Regulation of Effective B Cell Responses to Chronic Infection

University of Pennsylvania ScholarlyCommons

Publicly Accessible Penn Dissertations

2019

Regulation Of Effective B Cell Responses To Chronic Infection

Ryan Phillip Staupe University of Pennsylvania

Follow this and additional works at: https://repository.upenn.edu/edissertations

Part of the Allergy and Immunology Commons, Immunology and Infectious Disease Commons, and the Medical Immunology Commons

Recommended Citation Staupe, Ryan Phillip, "Regulation Of Effective B Cell Responses To Chronic Infection" (2019). Publicly Accessible Penn Dissertations. 3637. https://repository.upenn.edu/edissertations/3637

This paper is posted at ScholarlyCommons. https://repository.upenn.edu/edissertations/3637 For more information, please contact [email protected]. Regulation Of Effective B Cell Responses To Chronic Infection

Abstract Chronic viral infections disrupt B cell responses leading to impaired affinity maturation and delayed control of viremia. Previous studies have identified early pre-germinal center (GC) B cell attrition but the impact of chronic infections on B cell fate decisions in GCs remains poorly understood. To address this question, we used single-cell transcriptional profiling of virus-specific GC B cells during chronic viral infection to test the hypothesis that chronic viral infection disrupted GC B cell fate decisions leading to suboptimal humoral immunity. These studies revealed a critical GC checkpoint disrupted by chronic infection specifically at the point of dark onez re-entry. During chronic viral infection, virus-specific GC B cells were shunted towards terminal plasma cell (PC) or memory B cell (MB) fates at the expense of continued participation in the GC. Early GC exit was associated with decreased B cell mutational burden and antibody quality. Mechanistically, persisting antigen and inflammation independently drove facets of dysregulation, with a key role for inflammation in directing premature terminal differentiation. Thus, these studies define GC defects during chronic viral infection and identify a critical GC checkpoint that is short- circuited, preventing optimal maturation of humoral immunity.

Degree Type Dissertation

Degree Name Doctor of Philosophy (PhD)

Graduate Group Immunology

First Advisor E. John Wherry

Keywords B cells, Cellular differentiation, Chronic viral infection, Germinal Center, Inflammation, Single cell RNAseq

Subject Categories Allergy and Immunology | Immunology and Infectious Disease | Medical Immunology

This dissertation is available at ScholarlyCommons: https://repository.upenn.edu/edissertations/3637 REGULATION OF EFFECTIVE B CELL RESPONSES TO CHRONIC INFECTION Ryan P. Staupe A DISSERTATION in Immunology Presented to the Faculties of the University of Pennsylvania in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy 2019 Supervisor of Dissertation ______E. John Wherry, Ph.D. Chair, Department of Systems Pharmacology and Translational Therapeutics, Richard and Barbara Schiffrin President’s Distinguished Professor Graduate Group Chairperson ______David Michael Allman, Ph.D. Professor of Pathology and Laboratory Medicine

Dissertation Committee Michael Paul Cancro, Ph.D., Professor of Pathology and Laboratory Medicine David Michael Allman, Ph.D., Professor of Pathology and Laboratory Medicine Daniel Douek, M.D., Ph.D., Chief, Human Immunology Section, VRC, NIAID, NIH Jorge Henao-Mejia, M.D., Ph.D., Assistant Professor of Pathology and Laboratory Medicine Christopher A. Hunter, B.Sc, Ph.D., Distinguished Professor of Pathobiology

REGULATION OF EFFECTIVE B CELL RESPONSES TO CHRONIC

INFECTION

2019

Ryan Phillip Staupe

This work is licensed under the

Creative Commons Attribution-

NonCommercial-ShareAlike 4.0

License

To view a copy of this license, visit https://creativecommons.org/licenses/by-nc-sa/4.0/us/ ACKNOWLEDGMENTS

My path through graduate school has been one of tremendous personal and scientific growth that would not have been possible without the support of those around me and, in particular, my mentor E. John Wherry. Thank you John for always pushing me to think deeper, edit more, take risks, and be generous with my time. Your passion for science and clarity of thought has been crucial to my growth as a scientist.

Next, I would like to thank the members of my thesis committee for their invaluable advice and support over the past few years. Thank you Michael Cancro, David Allman, Daniel Douek, Jorge Henao-Mejia, and Christopher Hunter. Your guidance has been instrumental for shaping the science contained within this dissertation.

A special thank you to the Penn IGG program for providing a rewarding and challenging environment in which to learn and grow as a scientist. In particular, I’d like to thank Mary Taylor for going above and beyond in ensuring that our lives run smoothly. Your hard work and dedication are something to be emulated.

