expression. WT MEFs were deprived of serum for 72 hours, and then re-supplied with medium containing 10% FBS for 24 hrs. Myc mRNA levels were determined by quantitative RT-PCR relative to Actb (β-actin). Data shown are the mean ± SD of two independent WT MEF cultures. No statistically significant differences were detected.

150

3.4.3 Gene-targeting of Anp32b in mice

To investigate the physiological functions of ANP32B in vivo, we generated mice in which the Anp32b gene was disrupted by replacing the genomic region containing exons

2, 3, and 4 with a pgk-neo expression cassette (Fig. 3.4A). This mutation results in an mRNA in which the last 17 codons of exon 1 are linked to an out-of-frame exon 5.

Southern blotting of the genomic DNA of E14K ES cell clones demonstrated a single appropriate insertion of the pgk-neo sequence (Fig. 3.4B, left). Anp32b+/- mice (mixed

129:C57BL/6 background) were then generated and intercrossed to create Anp32b-/- progeny. Genomic deletion was confirmed in primary MEFs derived from Anp32b-/- embryos using a flanking genomic probe (Fig. 3.4B, right). Immunoblotting of extracts of

Anp32b-/- MEFs demonstrated that there was no detectable ANP32B protein in nullizygous mutants (Fig. 3.4C).

151

Figure 3.4. Generation and validation of Anp32b-deficient mice.

152

Figure 3.4. Generation and validation of Anp32b-deficient mice.

(A) Targeting of the murine Anp32b gene. Top: Murine genomic Anp32b locus showing exons and regions of homology (grey). FP, flanking probe detecting intron 1. H, HindIII; E, EagI sites. The translation start site (ATG) and stop sites are shown. Middle: Targeting construct that replaced exons 2 to 4 with a pgk-neo cassette. DTA, diphtheria toxin A sequence for negative selection. Bottom: Targeted Anp32b allele. Diagnostic HindIII fragments for the WT Anp32b allele (5.4 kb) and the targeted Anp32b allele (5.2 kb), as well as the diagnostic EagI fragment for neo insertion (8.3 kb), are shown. (B) Confirmation of Anp32b gene deletion. Left: DNA from Anp32b+/- and +/+ ES cells was Southern-blotted with a neo probe to confirm a single vector insertion. Right: DNA from MEFs derived from Anp32b+/-, Anp32b-/- or Anp32b+/+ mice was Southern-blotted using the flanking probe. WT HindIII band = 5.4 kb; mutant (MT) HindIII band = 5.2 kb. (C) Loss of ANP32B protein. Anp32b+/+, +/- and -/- MEFs were immunoblotted to detect ANP32B protein. A and B are MEFs from two different Anp32b-/- littermate embryos. β- tubulin, loading control. Results shown in (B) and (C) are representative of three trials.

153

3.4.4 ANP32B protects against perinatal lethality

The intercrossing of our mixed-bred 129:C57BL/6 Anp32b+/- mice to generate Anp32b-/- mutants revealed a defect in mutant mouse viability. At the time of weaning

(approximately 21 days of age), Anp32b-/- mice were present at a far lower frequency

(7%) than the 25% expected according to the standard Mendelian distribution (p<0.01;

Table 3.2). To exclude the possibility that the observed defect was due to a linked mutation present in the ES cell clones, we backcrossed the Anp32b mutant allele into the

C57BL/6 background for six generations. The penetrance of the survival defect was increased in the pure C57BL/6 background (p<0.01; Table 3.3), strongly suggesting that the defect in viability is not attributable to a linked gene mutation.

154

Table 1. Reduced survival rate of Anp32b-deﬁcient mice Anp32b+/+ Anp32b+/− Anp32b−/−

Expected 39 78 39 Observed* 57 87 11

Anp32b+/− mice of mixed 129ola:C57BL/6 background were intercrossed and pups were genotyped at time of weaning (between postnatal 19 and 22 d). “Expected” indicates the distribution for the indicated total number of mice under conditions of normal Mendelian segregation. “Observed” indicates the actual numbers of mice of the indicated genotypes obtained. *P < 0.01 by χ2 analysis.

remained viable until at least E17.5. Taken together, our data suggest that the defect(s) killing Anp32b−/− mice occur perinatally, i.e., at or very near the time of birth. Rigorous pathological examinations of eight Anp32b−/− E17.5 embryos as well as three postnatal day 1 (P1) pups of both the mixed 129:C57BL/6 and the pure C57BL/6 backgrounds revealed no obvious gross abnormalities in a majority of embryos examined. Some Anp32b−/− E17.5 embryos did show subtle and sporadic craniofacial abnormalities, including overlarge ventricles in the brain, expanded inner ear cavities, and palate closure defects. The heart and lungs were not obviously affected. Interestingly, large hematomas in the liver, aortic arch, or umbilical artery were observed in three of eight E17.5 embryos and in one of three P1 pups of the mixed mutant background. However, no one defect appeared to be responsible for the perinatal lethality of all Anp32b−/− mice.

Surviving Anp32b−/− Mice Show Reduced Size and Decreased Life fi Fig. 3. Generation and validation of ANP32B-deficient mice. (A)Targeting Span. To investigate the effects of ANP32B de ciency in adult −/− of the murine Anp32b gene. (Top)MurinegenomicAnp32b locus showing mice, we monitored our surviving mixed-bred Anp32b mice exons and regions of homology (gray). FP: flanking probe detecting intron 1. and their littermates from time of weaning until age approxi- / H: HindIII; E: EagI sites. The translation start site (ATG) and stop sites are mately 1 y. At time of weaning, surviving Anp32b− − female mice shown. (Middle) Targeting construct that replaced exons 2–4withapgk-neo weighed, on average, 42% less than WT females and 35% less cassette. DTA: diphtheria toxin A sequence for negative selection. (Bottom) than their female Anp32b+/− littermates (P < 0.05; Fig. 4C). A Targeted Anp32b allele. Diagnostic HindIII fragments for the WT Anp32b similar trend was seen for males, although the differences did not allele (5.4 kb) and the targeted Anp32b allele (5.2 kb), as well as the di- reach statistical significance in this analysis (Fig. S2). When the agnostic EagI fragment for neo insertion (8.3 kb) are shown. (B)Confirma- +/ +/+ data for male and female sex-matched littermates were com- tion of Anp32b gene deletion. (Left)DNAfromAnp32b − and Anp32b ES fi fi bined, a statistically signi cant reduction in body weight was cells was Southern-blotted with a neo probe to con rm a single vector in- −/− +/− sertion. (Right) DNA from MEFs derived from Anp32b+/−, Anp32b−/−,or apparent when Anp32b and Anp32b littermates were fi−/− +/+ Anp32b+/+ mice was Southern-blotted using the flanking probe. WT HindIII Tablecompared 1. Reduced (P < 0.01), survival as well rate as of whenAnp32bAnp32b-de cientand miceAnp32b band = 5.4 kb; mutant (MT) HindIII band = 5.2 kb. (C) Loss of ANP32B pro- littermates were compared+/+ (P < 0.05). +/− −/− Anp32b−/− Anp32b Anp32b tein. Anp32b+/+, Anp32b+/− and Anp32b −/− MEFs were immunoblotted to Surviving Anp32b mice were not robust and showed signs −/− detect ANP32B protein. A and B are MEFs from two different Anp32b−/− Expectedof premature aging. Many 39Anp32b mutants 78 exhibited early 39 onset littermate embryos. β-Tubulin: loading control. Results shown in B and C are Observed*kyphosis in which the curvature57 of the upper 87 spine was evident 11 at representative of three trials. 4 mo of age and severe by 6 mo (Fig. 4D). Increased hepatocyte +/ polyploidy,Anp32b − anothermice of mixed hallmark 129ola:C57BL/6 of premature background aging, was were also intercrossed present andin the pups aged wereAnp32b genotyped−/− mice. at time However, of weaning premature (between alopecia, postnatal 19which and −/− Table “3.2. Reduced” survival rate of Anp32b-deficient mice. To define precisely when the majority of Anp32b mice died, 22isd). a commonExpected aspectindicates of early the distribution aging, was for not the seen indicated in the total kyphotic number fi of mice under conditions of normal Mendelian segregation. “Observed−/− ” GENETICS we monitored litters from the time of rst detection (12 h mice. Consistent+/- with these phenotypic features, Anp32b mice −/− Anp32bindicates the mice actual from numbers mixed of 129ola:C57BL/6 mice of the indicated background genotypes were obtained. intercrossed and pups postnatal) until weaning. However, virtually all Anp32b mice 2 < *hadP < 0.01 a reduced by χ analysis. life span (P 0.01; Fig. 4E), but no consistent that were evident at first detection survived to adulthood. This werepathology genotyped could at time be identiof weaninfiedg as(between the cause postnatal of their19 and premature 22 d). Expected numbers observation indicated that the developmental problem(s) killing aredeath. the distribution Two cases for of the megacolon indicated total and number two cases of mice of hydroureter under conditions of normal −/− most Anp32b mice occurred either during embryogenesis or Mendelianremained segregation.viable until Observed at least E17.5.are the Taken actual numberstogether, of our mice data of the indicated before normal litter detection. In the course of generating our genotypessuggest that obtain theed. defect(s) killing Anp32b−/− mice occur perina- −/− fl primary MEFs, we noted that Anp32b embryos were present at tally,Table i.e., 2. at Genetic or very background near the time in uence of birth. survival rate of Anp32b- fi approximately Mendelian rates on embryonic day 14.5 (E14.5). deRigorouscient mice pathological examinations of eight Anp32b−/− E17.5 −/− We therefore chose to follow the fate of Anp32b embryos from embryos as well asAnp32b three+/+ postnatal dayAnp32b 1 (P1)+/− pups ofAnp32b both the−/− late embryogenesis until early time points after birth. Among mixed 129:C57BL/6 and the pure C57BL/6 backgrounds revealed −/− litters examined immediately after birth, some Anp32b pups noExpected obvious gross abnormalities 36 in a majority 72 of embryos 36exam- never initiated breathing, whereas others were able to nurse, ined.Observed* Some Anp32b−/− 56E17.5 embryos did 87 show subtle and 1 spo- at least initially. Timed pregnancy studies revealed that E17.5 radic craniofacial+/− abnormalities, including overlarge ventricles in −/− Anp32b mice from congenic C57BL/6 background were intercrossed Anp32b embryos were present at approximately Mendelian Tabletheand brain, pups 3.3. wereThe expanded survival genotyped inner rate at of eartime Anp32b cavities, of weaning-deficient and (between palate mice depends closure postnatal defects.on 19 genetic and frequency (Table 3). No defect detectable by gross inspection was The22 d). heart“Expected and” lungsindicates were the not distribution obviously for the affected. indicated Interestingly, total number −/− background. evident in any E17.5 Anp32b embryos of the mixed back- largeof mice hematomas under conditions in the of liver,normal aortic Mendelian arch, segregation. or umbilical“Observed artery” −/− ground (Fig. 4 A and B). Examination of Anp32b embryos of wereindicates observed the actual in numbers three of of miceeight of E17.5 the indicated embryos genotypes and in obtained. one of +/- the pure C57BL/6 background showed that these animals also Anp32bthree*P < 0.01 P1 bymice pupsχ2 analysis. from of the congenic mixed C57BL/6 mutant background background. were However, intercrossed no and pups were genotypedone defect at appearedtime of weanin to beg (betwee responsiblen postnatal for the19 and perinatal 22 d). Expected lethality numbers are the of all Anp32b−/− mice. Reilly et al. distribution for the indicatedPNAS total| June number 21, 2011 of mice| vol. under 108 |conditionsno. 25 | of10245 normal Mendelian −/− segregation.Surviving Anp32b Observed Miceare the Show actual Reduced numbers Size of and mice Decreased of the indicated Life genotypes fi Fig. 3. Generation and validation of ANP32B-deficient mice. (A)Targeting Span. To investigate the effects of ANP32B de ciency in adult obtained. −/− of the murine Anp32b gene. (Top)MurinegenomicAnp32b locus showing mice, we monitored our surviving mixed-bred Anp32b mice exons and regions of homology (gray). FP: flanking probe detecting intron 1. and their littermates from time of weaning until age approxi- / H: HindIII; E: EagI sites. The translation start site (ATG) and stop sites are mately 1 y. At time of weaning, surviving Anp32b− − female mice shown. (Middle) Targeting construct that replaced exons 2–4withapgk-neo weighed, on average, 42% less than WT females and 35% less cassette. DTA: diphtheria toxin A sequence for negative selection. (Bottom) than their female Anp32b+/− littermates (P < 0.05; Fig. 4C). A Targeted Anp32b allele. Diagnostic HindIII fragments for the WT Anp32b similar trend was seen for males, although the differences did not allele (5.4 kb) and the targeted Anp32b allele (5.2 kb), as well as the di- reach statistical significance in this analysis (Fig. S2). When the agnostic EagI fragment for neo insertion (8.3 kb) are shown. (B)Confirma- +/ +/+ data for male and female sex-matched littermates were com- tion of Anp32b gene deletion. (Left)DNAfromAnp32b − and Anp32b ES fi fi bined, a statistically signi cant reduction in body weight was cells was Southern-blotted with a neo probe to con rm a single vector in- −/− +/− +/− −/− apparent when Anp32b and Anp32b littermates were sertion. (Right) DNA from MEFs derived from Anp32b , Anp32b ,or −/− +/+ Anp32b+/+ mice was Southern-blotted using the flanking probe. WT HindIII compared (P < 0.01), as well as when Anp32b and Anp32b band = 5.4 kb; mutant (MT) HindIII band = 5.2 kb. (C) Loss of ANP32B pro- littermates were compared (P < 0.05). −/− tein. Anp32b+/+, Anp32b+/− and Anp32b −/− MEFs were immunoblotted to Surviving Anp32b mice were not robust and showed signs / detect ANP32B protein. A and B are MEFs from two different Anp32b−/− of premature aging. Many Anp32b− − mutants exhibited early onset littermate embryos. β-Tubulin: loading control. Results shown in B and C are kyphosis in which the curvature of the upper spine was evident at 155 representative of three trials. 4 mo of age and severe by 6 mo (Fig. 4D). Increased hepatocyte

polyploidy, another hallmark of premature aging, was also present in the aged Anp32b−/− mice. However, premature alopecia, which −/− To define precisely when the majority of Anp32b mice died, is a common aspect of early aging, was not seen in the kyphotic GENETICS we monitored litters from the time of first detection (12 h mice. Consistent with these phenotypic features, Anp32b−/− mice −/− postnatal) until weaning. However, virtually all Anp32b mice had a reduced life span (P < 0.01; Fig. 4E), but no consistent that were evident at first detection survived to adulthood. This pathology could be identified as the cause of their premature observation indicated that the developmental problem(s) killing death. Two cases of megacolon and two cases of hydroureter most Anp32b−/− mice occurred either during embryogenesis or before normal litter detection. In the course of generating our primary MEFs, we noted that Anp32b−/− embryos were present at Table 2. Genetic background influence survival rate of Anp32b- approximately Mendelian rates on embryonic day 14.5 (E14.5). deficient mice −/− We therefore chose to follow the fate of Anp32b embryos from Anp32b+/+ Anp32b+/− Anp32b−/− late embryogenesis until early time points after birth. Among litters examined immediately after birth, some Anp32b−/− pups Expected 36 72 36 never initiated breathing, whereas others were able to nurse, Observed* 56 87 1 at least initially. Timed pregnancy studies revealed that E17.5 +/− −/− Anp32b mice from congenic C57BL/6 background were intercrossed Anp32b embryos were present at approximately Mendelian and pups were genotyped at time of weaning (between postnatal 19 and frequency (Table 3). No defect detectable by gross inspection was 22 d). “Expected” indicates the distribution for the indicated total number −/− evident in any E17.5 Anp32b embryos of the mixed back- of mice under conditions of normal Mendelian segregation. “Observed” −/− ground (Fig. 4 A and B). Examination of Anp32b embryos of indicates the actual numbers of mice of the indicated genotypes obtained. the pure C57BL/6 background showed that these animals also *P < 0.01 by χ2 analysis.