Nothing in this dissertation would have been achieved if not for the incredible and dynamic members of the Wherry lab, both past and present, who have made coming to lab something I look forward to daily. In particular, I’d like to thank Pam Odorizzi and Kristen Pauken for showing me the ropes at a time when I truly knew nothing. Thank you to Sasi Manne, Josephine Giles, Ramin Herati, and Amy Baxter for your endless patience, scientific insight, and friendship over the years. Thank you to Omar Khan, Jennifer Wu, Zeyu Chen, and Jane Huang for being the most incredible group of graduate students anywhere and for all the laughs along the way. Finally, I’d like to thank Laura Vella who without her encouragement, patience, and generosity none of this work would have been possible. Finally, I’d like to thank my family for their love, support, and guidance during this journey. To my parents, thank you for instilling within me a drive and a passion for learning. Your hard work, dedication, and sacrifices over the years have allowed me to become the person I am today and for that I will be eternally thankful. To my wife Jenab, thank you for being my rock. Without your kindness, inspiration, and support over the years I would not be at this point today. I love you more than I can possibly put into words.

iii

ABSTRACT

REGULATION OF EFFECTIVE B CELL RESPONSES TO CHRONIC

INFECTION

Ryan P. Staupe

E. John Wherry

Chronic viral infections disrupt B cell responses leading to impaired affinity maturation and delayed control of viremia. Previous studies have identified early pre-germinal center (GC) B cell attrition but the impact of chronic infections on B cell fate decisions in GCs remains poorly understood. To address this question, we used single-cell transcriptional profiling of virus-specific GC B cells during chronic viral infection to test the hypothesis that chronic viral infection disrupted

GC B cell fate decisions leading to suboptimal humoral immunity. These studies revealed a critical GC checkpoint disrupted by chronic infection specifically at the point of dark zone re-entry. During chronic viral infection, virus-specific GC B cells were shunted towards terminal plasma cell (PC) or memory B cell (MB) fates at the expense of continued participation in the GC. Early GC exit was associated with decreased B cell mutational burden and antibody quality.

Mechanistically, persisting antigen and inflammation independently drove facets of dysregulation, with a key role for inflammation in directing premature terminal differentiation. Thus, these studies define GC defects during chronic viral

iv infection and identify a critical GC checkpoint that is short-circuited, preventing optimal maturation of humoral immunity.

TABLE OF CONTENTS

ACKNOWLEDGMENTS ...... iii ABSTRACT ...... iv LIST OF FIGURES ...... viii CHAPTER 1: Introduction ...... 1 1.1 Chronic viral infections: a unique challenge for humoral immunity ...... 1 1.2 Humoral immune responses to acute infection and immunization ...... 3 1.2.1 Initiation of the B cell response ...... 4 1.2.2 Establishment of the germinal center response ...... 7 1.2.3 Germinal center response ...... 8 1.3 Humoral Immunity to chronic viral infections ...... 15 1.4 LCMV model system as a tool to mechanistically study chronic viral infections ...... 17 CHAPTER 2: Chronic viral infections impair virus-specific germinal center responses .. 22 2.1 Introduction ...... 22 2.2 Methods ...... 24 2.2.1 Mice ...... 24 2.2.2 Viral infection ...... 25 2.2.3 LCMV-NP protein production and purification ...... 25 2.2.4 Single cell suspension preparation and LCMV-specific B cell staining and enrichment...... 26 2.2.5 Flow cytometry and cell sorting ...... 27 2.2.6 Immunofluorescent microscopy ...... 28 2.2.7 Antibody quantification and avidity ELISA ...... 29 2.2.8 B cell receptor sequencing (BCRseq) ...... 30 2.2.9 Bulk RNAseq sample preparation ...... 32 2.2.10 Bulk RNAseq data processing and analysis ...... 33 2.3 Results ...... 34 2.3.1 Chronic infection promotes terminal B cell differentiation ...... 34 2.3.2 Affinity maturation is impaired during chronic infection...... 43 2.3.3 Distinct molecular profiles of LCMV-specific GC, PC and MB during chronic infection...... 49 2.4 Discussion ...... 56 CHAPTER 3: Transcriptional regulation of germinal center responses during acute and chronic viral infection ...... 59 3.1 Introduction ...... 59 3.2 Methods ...... 61 3.2.1 Mice ...... 61 3.2.2 Infections and antibody treatments...... 61 3.2.3 LCMV-NP protein production and purification ...... 62 3.2.4 Single cell suspension preparation and LCMV-specific B cell staining and enrichment...... 62 3.2.5 Flow cytometry and cell sorting ...... 63

3.2.8 Single cell RNAseq sample preparation ...... 64 3.2.9 Single cell RNAseq data processing and analysis ...... 65 3.3 Results ...... 67 3.3.1 scRNAseq reveals germinal center organization...... 67 3.3.2 Chronic infection promotes terminal B cell differentiation and GC exit...... 73 3.3.3 Chronic inflammation and altered antigen sensing in GC responses during chronic viral infection...... 84 3.4 Discussion ...... 95 CHAPTER 4: Concluding Remarks and Future Directions ...... 98 4.1 Overview of results ...... 98 4.2 The role of chronic antigen in the regulation of the humoral immune responses to chronic infection ...... 101 4.3 Inflammation: Dominant role of persistent inflammation over persistent antigen as a driver of B cell dysregulation in chronic infection ...... 106 4.4 B cell dysfunction in other settings of chronic antigen exposure – similarities and differences in lessons learned from chronic LCMV ...... 108 4.5 Impact of these studies on approaches to vaccination ...... 111 4.6: Adding a fourth dimension - Spatial dysregulation at the tissue level during chronic infection is a big black box and may underly many aspects of adaptive immune dysfunction during chronic infection ...... 113 4.7 Concluding thoughts ...... 116 BIBLIOGRAPHY ...... 118

vii

LIST OF FIGURES

Figure 2.1: Chronic infection promotes terminal B cell differentiation.

Figure 2.2: Gating and sort strategies for analysis of LCMV-specific B cells

Figure 2.3: LCMV-specific B cell responses are diminished during chronic

infection. Related to Figure 2.1.

Figure 2.4: Affinity maturation is impaired during chronic infection.

Figure 2.5: Chronic infection impairs affinity maturation but BCR repertoire

usage is largely unaffected. Related to Figure 2.4

Figure 2.6: Distinct molecular profiles of LCMV-specific GC, PC and MB during

chronic infection.

Figure 2.7: Transcriptional regulation of LCMV-specific GC B cells, PC, and MB

is temporally dynamic

Figure 3.1: scRNAseq reveals germinal center organization.

Figure 3.2: Unsupervised clustering of LCMV-specific GC B cells identifies

clusters with unique biology and known GC B cell function.

Figure 3.3: Chronic infection promotes terminal B cell differentiation and GC exit.

Figure 3.4: Pseudotime analysis identifies GC B cell fate decision points within

the LZ and correlated gene expression signatures.

Figure 3.5: BCR signaling pathways are altered in LZ GC B cells during chronic

infection.

Figure 3.6: LCMV-specific B cell differentiation is not sensitive to CD40 pathway

modulation during chronic infection.

viii

Figure 3.7: Inhibition of BCR signaling does not impact LCMV-specific B cell

responses.

Figure 3.8: Chronic antigen and inflammation impair the germinal center

response during chronic viral infection.

CHAPTER 1: Introduction

1.1 Chronic viral infections: a unique challenge for humoral immunity

Chronic viral infections are a ubiquitous part of human life. It is estimated that each of the 7.5 billion people on our planet is infected with 8-10 chronic viral infections (Virgin et al., 2009). While most of these infections cause little to no clinical disease, some, like human immunodeficiency virus (HIV) and hepatitis C virus (HCV), cause significant morbidity and mortality worldwide (WHO, 2019).

Evasion and subversion of the adaptive immune system is fundamental to the establishment of chronicity during an infection. Chronic viral infections, which are defined by the presence of chronic antigen and persistent inflammation, represent a particular challenge for the adaptive immune system which must balance pathogen control with host immunopathology in establishing homeostasis (Virgin et al., 2009). This balance between viral clearance and immunopathology results in adaptive immune responses that are altered—and typically less functional—than are adaptive responses to acute infection.

The study of adaptive immune dysregulation in chronic viral infections has largely focused on dysregulation of the CD8 T cell response. The phenotype of

CD8 T cells exposed to chronic viral infection and chronic antigen is termed exhaustion and has been the basis for revolutionary basic and translational immunology. Work on CD8 T cells in chronic viral infections and chronic cancer

1 antigen exposure has defined key mechanisms of CD8 T cell dysfunction at the protein, transcriptional, and epigenetic levels. The discoveries in this field have now resulted in the current successes of cancer immunotherapy with tremendous potential for continued basic scientific knowledge and clinical impact.

Humoral defects during chronic infection are also well described, but much less is known about the mechanisms by which chronic viral infections impact the

B cell response. Hypergammaglobulinemia has long been noted as a hallmark of chronic viral infections such as human immunodeficiency virus (HIV) (Lane et al.,

1983; Milito et al., 2004; Moir and Fauci, 2017), hepatitis C virus (HCV) (Bjøro et al., 1994; Cashman et al., 2014; Racanelli et al., 2006), and during infection with the chronic strain of the lymphocytic choriomeningitis virus (LCMV) in mice

(Hunziker et al., 2003; Oldstone and Dixon, 1969). However, despite increased total antibody titers, the antibodies made during chronic viral infections are poor quality, displaying lower affinity maturation compared to acutely resolved infection and delayed development of virus-neutralizing function (Battegay et al.,

1993; Gray et al., 2007; Logvinoff et al., 2004). In addition, vaccine-induced antibody responses are worse in individuals with pre-existing chronic viral infections compared to healthy control subjects (McKittrick et al., 2013; Shive et al., 2018; Wiedmann et al., 2000). Together, these observations suggest that chronic viral infection impairs B cell differentiation and fate decisions, leading to dysregulated humoral immunity. Understanding how chronic viral infections alter the normal humoral immune response can therefore (i) define targets to

2 modulate defective humoral immunity and (ii) give mechanistic insight into the signals that govern the B cell response.

1.2 Humoral immune responses to acute infection and immunization

The idea that chronic viral infection leads to dysregulated humoral immunity is based on comparisons made to the canonical, effective humoral immune response to acute infection and immunization. Effective humoral immunity is the establishment of antibodies and B cell memory that both control ongoing infection and, as is the goal in vaccination, prevent establishment of infection upon re-exposure to the pathogen (Cyster and Allen, 2019; Plotkin,

2008). The humoral immune response is initially defined by the induction of low- affinity antibodies (MacLennan et al., 2003) that bind their target specifically and well, and is eventually defined by the generation of plasma cells that secrete matured antibody for years as well as the generation of B cell memory (Slifka et al., 1998). Tight regulation of B cell differentiation into these tasks is critical for achieving effective and long-lasting humoral immunity. Failure to properly regulate humoral immune responses to stimuli can lead to vaccine failure or loss of viral control (Barnett et al., 2016; Chen et al., 2018; Chou et al., 2016; Haynes et al., 2012; Szczawinska-Poplonyk et al., 2015). This section will discuss the initiation of the B cell response to acute infection or immunization as well as our

3 current understanding of the signals that govern B cell differentiation and antibody formation.