Reilly et al. PNAS | June 21, 2011 | vol. 108 | no. 25 | 10245

In order to define precisely when the majority of Anp32b-/- mice died, we monitored litters from the time of first detection (12 hrs postnatal) until weaning.

However, virtually all Anp32b-/- mice that were evident at first detection survived to adulthood. This observation indicated that the developmental problem(s) killing most

Anp32b-/- mice occurred either during embryogenesis or prior to normal litter detection.

In the course of generating our primary MEFs, we noted that Anp32b-/- embryos were present at approximately Mendelian rates on embryonic day 14.5 (E14.5). We therefore chose to follow the fate of Anp32b-/- embryos from late embryogenesis up until early time points after birth. Among litters examined immediately after birth, some Anp32b-/- pups never initiated breathing, whereas others were able to nurse, at least initially. Timed pregnancy studies revealed that E17.5 Anp32b-/- embryos were present at approximately

Mendelian frequency (Table 3.4). No defect detectable by gross inspection was evident in any E17.5 Anp32b-/- embryos of the mixed background (Fig. 2.5A, B). Examination of

Anp32b-/- embryos of the pure C57BL/6 background showed that these animals also remained viable up until at least E17.5. Taken together, our data suggest that the defect(s) killing Anp32b-/- mice occur perinatally, i.e. at or very near the time of birth.

156

Table 3. Normal survival rate of Anp32b-deficient embryos thymocytes and transformed MEFs to a broad array of intrinsic at E17.5 and extrinsic apoptotic stimuli. However, no statistically signifi- +/+ −/− Anp32b+/+ Anp32b+/− Anp32b−/− cant differences between Anp32b and Anp32b cells of either type were observed (Fig. S3). Similarly, prompted by the Expected 6 11 6 conclusions of previous reports (15, 17), we investigated whether Observed* 8 8 6 loss of ANP32B affected cell proliferation, but, again, no defects in in vitro proliferation were found for Anp32b−/− primary +/− Anp32b mice were intercrossed and embryos were genotyped at time MEFs, thymocytes, or splenocytes (Fig. S4). E17.5. “Expected” indicates the distribution for the indicated total number of Tableembryos 3.4. under Normal conditions survival of rate normal of Anp32b Mendelian-deficient segregation. embryos“ Observedat E17.5.” ANP32B Masks a Role for ANP32A in Essential Development. We have indicates the actual numbers of embryos of the indicated genotypes obtained. previously reported that neither the ANP32A-deficient nor the *No signi+/- ficant differences were found. Anp32b mice from mixed 129ola:C57BL/6 background were intercrossedANP32E-de in timed ficient mouse has any obvious abnormal phenotype matings and embryos were examined at E17.5. Expected numbers are the distribution(35). To for determine whether deficiency for another Anp32 family fi member might expose the cause of the viability defect of our thecausing indicated severe total morbiditynumber of embryos were observed, under conditions but the of other normal veMendelian aged segregation.−/ − +/− +/− Anp32b−/− mice died without identifiable cause. Anp32b mice, we generated Anp32b ;Anp32a and Observed are the actual numbers of embryos of the indicated genotypes obtained.Anp32b +/−;Anp32e+/− double-mutant strains and intercrossed them. As shown in Table 4, heterozygosity for Anp32a further Normal Cellular Apoptosis and Proliferation in the Absence of compromised development in the Anp32b−/− background (P < ANP32B. We next turned to in vitro studies in an attempt to −/− +/− −/− −/− identify the cause of the viability defects in Anp32b−/− mice. 0.01). Indeed, no Anp32b ;Anp32a or Anp32b ;Anp32a Because Anp32b has been implicated in the regulation of caspase mouse survived to weaning. Thus, contrary to prevailing belief, activity (10, 33), we first investigated the response of Anp32b−/− ANP32A does have a subtle role in murine embryogenesis that is revealed only in the absence of ANP32B. In contrast, loss of ANP32E did not further exacerbate the lethality associated with Anp32b deletion (Table 5). Although only A B one Anp32b−/−;Anp32e−/− mouse survived to weaning in this +/+ particular analysis, this rate was not statistically different from the survival rate expected for a mouse lacking only Anp32b−/− (Table 1). Furthermore, Anp32b−/−;Anp32e+/− mice exhibited a robust survival that stood in marked contrast to the fully penetrant lethality of Anp32b−/−;Anp32a+/− mice. When we attempted to -/- generate triple mutants lacking ANP32A, ANP32B, and ANP32E, none were viable (as expected). Strikingly, a single functional Anp32b allele was sufficient to allow survival to weaning age in the +/+ +/- -/- mixed-bred background (Table S1), but a single functional Anp32a or Anp32e allele could not support mouse survival in the C 18 D 16 absence of ANP32B. These data indicate that there is a hierarchy

14 of ANP32 protein functions in which ANP32B is the most im- 6monthsold ) 12 portant family member for embryogenesis, with ANP32A being of (g t

h 10 moderate importance and ANP32E being of least importance. g i 8 We Discussion 6 In this study, we provide evidence that ANP32B may be a useful 4 4 months old prognostic indicator in human breast cancer and describe the 2 generation and characterization of the ANP32B-deﬁcient mouse. 0 +/+ +/– +/– We carried out a meta-analysis of ANP32B mRNA levels in genotype +/+ -/- human157 breast cancers using three independent studies in which the patients were subject to different exclusion criteria and treat- E 100 ment regimens (37–39). Although ANP32B was not highlighted in 80 the original individual analyses, our examination of the combined

60 data indicates that elevated ANP32B expression correlates with shortened patient survival. We hypothesize that ANP32B does not Survival (%) 40 control tumorigenesis itself but rather increases, or is a marker of, 20 the robustness of the resulting tumor cells; that is, ANP32B is 0 17131925313743495561 elevated in tumor cells exhibiting increased proliferation and thus aggression. Consistent with this hypothesis, we found that Anp32b Age (weeks) mRNA was up-regulated in murine cells exhibiting enhanced Fig. 4. Phenotypes of ANP32B-deficient mice. (A and B)Normalgrossap- proliferation in tissue culture and in vivo. pearance. Embryos of the indicated Anp32b genotypes were collected at Our report identifies an ANP32 mutation with a strong ad- E17.5 and subjected to gross examination. (A) Rostral view. (B) Left side view. verse effect on development. We have shown that null mutation Tail fragments were removed for genotyping purposes. No obvious abnor- of Anp32b in mice results in highly penetrant perinatal lethality malities were seen. Embryos shown were derived from two different litters in a mixed genetic background and in fully penetrant lethality in and are representative of a total of 10 examined per genotype. (C)Reduced fi −/− +/+ a pure C57BL/6 background. The speci c cause of this lethality body weight. Anp32b mice were weighed alongside their Anp32b or remains unclear because surviving mixed-bred Anp32b−/− mice Anp32b+/− littermates at 3 wk of age. Results shown are values for individual suffered from a range of pathologies. We speculate that, during mice. Horizontal bars are mean values. *P < 0.05. (D) Kyphosis. X-ray analysis was performed on littermate pairs of female Anp32b+/+ and Anp32b−/− mice the course of evolution, the ANP32 proteins, and particularly at 4 and 6 mo of age. Results shown are representative of over 10 mice ANP32B, took on functions that were less dispensable for em- examined per genotype per age. (E) Decreased life span. Survival curves of bryogenesis. The nature of these functions remains obscure, as Anp32b+/+, Anp32b+/− and Anp32b−/− littermates that lived past weaning our data indicate that loss of ANP32B does not consistently af- age are shown. The reduced survival of Anp32b−/− mice is statistically signifi- fect one organ and that the majority of these animals die just cant, as determined by log-rank analysis (P < 0.01). after E17.5, when most murine organogenesis is complete.

10246 | www.pnas.org/cgi/doi/10.1073/pnas.1106211108 Reilly et al.

Rigorous pathological examinations of eight Anp32b-/- E17.5 embryos as well as three postnatal day 1 (P1) pups of both the mixed 129:C57BL/6 and pure C57BL/6 backgrounds revealed no obvious gross abnormalities in a majority of embryos examined.

Some Anp32b-/- E17.5 embryos did show subtle and sporadic craniofacial abnormalities, including overlarge ventricles in the brain, expanded inner ear cavities, and palate closure defects. The heart and lungs were not obviously affected. Interestingly, large hematomas in the liver, aortic arch or umbilical artery were observed in 3/8 E17.5 embryos and in 1/3

P1 pups of the mixed mutant background. However, no one defect appeared to be responsible for the perinatal lethality of all Anp32b-/- mice.

3.4.5 Surviving Anp32b-/- mice show reduced size and decreased lifespan

To investigate the effects of ANP32B deficiency in adult mice, we monitored our surviving mixed-bred Anp32b-/- mice and their littermates from time of weaning until age approximately one year. At time of weaning, surviving Anp32b-/- female mice weighed, on average, 42% less than WT females, and 35% less than their female Anp32b+/- littermates (p<0.05; Fig. 3.5C). A similar trend was seen for males, although the differences did not reach statistical significance in this analysis (Fig. 3.6). When the data for male and female sex-matched littermates were combined, a statistically significant reduction in body weight was apparent when Anp32b-/- and Anp32b+/- littermates were compared (p<0.01), as well as Anp32b-/- and Anp32b+/+ littermates (p<0.05).

Surviving Anp32b-/- mice were not robust and showed signs of premature aging.

Many Anp32b-/- mutants exhibited early-onset kyphosis in which the curvature of the upper spine was evident at 4 months of age and severe by 6 months (Fig. 3.5D). Increased

158

hepatocyte polyploidy, another hallmark of premature aging, was also present in the aged

Anp32b-/- mice. However, premature alopecia, which is a common aspect of early aging, was not seen in the kyphotic mice. Consistent with these phenotypic features, Anp32b-/- mice had a reduced lifespan (p<0.01; Fig. 3.5E) but no consistent pathology could be identified as the cause of their premature death. Two cases of megacolon and two cases of hydroureter causing severe morbidity were observed, but the other five aged Anp32b-/- mice died without identifiable cause.

159

Figure 3.5. Phenotypes of Anp32b-deficient mice.

160

Figure. 3.5. Phenotypes of Anp32b-deficient mice.

(A, B) Normal gross appearance. Embryos of the indicated Anp32b genotypes were collected at E17.5 and subjected to gross examination: (A) rostral view; (B) left side view. Tail fragments were removed for genotyping purposes. No obvious abnormalities were seen. Embryos shown were derived from two different litters and are representative of a total of 10 examined per genotype. (C) Reduced body weight. Anp32b-/- mice were weighed alongside their Anp32b+/+ or +/- littermates at 3 weeks of age. Results shown are values for individual mice. Horizontal bars are mean values. *p<0.05 (D) Kyphosis. X- ray analysis was performed on littermate pairs of female Anp32b+/+ and -/- mice at 4 months and 6 months of age. Results shown are representative of over 10 mice examined per genotype per age. (E) Decreased lifespan. Survival curves of Anp32b+/+, +/- and -/- littermates that lived past weaning age are shown. The reduced survival of Anp32b-/- mice is statistically significant, as determined by log-rank analysis (p<0.01).

161

Figure 3.6. Weights of Anp32b-deficient male mice.

162

Figure 3.6. Phenotypes of Anp32b-deficient mice.

Anp32b-/- male mice were weighed alongside their Anp32b+/+ or +/- littermates at 3 weeks of age. Results shown are values per individual mouse. Horizontal bars are mean values. Although no statistically significant difference was apparent for male mice in this analysis, when these data are combined with those for female Anp32b-/- mice, a statistically significant relationship between loss of Anp32b and reduced body weight became apparent.

163

3.4.6 Normal cellular apoptosis and proliferation in the absence of ANP32B

We next turned to ex vivo studies in an attempt to identify the cause of the viability defects in Anp32b-/- mice. Because Anp32b has been implicated in the regulation of caspase activity [164, 186], we first investigated the response of Anp32b-/- thymocytes and transformed MEFs to a broad array of intrinsic and extrinsic apoptotic stimuli.

However, no statistically significant differences between Anp32b+/+ and Anp32b-/- cells of either type were observed (Fig.3.7). Similarly, prompted by the conclusions of previous reports [169, 171], we investigated whether loss of ANP32B affected cell proliferation but, again, no defects in in vitro proliferation were found for Anp32b-/- primary MEFs, thymocytes or splenocytes (Fig.3.8).

164

Figure 3.7. No aberrant apoptotic response in Anp32b-deficient cells.

165

Figure 3.7. No aberrant apoptotic response in Anp32b-deficient cells.

Viability of (A) thymocytes and (B) transformed MEFs from Anp32b+/+ (black bars), Anp32b+/- (gray bars), or Anp32b-/- (open bars) mice as determined by propidium iodide exclusion. Cells were exposed to the following apoptotic stimuli for 18 hrs: 1.0 Gy γ- irradiation; 0.2 or 1.0 µM etoposide; 0.1 or 1.0 µM dexamethasone; 50 ηg/ml phorbol myristate acetate (PMA); 1 µg/ml anti-Fas antibody (clone Jo2, BD Biosciences) plus 0.01, 0.1 or 1.0 µg/ml cycloheximide (CHX); 1 µM ionomycin; 20 or 60 mJ/cm2 ultraviolet light; 1.0 or 5.0 µM cis-platin; culture for 20 hrs in 0.2% oxygen (hypoxia); or 10 µM staurosporine. Results shown are the mean ± SD (n=3/stimulus) and represent two independent trials. No significant differences were detected.

166

Figure 3.8. No proliferation defects in Anp32b-deficient cells.

(A) Normal T cell proliferation. Peripheral T cells purified from Anp32b+/+ (black bars) or Anp32b-/- (gray bars) mice were stimulated with plate-bound anti-CD3 antibody with or

167

without 0.1µg/ml anti-CD28 antibody. Proliferation was determined by [3H]-thymidine incorporation assays. Results shown are the mean ± SD (n=3). (B) Normal primary MEF proliferation. Primary MEFs were established from Anp32b+/+ (black symbols), Anp32b+/- (gray symbols), or Anp32b-/- (open symbols) mice and proliferation in culture was monitored for two days. Nt/No represents the cell number at a given time normalized to the cell number at day 0. Three experiments involving three independent MEF isolations are shown. Results shown for each data point are the mean ± SD of technical duplicates. For (A) and (B), no significant differences were determined.

168

3.4.7 ANP32B masks a role for ANP32A in essential development

We have previously reported that neither the ANP32A-deficient nor ANP32E- deficient mouse has any obvious abnormal phenotype [200]. To determine whether deficiency for another Anp32 family member might expose the cause of the viability defect of our Anp32b-/- mice, we generated Anp32b+/-;Anp32a+/- and Anp32b+/-;Anp32e+/- double mutant strains and intercrossed them. As shown in Table 3.5, heterozygosity for

Anp32a further compromised development in the Anp32b-/- background (p<0.01). Indeed, no Anp32b-/-;Anp32a+/- or Anp32b-/-;Anp32a-/- mouse survived to weaning. Thus, contrary to prevailing belief, ANP32A does have a subtle role in murine embryogenesis that is revealed only in the absence of ANP32B.

169

Table 4. Loss of Anp32b reveals a moderate viability defect attributable to Anp32a Anp32b+/+ Anp32b+/− Anp32b−/−

Anp32a+/+ Anp32a+/− Anp32a−/− Anp32a+/+ Anp32a+/− Anp32a−/− Anp32a+/+ Anp32a+/− Anp32a−/−

Mendelian 10 20 10 20 41 20 10 20 10 Expected 15 30 15 23 45 23 3 6 3 Observed* 15 25 20 28 43 23 8 0 0 Numbers of double-mutant mice of the indicated genotypes arising from the intercrossing of Anp32a+/−;Anp32b+/− mice. “Mendelian” indicates the distribution for the indicated total number of mice under conditions of normal Mendelian segregation. “Expected” indicates the distribution for the Mendelian distributed numbers with integration of the observed rates of Anp32b lethality (Table 1). “Observed” Tableindicates the3.5 actual. Loss numbers of Anp32b of mice of thereveals noted genotypes a moderate obtained. Statisticalviability probability defect is calculatedattributable against “Expected to Anp32a” ratios. . *P < 0.01 by χ2 analysis.