1.2.1 Initiation of the B cell response

B cell differentiation begins in the bone marrow, where cells recombine their VDJ gene segments to arrive at a novel B cell receptor (BCR) arrangement

(Hardy and Hayakawa, 2001). The resulting naïve B cells—which express a functional, rearranged antigen receptor—then traffic to secondary lymphoid organs (SLO) such as the spleen and lymph nodes (Junt et al., 2008; Pereira et al., 2010). The majority of naïve B cells in SLO reside within the B cell follicles, where they await activation and the signals to undergo competitive testing and mutation of the BCR to produce high quality antibody in a process known as affinity maturation (Cyster and Allen, 2019). However, not all effective humoral immunity occurs as a result of affinity maturation. Instead, the first humoral response occurs as a result of a second population of naïve B cells in the spleen, termed marginal zone (MZ) B cells. MZ B cells are an innate-like population that express high levels of IgM and CD21, have lowered activation thresholds, and represent the first line of defense against microbial infection (Pillai et al., 2005).

MZ B cells are strategically positioned around the red pulp of the spleen and represent a major source of IgM in the nascent antibody response. MZ B cells therefore play a critical role in bridging the gap between the early antibody response and the activation of follicular (Fo) B cells to produce high affinity antibodies. 4

In contrast to their non-circulating MZ cousins, most B cells are Fo B cells and express high levels of IgD, CD23, and CXCR5 (Batista and Harwood, 2009).

Naïve Fo B cells maintain the ability to circulate through the body’s SLOs looking for antigen, but as their name indicates, Fo B cells can organize into micro- anatomical structures known as B cell follicles. The follicular structure is formed in response to stromal cell secretion of the chemokine CXCL13, which is the ligand for the Fo B cell chemokine receptor CXCR5 (Pereira et al., 2010). Proper organization of naïve B cells with SLOs is critical for efficient encounter of B cells with antigen and proper induction of humoral immunity (Allen et al., 2004; Ansel et al., 2000). The remainder of this thesis will therefore focus on the follicular B cell, and ‘B cell’ hereafter will refer to these follicular populations.

Induction of the Fo B cell response begins with the flow of foreign antigen into SLO (Batista and Harwood, 2009). Naïve B cell engagement of antigen through the rearranged BCR initiates a complex series of processes that ultimately define the fate of each B cell. Upon cognate antigen encounter, naïve

B cells proliferate, capture the antigen, and process it for display on major histocompatibility complex class II (MHCII) (Batista and Neuberger, 2000; Kwak et al., 2019) Additionally, recently activated B cells upregulate costimulatory receptors (CD80 and CD86), cell adhesion molecules (ICAM1), and chemokine receptors (CCR7) (Cyster and Allen, 2019). Increased expression of CCR7 allows for B cell migration towards the border with the T cell zone and enhances interaction with CD4+ T cells (Reif et al., 2002). While B cell differentiation can happen in a T cell-independent manner (Allman et al., 2019), effective antibody 5 responses to viral infection or immunization are generally T cell-dependent

(Crotty, 2019).

Interaction with T cells at the T:B border is the first differentiation checkpoint that B cells face. At this checkpoint, B cells enter either the extrafollicular response or the intrafollicular response, where affinity maturation occurs. In extra follicular responses, naive B cells undergo little BCR mutation and instead directly transform into plasma cells (PCs) that secrete antibody

(MacLennan et al., 2003). In this path, speed and quantity is prioritized over antibody quality as extrafollicular responses are responsible for fast production of antibody which is not class switched (i.e. mostly IgM) and of low affinity (Kwak et al., 2019; Takahashi et al., 1998). However, the majority of B cell biology that is important for control of chronic infection and for the establishment of high quality, long-lived humoral immunity requires affinity maturation during the intrafollicular response. The intrafollicular process of affinity maturation occurs in a highly specialized and coordinated micro-anatomical structure known as the germinal center (GC) (Berek et al., 1991; Jacob et al., 1991). The exact signals regulating this checkpoint are currently unclear, but commitment to the GC pathway is known to require prolonged and sustained interaction with antigen-specific CD4 T cells at the T:B border and subsequent expression of the GC master transcriptional regulator Bcl6 (Kerfoot et al., 2011; Kitano et al., 2011).

Expression of Bcl6 leads to downregulation of CCR7 and upregulation of S1PR2 on recently activated B cells and therefore allows B cells (in partnership with the

6 responding T cells) to progress deeper into the follicle and to establish a nascent

GC (Haynes et al., 2007; Shaffer et al., 2000).

1.2.2 Establishment of the germinal center response

The GC response to immunization or infection evolves over time and typically begins during the first 7-10 days of an immune response. In addition to

GC B cell precursors and T cells (now termed T follicular helper cells or Tfh), The

GC contains specialized populations of stromal cells that become activated and begin to proliferate. These stromal cells include the CXCL13-expressing follicular dendritic cells (FDC) and the CXCL12-expressing reticular cells (CRC) (Cyster et al., 2000; Rodda et al., 2015, 2018). Together, expansion of the B cells, Tfh, and stromal cells leads to physical displacement of naïve B cells, to the formation of the follicular mantle zone, and to the physical establishment of the GC. Beyond their role in recruiting GC B cells and Tfh via CXCL13 expression, FDC have high expression of complement receptor I (CD35) and therefore serve as the major antigen depot for the GC response (Heesters et al., 2014, 2013). Chemokine expression from the two stromal cell populations is critical for the polarization of the GC into a T cell-, FDC-, Tfh-, and antigen-rich light zone (LZ) and a CRC- rich, CXCL12-rich, T cell-poor dark zone (DZ) (Allen et al., 2004; Wang et al.,

2011) GCs are not considered fully matured until this polarization of the GC is established.