Our histologicalNumbers examinations of double mutant did not showmice any of defectsthe indicate in Materialsd genotypes and Methods arising from the intercrossing vascular endothelium+/ or- smooth muscle,+/- as might have been Prognostic Marker Identification. Data were acquired from three publicly expectedof from Anp32a previous reports;Anp32b on ANP32B mice. functions“Mendelian” (32, 40). numbersavailable are datasets the (37 distribution–39). The prognostic for riskthe associated indicated with particular However, our data regarding the incidence of hematomas in gene expression was computed using application of the Cox proportional / Anp32b−total− embryos number do support of mice a potential under role conditions for ANP32B of in normalhazard regression Mendelian model. segregation. “Expected” vascular development. Considering the recognized function of KLF5 innumbers vasculogenesis are the(32), distribution and the reported for theinfluence Mendel of ianNorthern distributed Blotting. The numbers multiple murine with tissue integration blot for Northern of analysis was ANP32B on KLF5-activated transcription, it is possible that the acquired from Zyagen (catalog no. MN-MT-2). Anp32b mRNA was detected hematomasthe may observed have been rates due of to defectsAnp32b in ANP32B-assisted,lethality (Table 3.2).by hybridizing “Observed” to a labeled are probe the consisting actual of thenumbers EcoRI–BamHI of fragment at KLF5-activated transcription. the 5′ end of the ORF. Gapdh was detected by probing with a 214-bp Our resultsmice con offirm the genetically indicated that there genotypes is considerable obtained. func- Statisticalfragment that probability hybridized to exons is 5, calculated 6, and 7. against tional overlap among ANP32 proteins. If different ANP32 proteins functioned in discrete cell types, or in different processes within the Real-Time RT-PCR. RNA was extracted from specified cell populations using same cell,“Expected” we would not haveratios. expected to see the synergistic effects astandardprotocolandtheRNeasykit(Qiagen),quantified, and reverse- that we observed in our compound mutants. However, our most transcribed using a SuperScript II first-strand synthesis kit (Invitrogen). cDNA dramatic finding is that mice carrying Anp32b mutations are sen- samples were then used as templates for quantitative real-time PCR (qPCR) sitized to loss of Anp32a, a gene that has not previously been using an ABI 7900HT detection system and SYBR Green (Applied Biosystems). shown to have an effect on mouse development. These data es- Data were normalized to Actb (β-actin), Tbp mRNA, or Rn18S (18S rRNA). fi tablish that, contrary to previous assumptions, ANP32 family Primer sequences are provided in Table S2. Statistical signi cance of differ- members are not all functionally redundant in vivo. Although ences in normalized values were assessed by Student’s t test. mutations of multiple alleles of Anp32a or Anp32e are compatible Mice. Mice were maintained under specified-pathogen free conditions in with embryogenesis, at least one functional Anp32b allele must be −/− present in any compound mutant for normal rates of survival. Thus, individually ventilated cages and fed a 5% irradiated meal. Anp32b mice ANP32B is clearly the most important member of the ANP32 analyzed in this study were derived from two separate homologous recombinant clones and analyzed in approximately equal proportion for all family, at least in mice. experiments. Unless otherwise stated, analyses were performed in the The broad array of functions ascribed to various ANP32 family mixed-bred 129ola;C57BL/6 genetic background. Statistical analyses for members in the literature suggests that certain reported bio- weights and longevity were performed using Student’s t test and log-rank chemical functions may be artifactual. The data we have pre- analysis, respectively. sented here, as well as in a previous publication (35), demonstrate that the Anp32 genes are not likely to be broadly involved in fi Plasmids and Primers. Sequences directing the expression of diphtheria toxin apoptosis. Furthermore, no ANP32-de cient cells have yet shown A in mouse ES cells were cloned into pBluescript (pgk-neo) to give the plasmid proliferation defects, implying that ANP32-mediated inhibition of pBSneoDTA (35). Regions of the Anp32b gene adjacent to the targeted cell cycle control phosphatases, such as PP2A, may be redundant exons were cloned by high-fidelity PCR. Primers used to clone the upstream or have a limited impact on cell proliferation. Instead, we spec- and downstream arms of homology into the XhoI site and XbaI site, re- ulate that ANP32 proteins play a more direct role in controlling spectively, are shown in Table S2. gene expression, likely through their reported activities in chro-

matin regulation and/or selective mRNA transport. In any case, Gene Targeting. Targeting constructs were linearized and transfected by GENETICS regardless of the ANP32 protein function under investigation, our electroporation into E14K mouse ES cells as previously described (41, 42). results indicate that future research aimed at understanding Southern probes used to detect Anp32b genomic sequences were ampliﬁed the shared functions of ANP32 proteins should concentrate from genomic ES cell DNA. PCR primer sequences for Anp32b ﬂanking probe on ANP32B. generation are presented in Table S2.

Table 5. Loss of Anp32b does not reveal any defect attributable to Anp32e Anp32b+/+ Anp32b+/− Anp32b−/−

Anp32e+/+ Anp32e+/− Anp32e−/− Anp32e+/+ Anp32e+/− Anp32e−/− Anp32e+/+ Anp32e+/− Anp32e−/−

Mendelian 6 11 6 11 22 11 6 11 6 Expected 8 16 8 12 25 12 2 3 2 Observed 9 13 10 14 21 15 3 2 1170

Number of double-mutant mice arising from the intercrossing of Anp32e+/−;Anp32b+/− mice. “Mendelian” indicates the distribution for the indicated total number of mice under conditions of normal Mendelian segregation. “Expected” indicates the distribution for the Mendelian distributed numbers with integration of the observed rates of Anp32b lethality (Table 1). “Observed” indicates the actual numbers of mice of the noted genotypes obtained. No statistically signiﬁcant deviation from “Expected” values was observed using χ2 analysis.

Reilly et al. PNAS | June 21, 2011 | vol. 108 | no. 25 | 10247

In contrast, loss of Anp32e did not further exacerbate the lethality associated with

Anp32b deletion (Table 3.6). Although only one Anp32b-/-;Anp32e-/- mouse survived to weaning in this particular analysis, this rate was not statistically different from the survival rate expected for a mouse lacking only Anp32b-/- (Table 3.2). Furthermore,

Anp32b-/-;Anp32e+/- mice exhibited a robust survival that stood in marked contrast to the fully-penetrant lethality of Anp32b-/-;Anp32a+/- mice. When we attempted to generate triple mutants lacking ANP32A, ANP32B and ANP32E, none were viable (as expected).

Strikingly, a single functional Anp32b allele was sufficient to allow survival to weaning age in the mixed-bred background (Table 3.7), but a single functional Anp32a or Anp32e allele could not support mouse survival in the absence of ANP32B. These data indicate that there is a hierarchy of ANP32 protein functions in which ANP32B is the most important family member for embryogenesis, with ANP32A being of moderate importance, and ANP32E being of the least importance.

171

Table 3.6. Loss of Anp32b does not reveal any defect attributable to Anp32e.

Number of double-mutant mice arising from the intercrossing of Anp32e+/−;Anp32b+/− mice. “Mendelian” indicates the distribution for the indicated total number of mice under conditions of normal Mendelian segregation. “Expected” indicates the distribution for the Mendelian distributed numbers with integration of the observed rates ofAnp32b lethality (Table 3.2). “Observed” indicates the actual numbers of mice of the noted genotypes obtained. No statistically significant deviation from “Expected” values was observed using χ2 analysis.

172

Table 3.7: A single functional Anp32b allele is sufficient for mouse survival.

Numbers of triple mutant mice of the indicated genotypes arising from the intercrossing of triply heterozygous Anp32a+/-;Anp32b+/-;Anp32e+/- mice. “Mendelian” numbers are the distribution for the indicated total number of mice under conditions of normal Mendelian segregation. “Expected” numbers are the distribution for the Mendelian distributed numbers with integration of the observed rates of Anp32b lethality (Table 3.2). “Observed” are the actual numbers of mice of the indicated genotype

173

3.5 Discussion

In this study, we provide evidence that ANP32B may be a useful prognostic indicator in human breast cancer, and describe the generation and characterization of the ANP32B- deficient mouse.

We carried out a meta-analysis of ANP32B mRNA levels in human breast cancers using three independent studies in which the patients were subject to different exclusion criteria and treatment regimens [212-214]. Although ANP32B was not highlighted in the original individual analyses, our examination of the combined data indicates that elevated

ANP32B expression correlates with shortened patient survival. We hypothesize that

ANP32B does not control tumorigenesis itself but rather increases, or is a marker of, the robustness of the resulting tumor cells; that is, ANP32B is elevated in tumor cells exhibiting increased proliferation and thus aggression. Consistent with this hypothesis, we found that Anp32b mRNA and protein were upregulated in murine cells exhibiting enhanced proliferation in tissue culture and in vivo.

Our report identifies an ANP32 mutation with a strong adverse effect on development. We have shown that null mutation of Anp32b in mice results in highly penetrant perinatal lethality in a mixed genetic background, and fully penetrant lethality in a pure C57BL/6 background. The specific cause of this lethality remains unclear since surviving mixed-bred Anp32b-/- mice suffered from a range of pathologies. We speculate that during the course of evolution, the ANP32 proteins, and particularly ANP32B, took on functions that were less dispensable for embryogenesis. The nature of these functions remains obscure, as our data indicate that loss of ANP32B does not consistently affect

175

one organ, and that the majority of these animals die just after E17.5, when most murine organogenesis is complete.

Our histological examinations did not show any defects in vascular endothelium or smooth muscle, as might have been expected from previous reports on ANP32B functions [185, 219]. However, our data regarding the incidence of hematomas in

Anp32b-/- embryos do support a potential role for ANP32B in vascular development.

Considering the recognized function of KLF5 in vasculogenesis [185], and the reported influence of ANP32B on KLF5-activated transcription, it is possible that the hematomas may have been due to defects in ANP32B-assisted KLF5-activated transcription.

Our results confirm genetically that there is considerable functional overlap among ANP32 proteins. If different ANP32 proteins functioned in discrete cell types, or in different processes within the same cell, we would not have expected to see the synergistic effects we observed in our compound mutants. However, our most dramatic finding is that mice carrying Anp32b mutations are sensitized to loss of Anp32a, a gene that has not previously been shown to have an effect on mouse development. These data establish that, contrary to previous assumptions, ANP32 family members are not all functionally redundant in vivo. Although mutations of multiple alleles of Anp32a or

Anp32e are compatible with embryogenesis, at least one functional Anp32b allele must be present in any compound mutant for normal rates of survival. Thus, ANP32B is clearly the most important member of the ANP32 family, at least in mice.

The broad array of functions ascribed to various ANP32 family members in the literature suggests that certain reported biochemical functions may be artefactual. The

176

data we have presented here, as well as in a previous publication [200], demonstrate that the Anp32 genes are not likely to be broadly involved in apoptosis. Furthermore, no

ANP32-deficient cells have yet shown proliferation defects, implying that ANP32- mediated inhibition of cell cycle control phosphatases, such as PP2A, may be redundant or have a limited impact on cell proliferation. Instead, we speculate that ANP32 proteins play a more direct role in controlling gene expression of such genes as rpS6 likely through their reported activities in chromatin regulation and/or selective mRNA transport.

In any case, regardless of the ANP32 protein function under investigation, our results indicate that future research aimed at understanding the shared functions of ANP32 proteins should concentrate on ANP32B.

177

Chapter 4

ANP32B in the immune system

178

4 Investigating the Role of ANP32B in the Immune System 4.1 Abstract

The acidic nuclear phosphoprotein 32B (ANP32B) belongs to a highly conserved and broadly expressed family of ANP32 proteins that have been implicated in many cellular functions. The biochemical activities ascribed to ANP32B, namely post- transcriptional regulation of early response genes (ERGs), repression of Krüppel-like

Factor 5 (KLF-5) dependent gene expression, association with cell cycle progression in rat T and B cells, as well as its high expression in cells of the immune system, suggest means by which this gene could potentially influence the immune system.

In this chapter, we used the mixed bred ((Ola129:C57BL/6J) and backcrossed

(C57BL/6J) ANP32B-deficient mice (Anp32b-/- mice) to elucidate the precise biological function(s) of ANP32B in the immune system in vivo. Phenotypic analysis of the mixed bred mice revealed a largely normal immune system. By contrast, results from the backcrossed Anp32b-/- mice suggested that ANP32B is necessary for proper development of B cells and homeostasis of T cells in the periphery. Regardless of genetic background, however, ANP32B expression appeared to be necessary for normal differentiation of B cells in the periphery.

T cells from the mixed bred Anp32b-/- mice exhibited decreased levels of total ribosomal protein S6 (rpS6) at resting levels and showed delayed kinetics in its induction following activation. Moreover, dendritic cells (DCs) from Anp32b-/- mice failed to become fully activated in response to stimuli. To our knowledge, these are novel findings implicating ANP32B in the development, differentiation, and/or activation of various

179

immune cell subsets in vivo. The mechanism and the functional consequences of the observed defects in Anp32b-/- mice remain to be elucidated.

180

4.2 Introduction

4.2.1 ANP32B is a member of a conserved protein family.

The Acidic Nuclear Phosphoprotein 32 (ANP32) family of proteins are highly conserved and broadly expressed gene products that are distinguished by N-terminal leucine-rich repeats (LRRs) and C-terminal acidic tails [155]. The family consists of three different gene products: ANP32A (also known as pp32, I1PP2A, PHAPIa or LANP), ANP32B

(also known as PAL31, PHAPIb or APRIL), and ANP32E (also known as CPD1,

PHAPIII or LANPL) [156, 157, 220]. The importance of these genes in biology lies in the existence of their homologues in all metazoan species as well as their sequence conservation throughout vertebrate evolution [156, 157]. The aberrant expression of these proteins, including a highly significant correlation between ANP32B mRNA expression and poor breast cancer prognosis, suggests that ANP32 family plays a role in cancer progression [198]. In addition, the biological activities ascribed to this family, namely regulation of apoptosis [162-168], chromatin regulation [179-182], and post- transcriptional gene expression [176, 187-189], suggest means by which these proteins could influence tumorigenesis. However, it should be noted that the unusual amino acid composition of these proteins makes them prone to misleading purification and potentially artefactual results. Therefore, the precise non-redundant biochemical and cellular functions of the ANP32 proteins can best be determined using a genetic approach.

To that end, our laboratory generated gene-targeted Anp32a, Anp32b and Anp32e- deficient mice and obtained data indicating that Anp32b is the most important member of

181

the family with respect to mammalian development [198]. Unlike mice deficient for either ANP32A or ANP32E, animals lacking ANP32B showed a partially penetrant perinatal lethality in the mixed background. Although at the gestational stage, E17.5,

Anp32b-/- embryos exhibited normal weights and expected Mendelian ratios, they were both significantly reduced at the time of weaning. Moreover, surviving adult Anp32b-/- mice demonstrated symptoms of accelerated aging as well as dysfunctions in various organ systems. When these mice were backcrossed six generations onto the C57BL/6J strain, they demonstrated fully penetrant embryonic lethality [198]. Initial histopathological analysis of these embryos revealed consistent and severe craniofacial defects as well as abnormalities in eyelid closure and ear canals. As ANP32B appears to be the most important member for murine development, the focus of this chapter will be on elucidating the in vivo role of this gene.

ANP32B likely plays crucial roles in many different organ systems and cell types.

However, the immune system is perhaps best suited for this analysis as it provides a well- studied system for examining the function of any gene in cell survival, development, and differentiation. Hence, this chapter will outline several investigations into the immune system of the Anp32b null mice in an effort to elucidate the true biological and non- redundant functions of ANP32B in cell biology.

182

4.3 Materials and Methods

4.3.1 Mice

Mice were maintained under specified pathogen-free conditions in individually ventilated cages and fed 5% irradiated meal. Anp32b-/- mice analyzed in this study were derived from two separate homologous recombinant clones and analyzed in approximately equal proportion for all experiments. Unless otherwise stated, analyses were performed in the mixed-bred 129ola; C57BL/6 genetic background.

4.3.2 Immunoblot Analysis

Standard western blot analysis was used to analyse protein expression, using the Odyssey

LICOR Infrared Imaging System (Biosciences) and the Odyssey LICOR software [198].

4.3.3 Flow Cytometry

Single cell suspensions were prepared from thymus, spleen or lymph nodes. The latter two were treated with red blood lysis buffer to remove RBCs. Cells (1–2 × 106) were pre- incubated with Fc block for 15 min at 4°C and stained with antibodies recognizing the following antigens: CD16/CD32 (2.4G2), CD45R/B220 (RA3-6B2), CD4, CD8a, CD3,

NK1.1, CD11c, CD11b, GR-1, CD80, CD86, and CD83 (all obtained from BD or eBioscience unless otherwise specified). Flow cytometry data were acquired using either a FACSCalibur (BD) or FACSCanto (BD) flow cytometer, and analyzed with either the

CellQuest Pro software (BD) or the FlowJo analysis program (Tree Star).

4.3.4 T Cell Purification and Activation

Single cell suspensions derived from spleen or lymph nodes were treated with red blood cell lysis buffer (Sigma) to remove erythrocytes. B and T cells were purified using the

183

IMag cell separation system (BD). Briefly, total leukocytes were incubated with mouse anti-CD16/32 blocking antibodies, after which T cells were negatively selected using biotinylated anti-Ter119/B220/CD19/CD11b/Nk1.1/CD11c antibodies. For naïve T cell isolation, biotin anti-CD44 antibody was also added. Purified T cells were stimulated with plate-bound Rabbit anti-hamster, anti-CD3 (clone 2C11; BD Biosciences) plus anti-

CD28 antibodies (clone 37.51; BD Biosciences).

4.3.5 Fetal Liver Chimeras

E14.5 wild type and Anp32b-/- fetal liver cells (generated using plugged matings) were harvested and homogenized into single cell suspensions. Recipient mice (Rag-2-/-) were irradiated with 6Gy gamma irradiation. 3-5x106 wild type and Anp32b-/- fetal liver cells were transferred into Rag host intravenously. Chimeric mice were analyzed 6-16 weeks after reconstitution. Reconstitution was monitored by peripheral blood.

4.3.6 Generation of murine BMDCs

BMDCs were generated as previously described [221]. Activation was induced by LPS treatment (1mg/ml) for 24 hours. Cells were analyzed by flow cytometry.

4.3.7 Statistical Analysis

Where appropriate, all differences were evaluated using unpaired, 2-tailed Student’s t test, as calculated using GraphPad Prism. Data are presented as mean ± SEM.

Statistically significant differences are indicated as *, p< 0.05, **, p<0.01 or ***, p<0.001.

184

4.4 Results

4.4.1 ANP32B protein levels are linked to proliferative tissues in mice

Several studies have reported high mRNA expression of Anp32b in proliferative tissues

[169, 171, 198]. To determine if these findings are valid at the protein level, we performed immunoblot analysis (Fig.3.1) on a wide range of tissues from adult wild type

(WT) mice. We found that ANP32B protein expression was low (relative to β-tubulin) in tissues where less cell proliferation usually occurs, i.e. in brain (Fig. 4.1). In contrast, tissues that generally have high cell proliferation rates, i.e. spleen and thymus, showed high levels of ANP32B protein expression. Intriguingly, the spleen sample exhibited a bright band that was half the expected molecular weight of ANP32B (36kDa), indicating a potential cleavage of ANP32B protein. Taken together, our data further suggest a link between ANP32B expression level and rate of cell proliferation. They also provide a strong basis for investigating the role of ANP32B in immune cells as they exhibited the highest expression of ANP32B protein.

185

ANP32B)is)highly)expressed)in)lymphoid)organs.)

4 )tubulin0

ϐ 3

1 ANP32B/ 0 Brain Thymus Heart Lungs Liver Spleen Kidney Testis

36 ANP32B

β-tubulin

Which)immune)subsets)are)aﬀected)by)the)loss)of)this)gene?)Figure 4.1. ANP32B protein expression correlates with proliferative tissues.

Cells from the indicated tissues were lysed and analyzed by immunoblot for ANP32B and β-tubulin expression (lower panel). ANP32B expression was quantified by Odyssey Li- Cor instrument and normalized to β-tubulin (upper panel).

186

4.4.2 ANP32B protein appears to be cleaved in splenocytes

We next investigated the hypothesis that ANP32B is selectively cleaved in splenocytes.

To do so, we compared the protein expression of ANP32B in WT and ANP32B deficient splenocytes and thymocytes. Whereas the thymus band for ANP32B appeared at the expected molecular weight (36kDa), the band for spleen once again appeared much lower

(18kDa) (Fig. 4.2). Moreover, this band was specific to ANP32B as its expression was absent in ANP32B deficient splenocytes. Therefore, ANP32B appears to be selectively cleaved in splenocytes.

187

Figure 4.2. Immunoblot analysis of ANP32B expression in thymus and spleen of WT and Anp32b-/- deficient mice.

188

4.4.3 Anp32b-/- mice exhibit reduced splenic cellularity

The high expression of ANP32B in immune tissues and its potential cleavage in splenocytes led us to investigate the function of ANP32B in the immune system. To get a broad overview of hematopoiesis, we performed hematology analysis on a set of WT and mixed bred ANP32B deficient mice. Gross hematological examination demonstrated no differences in white blood cells or erythrocyte counts (Fig. 4.3A), suggesting that

ANP32B deficiency does not result in any major hematological abnormalities. We next examined the effect of ANP32B loss on the size of primary and secondary lymphoid organs. Specifically, cellularity of bone marrow (BM), thymus and spleens of the

Anp32b-/- mice was compared with their wild type counterparts. While BM and thymus cellularity were similar, splenic size was significantly reduced in animals lacking

ANP32B (Fig. 4.3B), suggestive of the involvement of ANP32B in immune cell development and/or homeostasis.

189

Figure 4.3. Gross analysis of Anp32b-/- immune system.

A) Hematology analysis of WT and mixed bred ANP32B deficient mice. Blood was collected in EDTA-coated tubes and used to do the following hematological analysis: white blood cells (WBC), red blood cells (RBC), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC). Results are mean+/- SEM and are representative of 6 WT and 9 KO mice. (B) Absolute cellularity of bone marrow (BM), thymus, and spleen in control (blue bars) and mixed Anp32b-/- (red bars) mice. Results are representative of 5-7 independent experiments involving one to four mice per genotype. Error bars are mean +/- SEM.

190

4.4.4 Mixed bred Anp32b-/- mice display a normal immune system

The reduced splenic cellularity prompted us to examine the immune system of Anp32b-/- mice to define the subset of immune cells most affected by the loss of this gene. We carried out a thorough flow-cytometric analysis to compare lymphoid and myeloid cell populations from the bone marrow, thymus, spleen and lymph nodes (LNs) of WT and mixed bred Anp32b-/- littermate mice. As shown in figure 4.4A, no significant abnormalities were observed in T cell development, maturation or peripheral homeostasis

(Fig. 4.4A, Table 4.1). Similarly, the development and maturation of B cells was intact in the absence of ANP32B (Fig. 4.4B,C, Table 4.1). Likewise, the development and homeostasis of myeloid cells was normal in the absence of ANP32B (Fig. 4.4D, Table

4.1). Intriguingly, B cell differentiation appeared to be the only investigated aspect of immune system that was significantly altered in the absence of ANP32B. Knockout mice exhibited increased proportion of marginal zone (MZ) B cells, as characterized by B220+,

CD23lo, CD21hi and decreased proportion of follicular zone (FO) B cells, as characterized by B220+, CD21lo, CD23hi), indicating a potential role for ANP32B in B cell differentiation (Table 4.1). Overall, these results from mixed genetic background mice imply that ANP32B expression may be dispensable for the development and homeostasis of major immune cell subsets but appears to modulate the proportions of B cell subsets in the periphery.

191

A# THY# SPL# LN#

WT#

KO# CD8$ CD4$

B# THY# SPL# LN#

WT#

KO# B220$ CD3$

192

C# BM# SPL#

WT#

KO# $ IgD IgM$

D# BM# SPL#

WT#

KO# CD11b$ GR11$

Figure 4.4. Anp32b-/- mice display a largely normal immune system.

193

Figure 4.4. Anp32b-/- mice display a largely normal immune system.

(A)Flow cytometric analysis of CD4 and CD8 T cells from thymus, spleen, and lymph node (LN) of littermate control (top panel) and ANP32B deficient mice (lower panel). Numbers shown are percentage of live lymphocytes. (B) Flow cytometric analysis of B cell populations from bone marrow, spleen and lymph nodes of littermate control (upper panel) and ANP32B deficient mice (lower panel). Numbers shown are percentage of live lymphocytes. (C) Flow cytometric profile of B cell development. WT and Anp32b deficient bone marrow and splenocytes were stained with B220 and IgD and IgM antibodies to distinguish developing and mature B cells. Numbers shown are percentage of live B220+ cells. (D) Flow cytometric analysis of myeloid cell populations from bone marrow and spleen of littermate control (upper panel) and Anp32b deficient mice (lower panel). Numbers shown are percentage of live cells. Results are representative of 7 wild type and 7 Anp32b-/-mice.

194

GaCng0 +/+0mean0 −/−0mean0 +/+0SD0 +/)0SD0 −/−0SD0 +/+,0n0 +/),0n0 −/−,0n0 T#test' THY0 CD4/8) 81) 80) 4.7) 2.1) 8.4) 6) 2) 7) 0.405) CD4) 11.2) 10.5) 3.2) 2.1) 1.4) 6) 2) 7) 0.31) CD8) 4.5) 2.9) 2.7) 0.5) 0.8) 6) 2) 7) 0.08) SPL0 CD4) 23.8) 23.3) 7.1) 0.2) 4.2) 6) 2) 7) 0.444) CD8) 12.4) 10.9) 1.8) 0.8) 1.5) 6) 2) 7) 0.062) B220) 38) 32.9) 8.7) 9.7) 7.3) 7) 2) 8) 0.118) THY0 CD4:CD8) 2.8) 3.9) 1) 1.5) 1.4) 6) 2) 7) 0.081) SPL0 CD4:CD8) 1.9) 2.2) 0.4) 0.1) 0.4) 6) 2) 7) 0.126) SPL0B220+0 MZ)CD21+) 6.4) 11.1) 3.1) 1.1) 5.1) 7) 2) 8) 0.03) FO)CD23+) 72.1) 60.9) 8.8) 6) 11.8) 7) 2) 8) 0.028) CD21JCD23J) 19.5) 25.9) 6.9) 4.2) 7.3) 7) 2) 8) 0.053) SPL0CD25+0 CD4) 8.9) 10.6) 1.5) 1.3) 2.1) 3) 2) 4) 0.147) SPL0CD69+0 CD4) 10.3) 10.9) 1.7) 0.7) 1.2) 3) 2) 4) 0.289) SPL0memory0 CD4) 21) 29.4) 6.8) 10) 14.1) 4) 2) 5) 0.156) SPL0B220+0 IgM)IgD) 57.3) 62.1) 25) 8.2) 17.2) 4) 2) 5) 0.302) GRJ1)+) BM0PMN0 CD11b+) 36.1) 39) 17.9) 6.8) 14.1) 5) 3) 6) 0.385) GRJ1)+) SPL0PMN0 CD11b+) 8.1) 9.2) 13.8) 11.1) 8.5) 4) 3) 5) 0.443)

Table 4.1. Analysis of immune system cell populations in wildtype and Anp32b-/- littermate mice.

Flow cytometric analysis was used to compare thymocyte, splenocyte, and bone marrow cell populations of 7 wild type and 7 Anp32b-/- mice. Yellow cells highlight populations with a statistical significance between the wild type and Anp32b-/- mice. SD= standard deviation.

195

4.4.5 Backcrossed ANP32B deficient mice show defects in B/T cell

development and homeostasis

Because a pure genetic background is preferred for immunological studies, it became necessary for us to examine the immune system of the backcrossed Anp32b-/- mice.

However, ANP32B deficient mice backcrossed onto the C57BL/6 background demonstrate a nearly complete penetrant lethality [198]. Incidentally, however, one backcrossed Anp32b-/- pup was born and survived to adulthood, allowing us to analyze the immune system of the backcrossed ANP32B deficient animals. Similar to data obtained from mixed bred mice, we observed normal T cell development in thymus of this mouse (Fig. 4.5, left panel). However, in contrast to the mixed bred mice, the proportion of CD4+ and CD8+T cells in the periphery appeared to be reduced in comparison to the littermate control animals (Fig. 4.5 middle panel). Similarly, peripheral

B cell numbers were also reduced in the periphery (Fig.4.5 right panel). In addition, B cell development appeared to be greatly altered in the absence of ANP32B (Fig. 4.6).

Taken together, contrary to the results from the mixed bred mice, these data from a single backcrossed mouse suggest an important role of ANP32B in B cell development and T cell homeostasis.

197

Figure 4.5. Dysregulated B/T cell development and homeostasis in backcrossed Anp32b-/-mice.

Flow cytometric analysis of CD4 and CD8 T cells from thymus and spleen of control (top panel) and Anp32b deficient mouse (lower panel). Numbers shown are percentage of live lymphocytes. (Right panel) Flow cytometric analysis of B cell populations from spleen of control (upper panel) and Anp32b deficient mouse (lower panel). Numbers shown are percentage of live lymphocytes. Numbers shown are percentage of live cells. Results are representative of 1 WT and 1 Anp32b-/- mouse.

198

Figure 4.6. Flow cytometric analysis of B cell development in the bone marrow.

B cell development in control (top panel) and ANP32B deficient mouse (lower panel). Numbers shown are percentage of live lymphocytes and results are representative of 1 WT and 1 Anp32b-/- mouse. Plots are gated on indicated populations.

4.4.6 ANP32B is indispensable for B/T cell development in fetal liver chimeras

The fully penetrant lethality in the backcrossed Anp32b-/- mice limited our ability to properly analyze the immune system of backcrossed mice. Fetal liver chimeras provide optimal means of studying the immune systems of embryonic lethal mice. Therefore, we generated chimeras to study the immune system of backcrossed Anp32b-/- embryos. E14.5 fetal liver cells from WT and Anp32b-/- embryos were used to reconstitute sub-lethally

199

irradiated B and T cell deficient RAG-2-/- mice and reconstitution was examined two months later. Recipient mice reconstituted with Anp32b-/- fetal liver cells demonstrated profound defects in B and T lymphocyte development and homeostasis in the periphery

(Fig.4.7). These findings are consistent with our analysis of the single backcrossed animal and therefore our data collectively suggest an important role for ANP32B in B/T cell development and homeostasis.

Figure 4.7. Dysregulated B/T cell homeostasis in fetal liver chimeras.

Flow cytometric analysis of CD4 and CD8 T cells from thymus and spleen of control (top panel) and Anp32b deficient fetal liver chimeras (lower panel). Numbers shown are percentage of live lymphocytes. (Right panel) Flow cytometric analysis of B cell populations from spleen of control (upper panel) and ANP32B deficient fetal liver chimeras (lower panel). Numbers shown are percentage of live lymphocytes. Numbers

200

shown are percentage of live cells. Results are representative of 5 WT and 3 KO chimeras.

4.4.7 B cell differentiation is dysregulated in Anp32b-/- mice

The most statistically significant alteration that we observed in the immune analysis of the mixed bred ANP32B deficient mice was differentiation of B cells. Anp32b-/- mice exhibited a significantly increased proportion of MZ B cells and a concomitantly proportional decrease in FO B cells (Fig. 4.8). Because knockout mice also demonstrated a decreased splenic cellularity compared to the WT mice, this suggests that the differences in percentages of immune cell populations are reflective of even greater differences in total MZ and FO B cell numbers.

We also examined the single backcrossed mouse and the fetal liver chimeras for similar defects in B cell differentiation. Consistent with the mixed background data, we observed an increased proportion of MZ B cells and decreased proportion of FO cells in the absence of ANP32B in the backcrossed mouse as well as fetal liver chimeras (Fig.

4.8A). Interestingly, the defect appeared to be enhanced with the backcrossing of the mice (Fig. 4.8B). Taken together, ANP32B appears to be involved in B cell fate decisions, regulating the proportions of the different B cell subsets in the periphery.

Importantly, this phenomenon was consistently observed in all ANP32B-deficient mice, independent of genetic background.