1.2.3 Germinal center response

Mature GCs are the primary location of antibody affinity maturation, a cyclic process by which B cells mutate their immunoglobulin genes, compete for antigen and Tfh cell help, and undergo clonal expansion following positive selection (Mesin et al., 2016). Affinity maturation occurs in two distinct steps which are uniquely partitioned between the LZ and the DZ (Victora et al., 2010,

2012). Within the LZ, GC B cells undergo positive selection wherein they capture antigen from FDCs, process and display it on MHCII (pMHCII), and present it to

Tfh (Schwickert et al., 2011; Victora et al., 2010) . Productive pMHCII:TCR interactions between B cells and Tfh induce expression of the pro-survival cytokines IL-4 and IL-21 which are sensed by their cognate receptors on GC B cells (Chevrier et al., 2017; Reinhardt et al., 2009; Weinstein et al., 2016; Yusuf et al., 2010; Zotos et al., 2010). Survival signaling through these cytokine pathways is vital for GC B cells; GC B cells are intrinsically pro-apoptotic, and cells that don’t receive pro-survival cytokines will apoptose (Mayer et al., 2017).

In addition to cytokine-derived survival signals, Tfh are the primary source of

CD40 signals to GC B cells (Foy* et al., 1996; Foy et al., 1994). CD40 signals from Tfh cells are essential in the GC reaction and (Kawabe et al., 1994) lead the

GC B cell to upregulate c-Myc and the MTORC1 signaling pathway, thereby allowing migration to the DZ and entry into cell cycle for division and accompanying mutation (Dominguez-Sola et al., 2012; Ersching et al., 2017; Luo et al., 2018).

Within the DZ, GC B cells undergo somatic hypermutation wherein in they mutate their immunoglobulin genes, re-express BCR, and—through mechanisms not yet understood—test for productive BCR signaling before returning to the LZ for further positive selection (Mcheyzer-Williams et al., 2015; Mesin et al., 2016;

Stewart et al., 2018; Victora et al., 2010). Mutation of immunoglobulin genes is accomplished by upregulation of AID, a cytosine deaminase that binds to variable regions of the immunoglobulin genes and deaminates cytosines

(Muramatsu et al., 2000, 2002; Wang et al., 2017). Repair of these DNA lesions by cellular DNA repair machinery (mismatch repair and base excision repair), results in the introduction of a DNA base somatic mutation (Odegard and Schatz,

2006). It is estimated that GC B cells make 1 somatic mutation per 103 bases during each round of cell division (Kleinstein et al., 2003). Mutations can result in changes to the protein sequence of the BCR (replacement mutations) or no change BCR protein sequence (silent mutations). However, not all mutations made by AID are beneficial. Mutations can result in productive BCRs that are able to signal or non-productive BCRs that cannot (Mayer et al., 2017). Although not yet fully understood, testing BCR signaling productivity is a key checkpoint regulating GC B cell DZ exit and LZ reentry, and GC B cells with non-productive

BCRs ultimately undergo apoptosis (Stewart et al., 2018).

Affinity maturation is ultimately achieved by linking cell division within the

DZ to the LZ process of intra- and inter-clonal competition for antigen and Tfh help. Competition between GC B cells is achieved by a compounding process.

First, the amount of antigen captured by GC B cells from FDC is directly 9 proportional to the affinity of the BCR (i.e. higher affinity BCR captures more antigen) (Gitlin et al., 2014). Second, GC B cells that have captured more antigen then process and present more pMHCII to Tfh, thereby inducing stronger help signals (i.e more CD40 signals) (Gitlin et al., 2014; Shulman et al., 2013; Victora et al., 2010). Stronger help signals from Tfh cause higher affinity GC B cells to undergo more cell divisions within the DZ in a shorter period of time, leading to an accumulation of more mutations than lower affinity GC B cells (Gitlin et al.,

2014, 2015). However, how antigen capture and Tfh cell help in the LZ molecularly translate into DZ reentry and increased GC B cell divisions is still only partially understood.

The current model for how positive selection and DZ reentry are linked has recently been summarized by Shlomchik et. al (Shlomchik et al., 2019). They propose that BCR signaling by GC B cells leads to the phosphorylation and inactivation of the transcription factor FOXO1 through the P13K-Akt pathway

(Luo et al., 2018), leading to loss of CXCR4 expression and retention of GC B cells within the LZ (Dominguez-Sola et al., 2015; Inoue et al., 2017; Sander et al.,

2015). Antigen capture and presentation followed by GC B cell interaction with

Tfh leads to CD40 stimulation. Unlike in naive B cells, where NFkB is downstream of BCR signaling, in the GC B cells, NfKB becomes downstream of

CD40 ligation (Luo et al., 2018). The CD40L-induced NFkB expression leads to transcription of c-Myc and entry into cell cycle (Ersching et al., 2017; Luo et al.,

2018). The degree of c-Myc expression has been observed to tightly correlate with cell proliferation; cells with more c-Myc are able to undergo more cell 10 divisions suggesting that c-Myc may act as a cell division timer within GC B cells

(Calado et al., 2012; Dominguez-Sola et al., 2012; Finkin et al., 2019).