201

Figure 4.8. Mixed bred and backcrossed Anp32b-/- mice exhibit increased MZ and decreased FO B cells populations.

(A) Splenocytes from wild type and Anp32b-/- mice were used to analyze B cell differentiation by flow cytometry. Analyses were performed after gating on B220+ B cells and then comparing the expression of CD21 and CD23. Graph depicts the cumulative data of B cell differentiation in the wild type (blue) and mixed mutant mice (red). A total of 7 mice were used in each group and statistical significance is indicated by asterisks (*) p<0.05. (B) B cell differentiation in backcrossed mice (left panel) and fetal liver chimeras (right panel). Cells are gated on B220+ population.

202

4.4.8 Anp32b regulates rpS6 expression in resting and activated T cells.

The rpS6 kinase 1 (S6K1) and rpS6 kinase 2 (S6K2) double knockout mice present an overall similar physical phenotype as our Anp32b-/- mice [198]. This observation and our original hypothesis that Anp32b likely plays an important role in cellular proliferation prompted us to assess the effect of the loss of this gene on the activation of the PI3K pathway, which is primarily responsible for activation of S6Ks. Hence, we compared T cell activation in wild type and Anp32b-/- mice after ex vivo anti-CD3 and anti-CD28 stimulation. S6K1 phosphorylation was identical in WT and Anp32b mutant mice (Fig.

4.9). However, when the activation of a bona fide substrate of S6K, rpS6, was checked, there was a marked reduction in phosphorylation of rpS6 after 30 minutes of T cell activation (Fig. 4.9). Total rpS6 was also significantly reduced in Anp32b-/- T cells at all time points. Even at steady state levels, ANP32B deficient T cells contained less rpS6 protein, implying that the defect in rpS6 phosphorylation was not due to defective upstream kinase activation but likely due to other factors affecting total rpS6 expression in the knockout T cells. Experiments with prolonged anti-CD3/CD28 stimulation showed that rpS6 phosphorylation defect is only present during the early phases of T cell activation (Fig. 4.10). After overnight stimulation, Anp32b-/- T cells exhibit similar levels of phospho-rpS6 and after 24 hours, they even exceed the WT phospho-rpS6 expression, suggesting a case of delayed kinetics of rpS6 induction. In essence, our data seems to suggest that rpS6 expression is regulated differentially during early and late phases of T cell activation and hints that ANP32B may be an early regulator of rpS6 expression.

203

ANP32B+/- ANP32B-/- α-CD3/CD28 (min) 0 5 30 60 0 5 30 60

p-S6K1 (Thr389)

p-rpS6 (Ser 235/236)

Total rpS6

actin

p-S6K1 (Thr421/Ser424)

p-rpS6 Ser240/244

actin

Figure 4.9. rpS6 expression is dysregulated in resting and activated T-lymphocytes.

Fig. 9: rpS6 expressionSplenic is dysregulatedand lymph node in resting T cells and were activated purified T-lymphocytes. using magnetic Splenic enrichmentand lymph node beads T cells and were purified using magnetic enrichment beads and subjected to activation by plate bound anti-CD3 and anti-CD28 antibodies for the indicated timepoints. Lysates were used for western blot analysessubjected using to theactivation indicated byantibodies. plate bound These antiblots- CD3are representative and anti-CD28 of data antibodies obtained from for three the independent experiments. indicated time points. Lysates were used for western blot analyses using the indicated antibodies. These blots are representative of data obtained from three independent experiments.

204

(A) (B)

ANP32B+/- ANP32B-/-

CD3+CD28 (time) 0 5 30 60 o/n 0 5 30 60 o/n

p-rpS6 (Ser235/236)

p-S6K1(Thr421/Ser 424)

p-S6K1(Thr 389)

p-4EBP1 (Thr 35/47)

p-Erk (Thr 202/204)

p-p38 (Thr180/Tyr 182)

actin

Figure 4.10. rpS6 is induced with a delayed kinetics in Anp32b -/- T cells after activation.

Fig. 10. rpS6 is induced with a delayed kinetics in ANP32B -/- T cells after activation. (A) Splenic and lymph node T cells were purified using magnetic(A) Splenic enrichment and beads lymph and subjected node to activation T cells by plateboundwere purified anti-CD3 and using anti-CD28 magnetic antibodies for enrichment the indicated timepoints. beads Lysates and were used for western blot analyses using various antibodies. (B) The graph is a quantification of rpS6 phosphorylation normalized to actin in each lane.subjected to activation by plate bound anti-CD3 and anti-CD28 antibodies for the indicated time points. Lysates were used for western blot analyses using various antibodies. Results shown are representative of two independent experiments. (B) The graph is a quantification of rpS6 phosphorylation normalized to β-actin in each lane.

205

4.4.9 ANP32A and ANP32E do not regulate the expression of phospho-rpS6 in T cells

Given the functional redundancy between ANP32 proteins, we wondered if ANP32A and

ANP32E deficient T cell exhibit similar defects in rpS6 phosphorylation in T cells. To investigate, we purified T cells from ANP32A and ANP32E deficient animals and examined the expression of phospho-rpS6 after 30 minutes of anti-CD3/CD28 stimulation. The dynamics of phospho-rpS6 appeared to be similar or enhanced compared to littermate controls suggesting that ANP32B exclusively regulates rpS6 expression in T cells (Fig. 4.11).

206

Figure 4.11. Phospho-rpS6 expression in ANP32A and ANP32E deficient T cells.

Splenic and lymph node T cells were purified using magnetic enrichment beads and subjected to activation by plate bound anti-CD3 and anti-CD28 antibodies for 30 minutes. Lysates were used for immunoblot analyses using the indicated antibodies. The graph in the lower panel is quantification of rpS6 phosphorylation in Anp32a-/- T cells (normalized to β-actin in each lane). Blot is representative of a single experiment.

207

4.4.10 Reduced activation of Anp32b null dendritic cells ex vivo.

A recent report has suggested that Anp32b regulates the expression of the dendritic cell

(DC) activation marker CD83 through control of mRNA transport and stability [176]. To determine whether Anp32b-/- mice show altered CD83 expression and DC function, we established ex vivo cultures of DCs obtained from WT and Anp32b-/- mice. We found that

DCs from Anp32b-/- mice display reduced activation in response to lipopolysaccharide

(LPS) stimulation, as determined by staining for the DC activation markers CD40, CD80 and CD83. Figure 4.12 shows the flow cytometric profile of one such experiment, where

Anp32b+/+ and Anp32b-/- DCs cells are shown in red and green, respectively (Fig. 4.12).

These data suggest that ANP32B’s influence on DC activation is general in nature and not strictly limited to regulation of CD83, as mRNAs of CD40 and CD80 do not contain binding sequences allowing for differential mRNA control.

208

mock LPS-stimulated

100 100

80 80

CD40 60 60 % of Max % of Max 40 40

20 20

0 0 0 102 103 104 105 0 102 103 104 105 : CD40 : CD40

100 100

80 80

60 60 % of Max CD80 % of Max 40 40

20 20

0 0 0 102 103 104 105 0 102 103 104 105 : CD80 : CD80

100 100

80 80

CD83 60 60 % of Max % of Max 40 40

20 20

0 0 0 102 103 104 105 0 102 103 104 105 : CD83 : CD83 Figure 5: Defective stimulation of ANP32B-/- dendritic cells in response to LPS FigureCell surface4.12 .staining Defective of DC cultures Stimulation from wild-type (red) of orAnp32b ANP32B-null-/- (green)dendritic for activation cells markers in CD40,response CD80, and to CD83. LPS. Panels on the left represent mock-stimulated and panels on the right represent LPS-stimulated cultures.

Cell surface staining of DC cultures form WT (red) and Anp32b-null (green) for activation markers CD40, CD80 and CD83. Panels on left represent mock-stimulated and panels on the right represent LPS-stimulated cultures. Plots are representative of data obtained from three experiments involving 1-2 mice per genotype.

209

4.5 Discussion In this chapter, we have used the mixed bred (Ola129:C57BL/6J) and backcrossed

(C57BL/6J) ANP32B deficient mice to elucidate the cellular functions of ANP32B in the immune system. Phenotypic analysis of the mixed bred mice revealed a largely normal immune system. In contrast, results from a single backcrossed Anp32b-/- mouse and fetal liver chimeras suggested that ANP32B may be required for the proper development and homeostasis B and T cells. Regardless of genetic background, however, ANP32B expression appears to be essential for optimal differentiation of B cells.

The analyses of the mixed bred and backcrossed immune system revealed that, compared to WT, B cells from Anp32b-/- mice are twice as likely to differentiate into the

MZ B cell lineage. Because MZ B cell localization to the marginal zone of spleen is not essential for the normal development of these cells [44-45], we can eliminate the possibility that the increased number of MZ B cells in Anp32b-/- mice might be due to alterations in their recruitment to their niche. A more likely explanation for the observed differences is inherent alterations in cell intrinsic mechanisms leading to increased production of MZ B cell precursors. Two potential candidates for this intrinsic effect are:

Nuclear factor kappa B (NFκB) and Notch signaling. NFκB plays a crucial role in B cell and lymphoid organ development. Constitutive activation of the alternative NFκB pathway favors MZ B-cell development and accumulation [46]. It is thus important to examine potential defects in NFκB signaling in Anp32b-/- B cells. In addition, Notch signaling directs the differentiation of MZ B cells and suppresses that of FO B cells in the mouse spleen [47]. Thus, it is also important to investigate the effects of Anp32b deficiency on Notch signaling using the Anp32b-/- mice. Because NFκB and notch

210

signaling are crucial for MZ B cell development, any differences in expression and/or activity may account for the increase in marginal zone B cells seen in Anp32b-/- mice.

Through further analyses of the mixed bred Anp32b-/- mice, we identified a striking defect in the expression of rpS6 in resting and activated Anp32b null T cells.

Upon anti-CD3 and anti-CD28 crosslinking, T lymphocytes activate the PI3K pathway and the phosphorylation of rpS6 in a time dependent manner. In comparison to the WT T cells, Anp32b-/-T cells showed much delayed kinetics in their induction of phosphor-rpS6.

Interestingly, the decrease in rpS6 protein levels was not accompanied with a similar decrease in mRNA levels suggesting that it is largely a post translational event (data not shown). Importantly, this defect in rpS6 expression and phosphorylation was not observed in Anp32a-/- and Anp32e-/- T cells, suggesting that ANP32B is the non- redundant regulator of this protein. This is an intriguing finding as it may provide us with answers to the cause of lethality, decreased weight and earlier deaths of the surviving knockout mice. Further work needs to be done to determine how ANP32B regulates the expression of rpS6 and to examine whether this defect has any consequences for T cell biology. It is also important to test the relevance of this finding in other cells systems, such as hepatocytes, where rpS6 expression may be induced.

Our preliminary studies revealed a defect in Anp32b-/- DC activation in response to LPS stimulation. The failure of Anp32b-/- DCs to activate appropriately in response to

LPS suggests a few different mechanisms by which ANP32B might influence DC stimulation. The simplest explanation may be that ANP32B regulates the surface expression of the LPS receptor TLR4, potentially through HuR-mediated mRNA regulation [188]. Another possibility is that Anp32b-/- DCs may not respond to LPS 211

stimulation due to a defect(s) in signaling downstream of TLR4. A third possibility is the potential dysregulation of rpS6 in DCs. Given the role of ANP32B in regulation of rpS6 in T cells, it is worth investigating if ANP32B may be regulation rpS6 in DCs and whether that may affect DC activation. Finally, ANP32B may be affecting DC activation via regulation of CD83 expression. Recently, it was published that ANP32B is integral for CD83 mRNA transport and stability [176]. As CD83 is an activation marker on mature dendritic cells, it is possible that ANP32B also regulates DC activation via CD83.

To understand this defect at molecular level, it is important to evaluate the various signaling pathways in DC activation in Anp32b-/- DCs.

As a consequence of the fully penetrant lethality in the backcrossed Anp32b-/- mice, initial immune analyses have been performed using the mixed Ola129:C57BL/6J background Anp32b-/- mice. However, since a pure genetic background is preferred for immunological studies, it became necessary for us to generate fetal liver chimeras to examine the immune system of the backcrossed mice. The results presented in this chapter are based on one such set of WT and Anp32b-/- chimeras. Upon comparison of the immune systems of the reconstituted mice and that of the single backcrossed null we had obtained, we observed a consistent defect in B and T lymphocyte homeostasis in the null population. However, there was also a great deal of variation within the small population of null reconstituted mice and this suggests the need to perform more chimeras to get a consensus on the immune phenotype. Alternatively, Anp32b deficient conditional mice may prove useful in elucidating the cell intrinsic functions of ANP32B. Conditional mice can be crossed with mice bearing Cre recombinase under the control of various immune

212

cell subset-specific markers to specifically delete Anp32b in a particular immune lineage and thus, better dissect the role of this gene in the immune system.

ANP32B belongs to the growing family of ANP32 proteins, which have been linked to manifold activities, including regulation of gene expression, cell signaling, proliferation and cell death. In this chapter, we have shown that ANP32B may be necessary for normal differentiation of B cells, for rpS6 expression in T cells and for proper activation of dendritic cells. Our identification of these novel functions of

ANP32B will likely enhance the current state of knowledge of the ANP32 family of proteins and ANP32B in particular. This understanding could also lead to the development of new therapeutic agents that target ANP32 family members and improve the survival of cancer patients.

213

Chapter 5

Final Perspectives

214

5 Chapter 5: Final Perspectives 5.1 Summary

5.1.1 UVRAG

The previous chapters have described the functional characterization of UVRAG and ANP32B through the use of gene deficient mice. The first chapter aimed to examine the in vivo role of UVRAG in T cells. To investigate the physiological function of

UVRAG, we generated conventional UVRAG deficient mice. Deletion of UVRAG led to early embryonic lethality prompting us to generate conditional UVRAG deficient mice.

We crossed these mice with Lck-Cre transgenic mice to specifically delete UVRAG in developing T cells. Loss of UVRAG during T cell development led to defects in the homeostasis of peripheral T cells, as well as defects in the development of iNKT cells.

We validated the intrinsic requirement of UVRAG in these processes by generating mixed bone marrow chimeras. Conditional UVRAG deficient mice also allowed us to examine the function of UVRAG in several different disease models: namely EAE, OVA induced asthma and LCMV. Loss of UVRAG led to a dampened immune response to

OVA induced asthma and LCMV infections. These studies provided valuable insight into the role of UVRAG in T cell responses in vivo.

5.1.2 ANP32B

In the case of ANP32B, the gene deficient mouse approach was useful in establishing the importance of ANP32B in the ANP32 family. We were able to genetically show that ANP32B is the least redundant and most important ANP32 family member. In a mixed genetic background, ANP32B-deficient mice displayed a partially penetrant perinatal lethality that became fully penetrant in a pure C57BL/6 background.

215

As pure background is considered ideal for examining the immune system, we circumvented the lethality of backcrossed mice by generating fetal liver chimeras. These chimeras proved to be useful in defining the immune cell functions that require ANP32B.

However, a great deal of variability between individual reconstituted mice has necessitated the use of a larger cohort of experimental mice to reach a consensus on the phenotype. Collectively, studies using gene deficient mice have allowed us to establish the indisputable importance of ANP32B in murine development and the ANP32 family.

5.2 Future Directions

5.2.1 UVRAG

5.2.1.1 T cell Development

The effect of UVRAG inactivation on T cell development remains unclear. In retrospect, the use of URfl/fl; Lck-Cre mice to study this process was not ideal, as Lck expression, and thus the consequent deletion of UVRAG, initiates relatively late during the DN2 to

DN3 stages of T cell development. Depending on the half-life of the UVRAG protein, it may or may not have been completely eliminated from maturing thymocytes. To examine

UVRAG’s role in T cell development from its beginning, URfl/fl mice should be bred to transgenic Vav-Cre mice, which would ensure UVRAG deletion in the earliest T cell progenitor populations. A phenotypic analysis of early and late stages of T cell development in the Vav-Cre mice, as was done for URfl/fl; Lck-Cre mice, should yield information about the requirement of UVRAG in T cell development. Given the potential role of UVRAG in regulating IL-7 mediated survival of T cells, it is tempting to speculate that the absence of UVRAG in Vav-cre setting may yield a stronger developmental defect than was observed in URfl/fl; Lck-Cre mice.

216

5.2.1.2 T cell proliferation

In this dissertation, we have assessed the role of UVRAG in anti-TCR induced proliferation and foreign antigen induced proliferation ex vivo. Investigation of anti-TCR signalling in vitro suggested that UVRAG is a negative regulator of TCR-induced signalling. UVRAG deficient naïve T cells appeared to have a lower threshold for activation as compared to WT T cells. There are several reasons for why this effect may have been observed. First possibility is that the lymphopenic state of the UVRAG- deficient mice sensitizes the remaining naïve T cells to hyper-proliferation. The second possibility is that UVRAG is intrinsically involved in fine-tuning the responsiveness of the TCR. One can distinguish between these two possibilities by comparing the proliferation of naïve T cells from non-lymphopenic mixed BM chimeras (described in

Chapter 2). If the proliferation rates are comparable to WT chimeric cells, this would rule out a direct role for UVRAG in TCR signalling and suggest that the lymphopenic microenvironment was responsible for making knockout cells more sensitive to stimulation. If, however, UVRAG-deficient naïve T cells from these chimeras continue to show hyper-proliferation upon anti-TCR induced stimulation in vitro, that would suggest that UVRAG intrinsically regulates TCR sensitivity. In this case, we would need to further explore how UVRAG fine-tunes TCR signalling. To do so, we aim to compare proximal and late TCR signalling events following anti-CD3/CD28 stimulation by flow cytometry and western blotting. The possibility that UVRAG may be an intrinsic regulator of TCR signalling raises interesting hypotheses to explain the observed effects of UVRAG on T cell development and on the differential effect of UVRAG on CD4+ and

CD8+ T cell subsets.

217

5.2.1.3 T cell homeostasis

Next we need to determine whether a defect in sp-MHC signalling or cytokine signaling led to a reduced recovery of knockout T cells in the lymphopenia induced homeostatic proliferation models. So far we have used foreign antigen induced and anti-TCR induced proliferation to infer that sp-MHC signalling is likely intact in UVRAG knockout T cells but we have not directly examined sp-MHC induced proliferation. In light of differences between different types of proliferation in T cells, it is important to directly investigate sp-MHC induced proliferation in UVRAG knockout T cells. Only after ruling out a role for UVRAG in sp-MHC induced proliferation, can we hypothesize that a defect in response to cytokines led to reduced recovery of knockout T cells in mixed adoptive transfer experiments. We also need to examine the role of survival in reduced recovery of knockout T cells. To look at survival, we will need to perform mixed adoptive transfer experiments in immuno-replete hosts. This will prevent proliferation of donor cells and allow us to specifically measure the role of survival in T cell recovery.

Most importantly, we need to mechanistically elucidate how UVRAG influences

IL-7 signalling. So far, we have examined and failed to detect an alteration in expression of IL-7Rα chain, phoshp-STAT5α, and Bcl-2 levels. A comprehensive analysis of IL-7 downstream signalling will also involve examining the expression of IL-7Rγ chain, Mcl-

1, Bim, Akt, FOXO proteins and p27, as they are key players in IL-7 signalling. We aim to perform CyTOF (single cell multi-parametric protein detection) analysis on WT and

UVRAG deficient T cells in the presence and absence of IL-7. Through this approach, we hope to identify proteins whose expression may be deregulated by the loss of UVRAG.

Interestingly, as the IL-7Rγ chain is a common gamma chain shared by other 218

cytokines including IL-2, IL-4, IL-9, IL-15 and IL-21. It will also be important to examine the responses downstream of these cytokines in mediating their functions in T cells. If UVRAG regulates common gamma chain shuttling/recycling, we expect downstream of all of the common gamma chain cytokines to be affected. This may hold true for at least IL-15 as we have seen severe defects in the generation, maintenance and function of memory CD8+ T cells.

5.2.1.4 Autophagy in T cells

Many studies have pointed to an indispensable role for UVRAG in autophagy. These reports show that UVRAG is required not only for the initiation of autophagosome formation but also autophagosome maturation. Contrary to these reports, our data shows that UVRAG may not be necessary for mitophagy or activation-induced autophagy in T cells. However, to fully rule out a role for UVRAG in T cell autophagy, we will also need to examine starvation-induced autophagy and cytokine-deprivation induced autophagy as

UVRAG may be specifically required for these types of autophagy in T cells. Besides the

LC3I-II conversion, we aim to examine autophagy by electron and fluorescent microscopy.

5.2.1.5 iNKT cells

Our characterization of the T cell specific UVRAG knockout mice revealed a severe reduction in iNKT cell numbers in the thymus and many of the peripheral organs.

This defect appeared in steady state mice and was more pronounced than the T cell defect we characterized. As the work in this dissertation was focused on T cells, the iNKT cell phenotype was left uncharacterized. Future work on these mice will involve studying the role of UVRAG in iNKT cell biology, including development, homeostasis and function. 219

In particular, iNKT cell development is dependent on positive selection by DP thymocytes. Further work would determine whether the iNKT cell defect is due to an effect of UVRAG on DP thymocyte population and/or an intrinsic role for UVRAG in iNKT cells. Our data showing hyper-proliferation of naïve T cells in the absence of

UVRAG suggests that one of the functions of UVRAG in immune cells may be to fine- tune TCR signalling. As iNKT cells are also dependent on TCR signalling for development and homeostasis, it would be important to investigate if TCR signalling is intact in these cells. To examine the role of UVRAG in early iNKT cell development,

UVRAG animals will be bred with mice expressing Vav-cre recombinase, allowing for the deletion of UVRAG in early hematopoetic progenitor populations.

5.2.1.6 B cells and APCs

UVRAGfl/fl;Vav-Cre mice will also be an invaluable tool in investigating the role of

UVRAG in other immune cell types including B cells, DCs and macrophages. Although we failed to detect a role for UVRAG in T cell autophagy, it is likely that UVRAG’s involvement in this process is cell type specific. Given the essential role of autophagy in antigen presentation by APCs, these cells should be examined for changes in basal and induced autophagy. Moreover, autophagic assays using UVRAG deficient cells should be performed alongside a positive control, or a bone fide autophagy knockout cells such as

ATG5, ATG7 or Beclin-1 deficient cells.

5.2.2 ANP32B

Our analysis of the physiological role of ANP32B, as described in Chapter 3 and

4, was limited by the embryonic lethality of the pure bred C57BL/6 homozygous

ANP32B deficient mice. Two potential solutions to this problem are conditionally 220

targeted mice or ANP32B knockout mice on another genetic background. The conditional knockout mice will allow tissue specific and/or inducible deletion of ANP32B. The severity of the mixed bred and C57BL/6 ANP32B phenotype suggests that conditional mice may have interesting phenotypes and provide a wealth of new experimental tools.

We have already started the latter strategy to see if ANP32B deficient mice are viable on a pure Balb/c background. Although some of these mice have reduced weight and undergo premature deaths, most survive, providing us with a useful tool to investigate the in vivo role of ANP32B in a pure background.

5.3 Concluding Remarks

In conclusion, the work presented in this dissertation has helped elucidate the basic in vivo functions of two genes, UVRAG and ANP32B. Our work on UVRAG has provided insights into the control of autophagy in T cells and identified UVRAG as a novel regulator of naïve T cell homeostasis. We demonstrate that UVRAG may be dispensable for T cell autophagy but it is required for the homeostatic survival and proliferation of naïve T cells. We also demonstrate that ANP32B is the most important member of

ANP32 family of proteins for murine development. Our work implicates ANP32B in many aspects of immune cell development and function. The exact mechanism underlying these effects will be the subject of future investigations.

221

Chapter 6

References

222

6 References

1. Okada, H., et al., Survivin loss in thymocytes triggers p53-mediated growth arrest and p53-independent cell death. J Exp Med, 2004. 199(3): p. 399-410.

2. Mak, T.W., et al., Brca1 required for T cell lineage development but not TCR loci rearrangement. Nat Immunol, 2000. 1(1): p. 77-82.

3. Zaugg, K., et al., Cross-talk between Chk1 and Chk2 in double-mutant thymocytes. Proc Natl Acad Sci U S A, 2007. 104(10): p. 3805-10.

4. Cuervo, A.M., Autophagy: in sickness and in health. Trends Cell Biol, 2004. 14(2): p. 70-7.

5. {Mizushima, M., N., Y. Ohsumi, and T. Yoshimori, Autophagosome formation in mammalian cells. Cell Struct Funct, 2002. 27(6): p. 421-9.

6. Xie, Z. and D.J. Klionsky, Autophagosome formation: core machinery and adaptations. Nat Cell Biol, 2007. 9(10): p. 1102-9.

7. Lum, J.J., et al., Growth factor regulation of autophagy and cell survival in the absence of apoptosis. Cell, 2005. 120(2): p. 237-48.

8. Mizushima, N., et al., In vivo analysis of autophagy in response to nutrient starvation using transgenic mice expressing a fluorescent autophagosome marker. Mol Biol Cell, 2004. 15(3): p. 1101-11.

9. Yorimitsu, T., et al., Endoplasmic reticulum stress triggers autophagy. J Biol Chem, 2006. 281(40): p. 30299-304.

10. Kouroku, Y., et al., ER stress (PERK/eIF2alpha phosphorylation) mediates the polyglutamine-induced LC3 conversion, an essential step for autophagy formation. Cell Death Differ, 2007. 14(2): p. 230-9.

11. Ravikumar, B., et al., Regulation of mammalian autophagy in physiology and pathophysiology. Physiol Rev, 2010. 90(4): p. 1383-435.

12. Scott, R.C., G. Juhasz, and T.P. Neufeld, Direct induction of autophagy by Atg1 inhibits cell growth and induces apoptotic cell death. Curr Biol, 2007. 17(1): p. 1- 11.

13. Wang, Y., et al., Loss of macroautophagy promotes or prevents fibroblast apoptosis depending on the death stimulus. J Biol Chem, 2008. 283(8): p. 4766- 77.

14. Levine, B. and G. Kroemer, Autophagy in the pathogenesis of disease. Cell, 2008. 132(1): p. 27-42. 223

15. Essick, E.E. and F. Sam, Oxidative stress and autophagy in cardiac disease, neurological disorders, aging and cancer. Oxid Med Cell Longev, 2010. 3(3): p. 168-77.

16. Klionsky, D.J., The molecular machinery of autophagy: unanswered questions. J Cell Sci, 2005. 118(Pt 1): p. 7-18.

17. Mizushima, N., et al., Autophagy fights disease through cellular self-digestion. Nature, 2008. 451(7182): p. 1069-75.

18. Munz, C., Enhancing immunity through autophagy. Annu Rev Immunol, 2009. 27: p. 423-49.

19. Yan, Y. and J.M. Backer, Regulation of class III (Vps34) PI3Ks. Biochem Soc Trans, 2007. 35(Pt 2): p. 239-41.

20. Petiot, A., et al., Distinct classes of phosphatidylinositol 3'-kinases are involved in signaling pathways that control macroautophagy in HT-29 cells. J Biol Chem, 2000. 275(2): p. 992-8.

21. Funderburk, S.F., Q.J. Wang, and Z. Yue, The Beclin 1-VPS34 complex--at the crossroads of autophagy and beyond. Trends Cell Biol, 2010. 20(6): p. 355-62.

22. Noda, T., et al., Regulation of membrane biogenesis in autophagy via PI3P dynamics. Semin Cell Dev Biol, 2010. 21(7): p. 671-6.

23. Klionsky, D.J., et al., Guidelines for the use and interpretation of assays for monitoring autophagy in higher eukaryotes. Autophagy, 2008. 4(2): p. 151-75.

24. Liang, C., et al., Autophagic and tumour suppressor activity of a novel Beclin1- binding protein UVRAG. Nat Cell Biol, 2006. 8(7): p. 688-99.

25. Sun, Q., et al., Identification of Barkor as a mammalian autophagy-specific factor for Beclin 1 and class III phosphatidylinositol 3-kinase. Proc Natl Acad Sci U S A, 2008. 105(49): p. 19211-6.

26. Zhong, Y., Q.J. Wang, and Z. Yue, Atg14L and Rubicon: yin and yang of Beclin 1-mediated autophagy control. Autophagy, 2009. 5(6): p. 890-1.

27. Matsunaga, K., et al., Two Beclin 1-binding proteins, Atg14L and Rubicon, reciprocally regulate autophagy at different stages. Nat Cell Biol, 2009. 11(4): p. 385-96.

28. Takahashi, Y., et al., Bif-1 interacts with Beclin 1 through UVRAG and regulates autophagy and tumorigenesis. Nat Cell Biol, 2007. 9(10): p. 1142-51.

29. Pattingre, S., et al., Bcl-2 antiapoptotic proteins inhibit Beclin 1-dependent autophagy. Cell, 2005. 122(6): p. 927-39. 224

30. Tanida, I., T. Ueno, and E. Kominami, LC3 conjugation system in mammalian autophagy. Int J Biochem Cell Biol, 2004. 36(12): p. 2503-18.

31. Komatsu, M., et al., Impairment of starvation-induced and constitutive autophagy in Atg7-deficient mice. J Cell Biol, 2005. 169(3): p. 425-34.

32. Mizushima, N., et al., A protein conjugation system essential for autophagy. Nature, 1998. 395(6700): p. 395-8.

33. Tanida, I., et al., The human homolog of Saccharomyces cerevisiae Apg7p is a Protein-activating enzyme for multiple substrates including human Apg12p, GATE-16, GABARAP, and MAP-LC3. J Biol Chem, 2001. 276(3): p. 1701-6.

34. Shintani, T., et al., Apg10p, a novel protein-conjugating enzyme essential for autophagy in yeast. EMBO J, 1999. 18(19): p. 5234-41.

35. Mizushima, N., et al., A new protein conjugation system in human. The counterpart of the yeast Apg12p conjugation system essential for autophagy. J Biol Chem, 1998. 273(51): p. 33889-92.

36. Mizushima, N., et al., Dissection of autophagosome formation using Apg5- deficient mouse embryonic stem cells. J Cell Biol, 2001. 152(4): p. 657-68.

37. Fujita, N., et al., The Atg16L complex specifies the site of LC3 lipidation for membrane biogenesis in autophagy. Mol Biol Cell, 2008. 19(5): p. 2092-100.

38. Hanada, T., et al., The Atg12-Atg5 conjugate has a novel E3-like activity for protein lipidation in autophagy. J Biol Chem, 2007. 282(52): p. 37298-302.

39. Kabeya, Y., et al., LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing. EMBO J, 2000. 19(21): p. 5720-8.

40. Kabeya, Y., et al., LC3, GABARAP and GATE16 localize to autophagosomal membrane depending on form-II formation. J Cell Sci, 2004. 117(Pt 13): p. 2805- 12.

41. Ichimura, Y., et al., A ubiquitin-like system mediates protein lipidation. Nature, 2000. 408(6811): p. 488-92.

42. Xie, Z., U. Nair, and D.J. Klionsky, Atg8 controls phagophore expansion during autophagosome formation. Mol Biol Cell, 2008. 19(8): p. 3290-8.

43. Nakatogawa, H., Y. Ichimura, and Y. Ohsumi, Atg8, a ubiquitin-like protein required for autophagosome formation, mediates membrane tethering and hemifusion. Cell, 2007. 130(1): p. 165-78.

44. Jager, S., et al., Role for Rab7 in maturation of late autophagic vacuoles. J Cell Sci, 2004. 117(Pt 20): p. 4837-48. 225

45. Gutierrez, M.G., et al., Rab7 is required for the normal progression of the autophagic pathway in mammalian cells. J Cell Sci, 2004. 117(Pt 13): p. 2687-97.

46. Liang, C., et al., Beclin1-binding UVRAG targets the class C Vps complex to coordinate autophagosome maturation and endocytic trafficking. Nat Cell Biol, 2008. 10(7): p. 776-87.

47. Tanida, I., et al., Lysosomal turnover, but not a cellular level, of endogenous LC3 is a marker for autophagy. Autophagy, 2005. 1(2): p. 84-91.