Importantly, GC B cells require both a BCR signal and a CD40 signal to undergo positive selection (Luo et al., 2018). CD40 induced c-Myc expression is directly repressed by the transcription factor FOXO1 but in the presence of BCR signals

FOXO1 is inactivated allowing c-Myc expression (Luo et al., 2018; Wilhelm et al.,

2016). Combined CD40 signals and BCR also induce activation of the MTORC1 pathway, as evidenced by increased expression of phospho-S6 and the accumulation of biomass following positive selection (Ersching et al., 2017). In this way, increased Tfh help is permitted along with BCR signaling to promote increased c-Myc expression in higher affinity GC B cells, allowing them to undergo more cell divisions in the DZ and leading to increased accumulation of somatic mutations in their immunoglobulin genes.

One aspect of GC B cell fate missing from this iterative model is that GC B cells can also differentiate into PC or MB from directly within the GC (Cyster and

Allen, 2019; Kurosaki et al., 2015; Mesin et al., 2016; Nutt et al., 2015; De Silva and Klein, 2015). The decision to leave the GC as a MB or PC or to continue participating in the GC occurs in the LZ and is the second major differentiation checkpoint that B cells face during a humoral immune response (Ise et al., 2018;

Kräutler et al., 2017; Laidlaw et al., 2017; Suan et al., 2017a). Exiting the GC response and undergoing differentiation prevents further participation in affinity maturation and suggests tight regulation of this cellular fate decision. Exiting the

GC reaction early as a memory B cell is likely beneficial as this would maintain a 11 relatively unmutated pool of antigen-specific MB that could mount a more diverse response to secondary challenge. Exiting the GC reaction as a PC is a more complicated decision. PC differentiation occurs in two distinct waves, one early after B cell encounter with a T cell at the T:B border and one much later, after the establishment of the GC (Shlomchik and Weisel, 2012; Weisel et al., 2016).

These two waves of PC differentiation have different BCR affinities, with the earlier wave having lower affinity than the later one (MacLennan et al., 2003;

Takahashi et al., 1998; Weisel et al., 2016). These two observations, that post-

GC MB are generally lower affinity and post-GC PC are generally higher affinity, suggests a link between affinity maturation of GC B cells and differentiation decisions within the GC.

The link between BCR affinity and GC B cell differentiation into PC or MB has been solidified in recent years due to the development of fate-tracking tools and identification of PC and MB precursor populations. Phan et. al. showed, using a PC fate tracking system (Blimp1-GFP) combined with BCR knock-in mice, that high affinity B cells are preferentially recruited into the PC pool (Phan et al., 2006). How high BCR affinity translates molecularly into PC differentiation is just beginning to be understood. One factor that influences GC B cell differentiation into PC is the transcription factor IRF4 (Klein et al., 2006; Ochiai et al., 2013; Sciammas et al., 2006; Xu et al., 2015). IRF4 is induced downstream of the BCR in a dose dependent way, with higher stimulation leading to more IRF4

(Ochiai et al., 2013). IRF4 at low intracellular concentrations is able to bind to and induce transcription and expression of Bcl6 and AID (Ochiai et al., 2013). 12

However, IRF4 at high intracellular concentrations is able to bind to and induce transcription of the PC master transcriptional regulator Blimp1 leading to repression of Bcl6 and PC differentiation (Ochiai et al., 2013). In this way, strong signals through the BCR can lead to PC differentiation. More recently, GC B cell commitment to the PC fate has also been shown to require CD40 signals from

Tfh help in addition to signals from antigen/BCR alone (Luo et al., 2018).

In contrast to PC, MB originate from low affinity GC B cells. Several groups have recently described MB precursors in the GC B cell population

(Laidlaw et al., 2017; Suan et al., 2017a); however, a consensus set of markers defining this population has yet to be proposed. One shared feature amongst the identified precursor populations is the upregulation of chemokine receptors consistent with GC egress and low BCR affinity (Laidlaw et al., 2017; Suan et al.,

2017a). GC B cell differentiation into MB is associated with expression of the transcriptional repressor Bach2 (Shinnakasu et al., 2016). Expression of Bach2 in GC B cells is inversely correlated with the strength of Tfh signaling

(Shinnakasu et al., 2016). Bach2 expression levels provide a potential link between BCR affinity, Tfh cell help, and MB differentiation and suggest a model where low strength BCR signals and T cell help promote MB differentiation.

The necessity for Tfh signals in GC B cell positive selection and MB and

PC differentiation suggests a “Goldilocks” model for GC B cell differentiation that is dependent on BCR affinity/signal strength (Ise and Kurosaki, 2019). GC B cells with high affinity BCRs undergo selection into the PC compartment whereas GC

B cells will low affinity BCRs undergo differentiation into MB. GC B cells with 13 moderate BCR affinity undergo positive selection and DZ reentry for further rounds of affinity maturation. How the threshold is set for what signals are too high or too low is currently unclear and whether/how these thresholds change over the course of an immune response will require further study.

The crux of all B cell fate decisions, before compounding by differences in

Tfh help and division potential, is the rearranged and subsequently somatically hypermutated BCR. The BCR has two primary functions, as both a signaling molecule and as an endocytic receptor (Kwak et al., 2019; Shlomchik et al.,

2019). It is likely that both the endocytic and signaling properties of the BCR are important in regulating GC B cell fate decisions, because GC B cell positive selection and differentiation decisions involve both transcriptional changes and presentation of antigen. However, despite the central role of BCR and its binding of antigen, it is unclear how antigen availability impacts the competitive GC process. Indeed, most studies investigating regulation of GC B cell responses have focused on studies of immunization (Suan et al., 2015, 2017b). In contrast to immunization, where antigen is introduced as a bolus and levels of antigen decay over time, antigen during acute infection logarithmically increases over time (Virgin et al., 2009). How increasing levels of antigen impact the BCR- affinity based model of GC B cell differentiation and selection is unclear. Immune responses to chronic infection are yet another case, where antigen persists even longer than in acute infection and far longer than in the case of traditional immunization (Virgin et al., 2009). The relative role of prolonged antigen on B cell fate decisions is a major topic of this thesis. However, abundant and prolonged 14 antigen is not the only feature of chronic infection, and chronic inflammation likely also plays a role in changing humoral immunity (Virgin et al., 2009). The next sections of the introduction will therefore discuss what is known about the effects of chronic infection on the humoral immune response and describe the mouse model used to interrogate chronic infection more broadly.

1.3 Humoral Immunity to chronic viral infections

Effective humoral immune responses to infection and immunization are defined by antibodies that bind pathogens and function to both prevent and, to a lesser extent, control infection (Cyster and Allen, 2019; Dörner and Radbruch,

2007; Plotkin, 2008). Optimal antibody function requires that the antibodies have high affinity for their pathogenic target. In contrast to humoral immune responses to acute viral infection, immune responses to chronic infection are often dysfunctional, characterized by increased antibody production and delayed development of virus-neutralizing function (Battegay et al., 1993; Bjøro et al.,

1994; Cashman et al., 2014; Eschli et al., 2007; Gray et al., 2007; Hunziker et al.,

2003; Lane et al., 1983; Logvinoff et al., 2004; Milito et al., 2004; Moir and Fauci,

2017; Oldstone and Dixon, 1969; Racanelli et al., 2006). How chronic viral infections dysregulate and evade antibody-mediated immunity is currently unclear. Limiting access to neutralizing epitopes (Klein and Bjorkman, 2010; Moir and Fauci, 2017; Sommerstein et al., 2015), high viral mutation rates (Bonsignori

15 et al., 2017), establishment of a latent reservoir (Barton et al., 2011), and direct infection of B cells (Barton et al., 2011) are just some of the immune evasion strategies used by chronic viral infections to blunt anti-viral antibody responses

(Virgin et al., 2009).

Although, dysregulated humoral immunity has long been appreciated during chronic viral infections, our understanding of the key steps in B cell differentiation and underlying mechanisms of the dysregulation remains limited.

Recent studies have shown a role for type I interferon in driving B cell differentiation away from the GC fate resulting in the acute depletion of B cells during chronic viral infection before GC B cell differentiation can be established

(Fallet et al., 2016; Moseman et al., 2016; Sammicheli et al., 2016). Moreover, chronic viral infections in mice and humans can result in lymphoid tissue disruption and/or fibrosis (Mueller et al., 2007; Schacker et al., 2006; Wilson et al., 2013a), perhaps impacting B cell responses. However, despite these challenges, GC still form during chronic infection (Barnett et al., 2016; Clingan and Matloubian, 2013; Daugan et al., 2016; Fukazawa et al., 2015), and antibodies have been shown to diversify over time during chronic HIV and HCV infections (Doria-Rose et al., 2014; Kinchen et al., 2018; Wu et al., 2015), suggesting GC activity. Moreover, despite generally smaller GCs during chronic infection, affinity-matured antibodies produced as a result of the GC response exert critical control over viral loads and, in some cases, lead to viral control

(Chen et al., 2018; Chou et al., 2016; Doria-Rose et al., 2016; Kinchen et al.,

2018; Seiler et al., 1998). Additionally, non-neutralizing antibodies that 16 incompletely control viral replication are required to partially contain infection

(Hangartner et al., 2006; Horwitz et al., 2017; Richter and Oxenius, 2013; Straub et al., 2013) and can apply selective pressure for viral escape and evolution in some settings (Doria-Rose et al., 2014; Kinchen et al., 2018; Wu et al., 2015).

Thus, humoral immune responses that occur during chronic infections likely involve considerable participation of GC B cell responses. Yet, the biology governing GC B cell differentiation and fate decisions in the setting of chronic viral infections remains to be elucidated. A better understanding of dysregulated humoral immunity and GC biology during chronic viral infections could provide valuable insights to improve vaccination and antibody responses during chronic viral infections.

1.4 LCMV model system as a tool to mechanistically study chronic viral infections

Humoral immune dysfunction during chronic viral infections in humans is well documented and generally defined by the increased production of less avid antibodies. These observations and additional alterations in peripheral B cell subsets (Moir et al., 2008; Oliviero et al., 2015) during chronic viral infection suggest that chronic viral infection may alter B cell differentiation within SLOs and in particular the GC response. While study of serum antibody and peripheral

B cell subsets have been instrumental in defining humoral immune dysfunction

17 during chronic infection, mechanistic study of how chronic viral infection impacts the GC response has been limited by access to lymphoid tissues in humans.

Thus, modeling chronic viral infection in experimental animal models, such as mice, provides a more tractable system for increased mechanistic understanding of the GC response to chronic viral infection.