48. Klionsky, D.J., Autophagy: from phenomenology to molecular understanding in less than a decade. Nat Rev Mol Cell Biol, 2007. 8(11): p. 931-7.

49. Barth, S., D. Glick, and K.F. Macleod, Autophagy: assays and artifacts. J Pathol, 2010. 221(2): p. 117-24.

50. Mizushima, N. and T. Yoshimori, How to interpret LC3 immunoblotting. Autophagy, 2007. 3(6): p. 542-5.

51. Kimura, S., et al., Monitoring autophagy in mammalian cultured cells through the dynamics of LC3. Methods Enzymol, 2009. 452: p. 1-12.

52. Kadowaki, M. and M.R. Karim, Cytosolic LC3 ratio as a quantitative index of macroautophagy. Methods Enzymol, 2009. 452: p. 199-213.

53. Kuma, A., M. Matsui, and N. Mizushima, LC3, an autophagosome marker, can be incorporated into protein aggregates independent of autophagy: caution in the interpretation of LC3 localization. Autophagy, 2007. 3(4): p. 323-8.

54. Kimura, S., T. Noda, and T. Yoshimori, Dissection of the autophagosome maturation process by a novel reporter protein, tandem fluorescent-tagged LC3. Autophagy, 2007. 3(5): p. 452-60.

55. Bjorkoy, G., et al., Monitoring autophagic degradation of p62/SQSTM1. Methods Enzymol, 2009. 452: p. 181-97.

56. Pankiv, S., et al., p62/SQSTM1 binds directly to Atg8/LC3 to facilitate degradation of ubiquitinated protein aggregates by autophagy. J Biol Chem, 2007. 282(33): p. 24131-45.

57. Komatsu, M., et al., Homeostatic levels of p62 control cytoplasmic inclusion body formation in autophagy-deficient mice. Cell, 2007. 131(6): p. 1149-63.

58. Waguri, S. and M. Komatsu, Biochemical and morphological detection of inclusion bodies in autophagy-deficient mice. Methods Enzymol, 2009. 453: p. 181-96.

226

59. Moscat, J. and M.T. Diaz-Meco, p62 at the crossroads of autophagy, apoptosis, and cancer. Cell, 2009. 137(6): p. 1001-4.

60. Tolkovsky, A.M., Mitophagy. Biochim Biophys Acta, 2009. 1793(9): p. 1508-15.

61. Pua, H.H., et al., Autophagy is essential for mitochondrial clearance in mature T lymphocytes. J Immunol, 2009. 182(7): p. 4046-55.

62. Stephenson, L.M., et al., Identification of Atg5-dependent transcriptional changes and increases in mitochondrial mass in Atg5-deficient T lymphocytes. Autophagy, 2009. 5(5): p. 625-35.

63. Jia, W. and Y.W. He, Temporal regulation of intracellular organelle homeostasis in T lymphocytes by autophagy. J Immunol, 2011. 186(9): p. 5313-22.

64. Willinger, T. and R.A. Flavell, Canonical autophagy dependent on the class III phosphoinositide-3 kinase Vps34 is required for naive T-cell homeostasis. Proc Natl Acad Sci U S A, 2012. 109(22): p. 8670-5.

65. Godfrey, D.I., et al., A developmental pathway involving four phenotypically and functionally distinct subsets of CD3-CD4-CD8- triple-negative adult mouse thymocytes defined by CD44 and CD25 expression. J Immunol, 1993. 150(10): p. 4244-52.

66. Fowlkes, B.J., et al., Early T lymphocytes. Differentiation in vivo of adult intrathymic precursor cells. J Exp Med, 1985. 162(3): p. 802-22.

67. Godfrey, D.I., et al., Onset of TCR-beta gene rearrangement and role of TCR-beta expression during CD3-CD4-CD8- thymocyte differentiation. J Immunol, 1994. 152(10): p. 4783-92.

68. Wada, H., et al., Adult T-cell progenitors retain myeloid potential. Nature, 2008. 452(7188): p. 768-72.

69. Bell, J.J. and A. Bhandoola, The earliest thymic progenitors for T cells possess myeloid lineage potential. Nature, 2008. 452(7188): p. 764-7.

70. Wu, L., et al., CD4 expressed on earliest T-lineage precursor cells in the adult murine thymus. Nature, 1991. 349(6304): p. 71-4.

71. Allman, D., et al., Thymopoiesis independent of common lymphoid progenitors. Nat Immunol, 2003. 4(2): p. 168-74.

72. Strauch, U.G., et al., Distinct binding specificities of integrins alpha 4 beta 7 (LPAM-1), alpha 4 beta 1 (VLA-4), and alpha IEL beta 7. Int Immunol, 1994. 6(2): p. 263-75.

227

73. Moore, K.B. and S.A. Moody, Animal-vegetal asymmetries influence the earliest steps in retina fate commitment in Xenopus. Dev Biol, 1999. 212(1): p. 25-41.

74. Zuniga-Pflucker, J.C., D. Jiang, and M.J. Lenardo, Requirement for TNF-alpha and IL-1 alpha in fetal thymocyte commitment and differentiation. Science, 1995. 268(5219): p. 1906-9.

75. Livak, F., et al., Characterization of TCR gene rearrangements during adult murine T cell development. J Immunol, 1999. 162(5): p. 2575-80.

76. Petrie, H.T., et al., T cell receptor gene recombination patterns and mechanisms: cell death, rescue, and T cell production. J Exp Med, 1995. 182(1): p. 121-7.

77. Petrie, H.T., et al., Development of immature thymocytes: initiation of CD3, CD4, and CD8 acquisition parallels down-regulation of the interleukin 2 receptor alpha chain. Eur J Immunol, 1990. 20(12): p. 2813-5.

78. Nikolic-Zugic, J., M.W. Moore, and M.J. Bevan, Characterization of the subset of immature thymocytes which can undergo rapid in vitro differentiation. Eur J Immunol, 1989. 19(4): p. 649-53.

79. Dudley, E.C., et al., T cell receptor beta chain gene rearrangement and selection during thymocyte development in adult mice. Immunity, 1994. 1(2): p. 83-93.

80. Linette, G.P., et al., Bcl-2 is upregulated at the CD4+ CD8+ stage during positive selection and promotes thymocyte differentiation at several control points. Immunity, 1994. 1(3): p. 197-205.

81. Mazzucchelli, R. and S.K. Durum, Interleukin-7 receptor expression: intelligent design. Nat Rev Immunol, 2007. 7(2): p. 144-54.

82. Moore, N.C., et al., Developmental regulation of bcl-2 expression in the thymus. Immunology, 1994. 81(1): p. 115-9.

83. Sudo, T., et al., Expression and function of the interleukin 7 receptor in murine lymphocytes. Proc Natl Acad Sci U S A, 1993. 90(19): p. 9125-9.

84. Alam, S.M., et al., T-cell-receptor affinity and thymocyte positive selection. Nature, 1996. 381(6583): p. 616-20.

85. Minter, L.M. and B.A. Osborne, Cell death in the thymus--it' s all a matter of contacts. Semin Immunol, 2003. 15(3): p. 135-44.

86. Starr, T.K., S.C. Jameson, and K.A. Hogquist, Positive and negative selection of T cells. Annu Rev Immunol, 2003. 21: p. 139-76.

87. Jameson, S.C., Maintaining the norm: T-cell homeostasis. Nat Rev Immunol, 2002. 2(8): p. 547-56. 228

88. Khaled, A.R. and S.K. Durum, Lymphocide: cytokines and the control of lymphoid homeostasis. Nat Rev Immunol, 2002. 2(11): p. 817-30.

89. Sprent, J., et al., T cell homeostasis. Immunol Cell Biol, 2008. 86(4): p. 312-9.

90. Ma, A., R. Koka, and P. Burkett, Diverse functions of IL-2, IL-15, and IL-7 in lymphoid homeostasis. Annu Rev Immunol, 2006. 24: p. 657-79.

91. Williams, M.A., A.J. Tyznik, and M.J. Bevan, Interleukin-2 signals during priming are required for secondary expansion of CD8+ memory T cells. Nature, 2006. 441(7095): p. 890-3.

92. Blattman, J.N., et al., Estimating the precursor frequency of naive antigen-specific CD8 T cells. J Exp Med, 2002. 195(5): p. 657-64.

93. Murali-Krishna, K., et al., Counting antigen-specific CD8 T cells: a reevaluation of bystander activation during viral infection. Immunity, 1998. 8(2): p. 177-87.

94. Hildeman, D.A., et al., Activated T cell death in vivo mediated by proapoptotic bcl-2 family member bim. Immunity, 2002. 16(6): p. 759-67.

95. Hughes, P.D., et al., Apoptosis regulators Fas and Bim cooperate in shutdown of chronic immune responses and prevention of autoimmunity. Immunity, 2008. 28(2): p. 197-205.

96. Willis, S.N. and J.M. Adams, Life in the balance: how BH3-only proteins induce apoptosis. Curr Opin Cell Biol, 2005. 17(6): p. 617-25.

97. Wojciechowski, S., et al., Bim mediates apoptosis of CD127(lo) effector T cells and limits T cell memory. Eur J Immunol, 2006. 36(7): p. 1694-706.

98. Swain, S.L., H. Hu, and G. Huston, Class II-independent generation of CD4 memory T cells from effectors. Science, 1999. 286(5443): p. 1381-3.

99. Bruno, L., H. von Boehmer, and J. Kirberg, Cell division in the compartment of naive and memory T lymphocytes. Eur J Immunol, 1996. 26(12): p. 3179-84.

100. Schluns, K.S. and L. Lefrancois, Cytokine control of memory T-cell development and survival. Nat Rev Immunol, 2003. 3(4): p. 269-79.

101. Surh, C.D. and J. Sprent, Homeostasis of naive and memory T cells. Immunity, 2008. 29(6): p. 848-62.

102. Maki, K., et al., Interleukin 7 receptor-deficient mice lack gammadelta T cells. Proc Natl Acad Sci U S A, 1996. 93(14): p. 7172-7.

103. Schluns, K.S., et al., Interleukin-7 mediates the homeostasis of naive and memory CD8 T cells in vivo. Nat Immunol, 2000. 1(5): p. 426-32.

229

104. Osborne, L.C., et al., Impaired CD8 T cell memory and CD4 T cell primary responses in IL-7R alpha mutant mice. J Exp Med, 2007. 204(3): p. 619-31.

105. Carrio, R., C.E. Rolle, and T.R. Malek, Non-redundant role for IL-7R signaling for the survival of CD8+ memory T cells. Eur J Immunol, 2007. 37(11): p. 3078- 88.

106. Buentke, E., et al., Do CD8 effector cells need IL-7R expression to become resting memory cells? Blood, 2006. 108(6): p. 1949-56.

107. Osborne, L.C. and N. Abraham, Regulation of memory T cells by gammac cytokines. Cytokine, 2010. 50(2): p. 105-13.

108. Pellegrini, M., et al., Adjuvant IL-7 antagonizes multiple cellular and molecular inhibitory networks to enhance immunotherapies. Nat Med, 2009. 15(5): p. 528- 36.

109. Melchionda, F., et al., Adjuvant IL-7 or IL-15 overcomes immunodominance and improves survival of the CD8+ memory cell pool. J Clin Invest, 2005. 115(5): p. 1177-87.

110. McCarthy, D.P., M.H. Richards, and S.D. Miller, Mouse models of multiple sclerosis: experimental autoimmune encephalomyelitis and Theiler's virus- induced demyelinating disease. Methods Mol Biol, 2012. 900: p. 381-401.

111. Yednock, T.A., et al., Prevention of experimental autoimmune encephalomyelitis by antibodies against alpha 4 beta 1 integrin. Nature, 1992. 356(6364): p. 63-6.

112. Kawakami, N., et al., The activation status of neuroantigen-specific T cells in the target organ determines the clinical outcome of autoimmune encephalomyelitis. J Exp Med, 2004. 199(2): p. 185-97.

113. Lees, J.R., et al., Regional CNS responses to IFN-gamma determine lesion localization patterns during EAE pathogenesis. J Exp Med, 2008. 205(11): p. 2633-42.

114. Kroenke, M.A., et al., IL-12- and IL-23-modulated T cells induce distinct types of EAE based on histology, CNS chemokine profile, and response to cytokine inhibition. J Exp Med, 2008. 205(7): p. 1535-41.

115. Panitch, H.S., et al., Exacerbations of multiple sclerosis in patients treated with gamma interferon. Lancet, 1987. 1(8538): p. 893-5.

116. Cua, D.J., et al., Interleukin-23 rather than interleukin-12 is the critical cytokine for autoimmune inflammation of the brain. Nature, 2003. 421(6924): p. 744-8.

117. Langrish, C.L., et al., IL-23 drives a pathogenic T cell population that induces autoimmune inflammation. J Exp Med, 2005. 201(2): p. 233-40. 230

118. Hofstetter, H.H., et al., Therapeutic efficacy of IL-17 neutralization in murine experimental autoimmune encephalomyelitis. Cell Immunol, 2005. 237(2): p. 123- 30.

119. Nials, A.T. and S. Uddin, Mouse models of allergic asthma: acute and chronic allergen challenge. Dis Model Mech, 2008. 1(4-5): p. 213-20.

120. Harty, J.T., A.R. Tvinnereim, and D.W. White, CD8+ T cell effector mechanisms in resistance to infection. Annu Rev Immunol, 2000. 18: p. 275-308.

121. Doedens, A.L., et al., Hypoxia-inducible factors enhance the effector responses of CD8(+) T cells to persistent antigen. Nat Immunol, 2013. 14(11): p. 1173-82.

122. Altman, J.D., et al., Phenotypic analysis of antigen-specific T lymphocytes. Science, 1996. 274(5284): p. 94-6.

123. Kirkegaard, K., M.P. Taylor, and W.T. Jackson, Cellular autophagy: surrender, avoidance and subversion by microorganisms. Nat Rev Microbiol, 2004. 2(4): p. 301-14.

124. Nakagawa, I., et al., Autophagy defends cells against invading group A Streptococcus. Science, 2004. 306(5698): p. 1037-40.

125. Gutierrez, M.G., et al., Autophagy is a defense mechanism inhibiting BCG and Mycobacterium tuberculosis survival in infected macrophages. Cell, 2004. 119(6): p. 753-66.

126. Ogawa, M., et al., Escape of intracellular Shigella from autophagy. Science, 2005. 307(5710): p. 727-31.

127. Schmid, D., M. Pypaert, and C. Munz, Antigen-loading compartments for major histocompatibility complex class II molecules continuously receive input from autophagosomes. Immunity, 2007. 26(1): p. 79-92.

128. Jagannath, C., et al., Autophagy enhances the efficacy of BCG vaccine by increasing peptide presentation in mouse dendritic cells. Nat Med, 2009. 15(3): p. 267-76.

129. Virgin, H.W. and B. Levine, Autophagy genes in immunity. Nat Immunol, 2009. 10(5): p. 461-70.

130. Li, C., et al., Autophagy is induced in CD4+ T cells and important for the growth factor-withdrawal cell death. J Immunol, 2006. 177(8): p. 5163-8.

131. Pua, H.H., et al., A critical role for the autophagy gene Atg5 in T cell survival and proliferation. J Exp Med, 2007. 204(1): p. 25-31.

231

132. Yu, L., et al., Regulation of an ATG7-beclin 1 program of autophagic cell death by caspase-8. Science, 2004. 304(5676): p. 1500-2.

133. Nedjic, J., et al., Autophagy in thymic epithelium shapes the T-cell repertoire and is essential for tolerance. Nature, 2008. 455(7211): p. 396-400.

134. Hubbard, V.M., et al., Macroautophagy regulates energy metabolism during effector T cell activation. J Immunol, 2010. 185(12): p. 7349-57.

135. McLeod, I.X., et al., The class III kinase Vps34 promotes T lymphocyte survival through regulating IL-7Ralpha surface expression. J Immunol, 2011. 187(10): p. 5051-61.

136. Bell, B.D., et al., FADD and caspase-8 control the outcome of autophagic signaling in proliferating T cells. Proc Natl Acad Sci U S A, 2008. 105(43): p. 16677-82.

137. Feng, C.G., et al., The immunity-related GTPase Irgm1 promotes the expansion of activated CD4+ T cell populations by preventing interferon-gamma-induced cell death. Nat Immunol, 2008. 9(11): p. 1279-87.

138. Espert, L., et al., Autophagy is involved in T cell death after binding of HIV-1 envelope proteins to CXCR4. J Clin Invest, 2006. 116(8): p. 2161-72.

139. Kovacs, J.R., et al., Autophagy promotes T-cell survival through degradation of proteins of the cell death machinery. Cell Death Differ, 2012. 19(1): p. 144-52.

140. Arsov, I., et al., A role for autophagic protein beclin 1 early in lymphocyte development. J Immunol, 2011. 186(4): p. 2201-9.

141. Parekh, V.V., et al., Impaired autophagy, defective T cell homeostasis, and a wasting syndrome in mice with a T cell-specific deletion of Vps34. J Immunol, 2013. 190(10): p. 5086-101.

142. Perelman, B., et al., Molecular cloning of a novel human gene encoding a 63-kDa protein and its sublocalization within the 11q13 locus. Genomics, 1997. 41(3): p. 397-405.

143. Bekri, S., et al., Detailed map of a region commonly amplified at 11q13-->q14 in human breast carcinoma. Cytogenet Cell Genet, 1997. 79(1-2): p. 125-31.

144. Kim, M.S., et al., Frameshift mutation of UVRAG, an autophagy-related gene, in gastric carcinomas with microsatellite instability. Hum Pathol, 2008. 39(7): p. 1059-63.

145. Ionov, Y., et al., Manipulation of nonsense mediated decay identifies gene mutations in colon cancer Cells with microsatellite instability. Oncogene, 2004. 23(3): p. 639-45. 232

146. Goi, T., et al., Ascending colon cancer with hepatic metastasis and cholecystolithiasis in a patient with situs inversus totalis without any expression of UVRAG mRNA: report of a case. Surg Today, 2003. 33(9): p. 702-6.

147. Itakura, E., et al., Beclin 1 forms two distinct phosphatidylinositol 3-kinase complexes with mammalian Atg14 and UVRAG. Mol Biol Cell, 2008. 19(12): p. 5360-72.

148. Song, Z., et al., Essential Role for UVRAG in Autophagy and Maintenance of Cardiac Function. Cardiovasc Res, 2013.

149. Lee, G., et al., UVRAG is required for organ rotation by regulating Notch endocytosis in Drosophila. Dev Biol, 2011. 356(2): p. 588-97.

150. He, S., et al., PtdIns(3)P-bound UVRAG coordinates Golgi-ER retrograde and Atg9 transport by differential interactions with the ER tether and the beclin 1 complex. Nat Cell Biol, 2013. 15(10): p. 1206-19.

151. Yang, W., et al., Protein kinase B/Akt1 inhibits autophagy by down-regulating UVRAG expression. Exp Cell Res, 2013. 319(3): p. 122-33.

152. Knaevelsrud, H., et al., UVRAG mutations associated with microsatellite unstable colon cancer do not affect autophagy. Autophagy, 2010. 6(7): p. 863-70.

153. Zhao, Z., et al., A dual role for UVRAG in maintaining chromosomal stability independent of autophagy. Dev Cell, 2012. 22(5): p. 1001-16.

154. Yin, X., et al., A critical role for UVRAG in apoptosis. Autophagy, 2011. 7(10): p. 1242-4.

155. Huyton, T. and C. Wolberger, The crystal structure of the tumor suppressor protein pp32 (Anp32a): structural insights into Anp32 family of proteins. Protein Sci, 2007. 16(7): p. 1308-15.

156. Matilla, A. and M. Radrizzani, The Anp32 family of proteins containing leucine- rich repeats. Cerebellum, 2005. 4(1): p. 7-18.

157. Santa-Coloma, T.A., Anp32e (Cpd1) and related protein phosphatase 2 inhibitors. Cerebellum, 2003. 2(4): p. 310-20.

158. Anisimov, S.V., et al., SAGE identification of gene transcripts with profiles unique to pluripotent mouse R1 embryonic stem cells. Genomics, 2002. 79(2): p. 169-76.

159. Brody, J.R., et al., pp32 reduction induces differentiation of TSU-Pr1 cells. Am J Pathol, 2004. 164(1): p. 273-83.

233

160. Kular, R.K., et al., Neuronal differentiation is regulated by leucine-rich acidic nuclear protein (LANP), a member of the inhibitor of histone acetyltransferase complex. J Biol Chem, 2009. 284(12): p. 7783-92.

161. Puente, L.G., et al., Comparative analysis of phosphoprotein-enriched myocyte proteomes reveals widespread alterations during differentiation. FEBS Lett, 2004. 574(1-3): p. 138-44.

162. Adegbola, O. and G.R. Pasternack, Phosphorylated retinoblastoma protein complexes with pp32 and inhibits pp32-mediated apoptosis. J Biol Chem, 2005. 280(16): p. 15497-502.

163. Hoffarth, S., et al., pp32/PHAPI determines the apoptosis response of non-small- cell lung cancer. Cell Death Differ, 2008. 15(1): p. 161-70.

164. Jiang, X., et al., Distinctive roles of PHAP proteins and prothymosin-alpha in a death regulatory pathway. Science, 2003. 299(5604): p. 223-6.

165. Mazroui, R., et al., Caspase-mediated cleavage of HuR in the cytoplasm contributes to pp32/PHAP-I regulation of apoptosis. J Cell Biol, 2008. 180(1): p. 113-27.

166. Pan, W., et al., PHAPI/pp32 suppresses tumorigenesis by stimulating apoptosis. J Biol Chem, 2009. 284(11): p. 6946-54.

167. Schafer, Z.T., et al., Enhanced sensitivity to cytochrome c-induced apoptosis mediated by PHAPI in breast cancer cells. Cancer Res, 2006. 66(4): p. 2210-8.

168. Sun, W., et al., Proliferation related acidic leucine-rich protein PAL31 functions as a caspase-3 inhibitor. Biochem Biophys Res Commun, 2006. 342(3): p. 817- 23.

169. Amasaki, H., et al., Distributional changes of BrdU, PCNA, E2F1 and PAL31 molecules in developing murine palatal rugae. Ann Anat, 2003. 185(6): p. 517- 23.

170. Fukukawa, C., et al., pp32/ I-1(PP2A) negatively regulates the Raf-1/MEK/ERK pathway. Cancer Lett, 2005. 226(2): p. 155-60.

171. Sun, W., et al., PAL31, a nuclear protein required for progression to the S phase. Biochem Biophys Res Commun, 2001. 280(4): p. 1048-54.

172. Costanzo, R.V., et al., Anp32e/Cpd1 regulates protein phosphatase 2A activity at synapses during synaptogenesis. Eur J Neurosci, 2006. 23(2): p. 309-24.

173. Li, M., A. Makkinje, and Z. Damuni, Molecular identification of I1PP2A, a novel potent heat-stable inhibitor protein of protein phosphatase 2A. Biochemistry, 1996. 35(22): p. 6998-7002. 234

174. Radrizzani, M., et al., Differential expression of CPD1 during postnatal development in the mouse cerebellum. Brain Res, 2001. 907(1-2): p. 162-74.

175. Hill, M.M., et al., Analysis of the composition, assembly kinetics and activity of native Apaf-1 apoptosomes. EMBO J, 2004. 23(10): p. 2134-45.

176. Fries, B., et al., Analysis of nucleocytoplasmic trafficking of the HuR ligand APRIL and its influence on CD83 expression. J Biol Chem, 2007. 282(7): p. 4504- 15.

177. Itin, C., et al., Mapmodulin, cytoplasmic dynein, and microtubules enhance the transport of mannose 6-phosphate receptors from endosomes to the trans-golgi network. Mol Biol Cell, 1999. 10(7): p. 2191-7.

178. Opal, P., et al., Mapmodulin/leucine-rich acidic nuclear protein binds the light chain of microtubule-associated protein 1B and modulates neuritogenesis. J Biol Chem, 2003. 278(36): p. 34691-9.

179. Kutney, S.N., et al., A signaling role of histone-binding proteins and INHAT subunits pp32 and Set/TAF-Ibeta in integrating chromatin hypoacetylation and transcriptional repression. J Biol Chem, 2004. 279(29): p. 30850-5.

180. Schneider, R., et al., Direct binding of INHAT to H3 tails disrupted by modifications. J Biol Chem, 2004. 279(23): p. 23859-62.

181. Seo, S.B., et al., Regulation of histone acetylation and transcription by nuclear protein pp32, a subunit of the INHAT complex. J Biol Chem, 2002. 277(16): p. 14005-10.

182. Tochio, N., et al., Solution structure of histone chaperone ANP32B: interaction with core histones H3-H4 through its acidic concave domain. J Mol Biol, 2010. 401(1): p. 97-114.

183. Cvetanovic, M., et al., The role of LANP and ataxin 1 in E4F-mediated transcriptional repression. EMBO Rep, 2007. 8(7): p. 671-7.

184. Loven, M.A., et al., A novel estrogen receptor alpha-associated protein alters receptor-deoxyribonucleic acid interactions and represses receptor-mediated transcription. Mol Endocrinol, 2004. 18(11): p. 2649-59.

185. Munemasa, Y., et al., Promoter region-specific histone incorporation by the novel histone chaperone ANP32B and DNA-binding factor KLF5. Mol Cell Biol, 2008. 28(3): p. 1171-81.

186. Shen, S.M., et al., Downregulation of ANP32B, a novel substrate of caspase-3, enhances caspase-3 activation and apoptosis induction in myeloid leukemic cells. Carcinogenesis, 2010. 31(3): p. 419-26.

235

187. Gallouzi, I.E., C.M. Brennan, and J.A. Steitz, Protein ligands mediate the CRM1- dependent export of HuR in response to heat shock. RNA, 2001. 7(9): p. 1348-61.

188. Lin, F.Y., et al., The role of human antigen R, an RNA-binding protein, in mediating the stabilization of toll-like receptor 4 mRNA induced by endotoxin: a novel mechanism involved in vascular inflammation. Arterioscler Thromb Vasc Biol, 2006. 26(12): p. 2622-9.

189. Chemnitz, J., et al., Phosphorylation of the HuR ligand APRIL by casein kinase 2 regulates CD83 expression. Eur J Immunol, 2009. 39(1): p. 267-79.

190. Brennan, C.M., I.E. Gallouzi, and J.A. Steitz, Protein ligands to HuR modulate its interaction with target mRNAs in vivo. J Cell Biol, 2000. 151(1): p. 1-14.

191. Seo, S.B., et al., Regulation of histone acetylation and transcription by INHAT, a human cellular complex containing the set oncoprotein. Cell, 2001. 104(1): p. 119-30.

192. Hublitz, P., et al., NIR is a novel INHAT repressor that modulates the transcriptional activity of p53. Genes Dev, 2005. 19(23): p. 2912-24.

193. Chanchevalap, S., et al., Kruppel-like factor 5 is an important mediator for lipopolysaccharide-induced proinflammatory response in intestinal epithelial cells. Nucleic Acids Res, 2006. 34(4): p. 1216-23.

194. Zhu, L., et al., Cloning and characterization of a new silver-stainable protein SSP29, a member of the LRR family. Biochem Mol Biol Int, 1997. 42(5): p. 927- 35.

195. Panic, L., et al., Ribosomal protein S6 gene haploinsufficiency is associated with activation of a p53-dependent checkpoint during gastrulation. Mol Cell Biol, 2006. 26(23): p. 8880-91.

196. Volarevic, S., et al., Proliferation, but not growth, blocked by conditional deletion of 40S ribosomal protein S6. Science, 2000. 288(5473): p. 2045-7.

197. Sulic, S., et al., Inactivation of S6 ribosomal protein gene in T lymphocytes activates a p53-dependent checkpoint response. Genes Dev, 2005. 19(24): p. 3070-82.

198. Reilly, P.T., et al., Acidic nuclear phosphoprotein 32kDa (ANP32)B-deficient mouse reveals a hierarchy of ANP32 importance in mammalian development. Proc Natl Acad Sci U S A, 2011. 108(25): p. 10243-8.

199. Opal, P., et al., Generation and characterization of LANP/pp32 null mice. Mol Cell Biol, 2004. 24(8): p. 3140-9.

236

200. Reilly, P.T., et al., Generation and characterization of the Anp32e-deficient mouse. PLoS One, 2010. 5(10): p. e13597.

201. Buszczak, M., et al., The carnegie protein trap library: a versatile tool for Drosophila developmental studies. Genetics, 2007. 175(3): p. 1505-31.

202. Teitz, T., et al., Isolation by polymerase chain reaction of a cDNA whose product partially complements the ultraviolet sensitivity of xeroderma pigmentosum group C cells. Gene, 1990. 87(2): p. 295-8.

203. Kang, R., et al., The Beclin 1 network regulates autophagy and apoptosis. Cell Death Differ, 2011. 18(4): p. 571-80.

204. Yin, X., et al., UV irradiation resistance-associated gene suppresses apoptosis by interfering with BAX activation. EMBO Rep, 2011. 12(7): p. 727-34.

205. Hao, Z. and K. Rajewsky, Homeostasis of peripheral B cells in the absence of B cell influx from the bone marrow. J Exp Med, 2001. 194(8): p. 1151-64.

206. Brustle, A., et al., The NF-kappaB regulator MALT1 determines the encephalitogenic potential of Th17 cells. J Clin Invest, 2012. 122(12): p. 4698- 709.

207. Battegay, M., et al., Quantification of lymphocytic choriomeningitis virus with an immunological focus assay in 24- or 96-well plates. J Virol Methods, 1991. 33(1- 2): p. 191-8.

208. Hao, Z., et al., Fas receptor expression in germinal-center B cells is essential for T and B lymphocyte homeostasis. Immunity, 2008. 29(4): p. 615-27.

209. Ge, Q., et al., Homeostatic T cell proliferation in a T cell-dendritic cell coculture system. Proc Natl Acad Sci U S A, 2002. 99(5): p. 2983-8.

210. Theofilopoulos, A.N., W. Dummer, and D.H. Kono, T cell homeostasis and systemic autoimmunity. J Clin Invest, 2001. 108(3): p. 335-40.

211. Shin, H., et al., A role for the transcriptional repressor Blimp-1 in CD8(+) T cell exhaustion during chronic viral infection. Immunity, 2009. 31(2): p. 309-20.

212. Hu, Z., et al., The molecular portraits of breast tumors are conserved across microarray platforms. BMC Genomics, 2006. 7: p. 96.

213. Miller, L.D., et al., An expression signature for p53 status in human breast cancer predicts mutation status, transcriptional effects, and patient survival. Proc Natl Acad Sci U S A, 2005. 102(38): p. 13550-5.

214. van de Vijver, M.J., et al., A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med, 2002. 347(25): p. 1999-2009. 237

215. Okada, H., et al., Generation and characterization of Smac/DIABLO-deficient mice. Mol Cell Biol, 2002. 22(10): p. 3509-17.

216. Ruland, J., et al., p53 accumulation, defective cell proliferation, and early embryonic lethality in mice lacking tsg101. Proc Natl Acad Sci U S A, 2001. 98(4): p. 1859-64.

217. Lin, A.W., et al., Premature senescence involving p53 and p16 is activated in response to constitutive MEK/MAPK mitogenic signaling. Genes Dev, 1998. 12(19): p. 3008-19.

218. Dufner, A., et al., CARD6 is interferon inducible but not involved in nucleotide- binding oligomerization domain protein signaling leading to NF-kappaB activation. Mol Cell Biol, 2008. 28(5): p. 1541-52.

219. Zippo, A., et al., Identification of Flk-1 target genes in vasculogenesis: Pim-1 is required for endothelial and mural cell differentiation in vitro. Blood, 2004. 103(12): p. 4536-44.

220. Ulitzur, N., M. Humbert, and S.R. Pfeffer, Mapmodulin: a possible modulator of the interaction of microtubule-associated proteins with microtubules. Proc Natl Acad Sci U S A, 1997. 94(10): p. 5084-9.

221. Lechmann, M., et al., The CD83 reporter mouse elucidates the activity of the CD83 promoter in B, T, and dendritic cell populations in vivo. Proc Natl Acad Sci U S A, 2008. 105(33): p. 11887-92.

238