The LCMV mouse model of viral infection has been an invaluable tool for understanding anti-viral adaptive immune responses and host-pathogen interactions. LCMV is a single-stranded negative sense RNA virus of the family

Arenaviridae whose natural reservoir is the mouse. The LCMV genome is bisegmented and consists of coding segments for 5 viral proteins (Urata and

Yasuda, 2012). Viral entry into cells is mediated through binding of the LCMV glycoprotein GP-1 to its receptor, a-dystroglycan (Cao et al., 1998; Kunz, 2009).

Due to the broad expression of a-dystroglycan by cells in many tissues, infection with LCMV is systemic.

Many strains of LCMV exist, but perhaps the most widely used are LCMV-

Armstrong (LCMV-Arm, acute) which leads to an acute resolving infection and

LCMV-Clone13 (LCMV-Cl13, chronic) which leads to a persistent infection

(Ahmed, 1984; Wherry et al., 2003a). The acute and chronic strains of LCMV differ by only three amino acids (Matloubian et al., 1990, 1993; Sullivan et al.,

2011), but these mutations allow chronic LCMV to evade immunity and establish chronicity. A mutation in the LCMV polymerase protein (L protein) allows chronic

LCMV to replicate faster in dendritic cells and macrophages (Matloubian et al.,

1993). Two additional mutations in the viral glycoprotein GP-1 result in higher affinity binding of GP-1 to a-dystroglycan allowing increased early infection of

FDCs and disruption of lymphoid architecture (Matloubian et al., 1990; Sullivan et al., 2011).

Historically, the LCMV model system has been a useful tool to study how persistent antigen and chronic inflammation alter adaptive immune responses, particularly in CD8 T cells and within the field of T cell exhaustion (McLane et al.,

2019); however, little is known about how persistent antigen and inflammation impact humoral immunity. Humoral immune responses to chronic LCMV mirror many of the immune dysfunctions observed in humans with chronic viral infections such as HIV or HCV. Chronic LCMV infection causes hypergammaglobulinemia (Hunziker et al., 2003) and infected mice fail to make neutralizing antibodies until approximately 100 days post-infection (Battegay et al., 1993). Similar to HIV, LCMV uses glycan shielding to limit B cell and antibody access to the neutralizing determinant on GP-1 (Sommerstein et al., 2015).

Despite the delayed development of neutralizing antibodies during chronic LCMV infection, humoral immunity is critical for viral control as mice without CD4 T cells or lacking the antibody isotype IgG2a/c fail to control LCMV-CL13 viremia

(Barnett et al., 2016; Matloubian et al., 1994).

Our understanding of how LCMV-CL13 subverts the B cell response is still poorly understood due to the limited ability to track LCMV-specific B cell responses. Traditional study of humoral immunity has been focused on antigen-

19 specific antibodies, but much less attention has been paid to the B cells that produce those antibodies. This is evident even in the common naming of the two arms of the adaptive immune system, humoral and cellular. Of course, humoral immunity is entirely dependent on a form of cellular immunity. However, while antigen-specific T cell responses have long been studied with the use of fluorescently labeled pMHC tetramers, until recently B cell specificity remained limited to either serologic assays of the resulting secreted antibodies or of B cell

ELISPOTS that do not allow for any cellular characterization of the antibody- secreting cell. For B cells that do not yet secrete antibody, few antigen-specific assays exist. Therefore, one major barrier to the study of cellular aspects of humoral immunity is the lack of tools to identify virus-specific B cells.

Recent work using LCMV-specific BCR transgenic mice has identified an alteration of the pre-GC extrafollicular B cell differentiation checkpoint during chronic LCMV infection (Fallet et al., 2016; Moseman et al., 2016; Sammicheli et al., 2016). These studies found that chronic LCMV infection causes massive attrition of LCMV-specific B cell responses through promotion of plasma cell differentiation, extrafollicular localization, and subsequent deletion by inflammatory macrophages. However, despite alterations in this pre-GC checkpoint, mice infected with chronic LCMV still make GCs. Importantly, these

GC responses exert critical pressure during chronic LCMV infection as mice lacking factors critical for positive selection in the GC fail to control chronic

LCMV. Recently, recombinant protein tools have been developed that allow identification of LCMV-NP-specific B cell responses (Fallet et al., 2016; Schweier 20 et al., 2019; Sommerstein et al., 2015). Recombinant expression of LCMV-NP in bacterial systems allows tracking of LCMV-NP-specific B cells in vivo and allows more careful dissection of the mechanisms underlying humoral immune dysfunction during chronic infection. Thus, acute and chronic LCMV infection of mice is an ideal model for the investigation of how chronic infection alters B cell differentiation.

CHAPTER 2: Chronic viral infections impair virus-specific

germinal center responses

2.1 Introduction

Effective humoral immune responses to infection and immunization are defined by high-affinity antibodies generated as a result of B cell differentiation and selection that occurs within germinal centers (GC) (Cyster and Allen, 2019;

Plotkin, 2008). Within the GC, B cells undergo affinity maturation, an iterative and competitive process wherein B cells mutate their immunoglobulin genes (somatic hypermutation) and undergo clonal selection by competing for T cell help (Cyster and Allen, 2019; Mesin et al., 2016; Shlomchik et al., 2019). Balancing the decision to remain within the GC and continue participating in affinity maturation or to exit the GC as a plasma cell (PC) or memory B cell (MB) is critical for achieving optimal antibody avidity, antibody quantity, and establishing immunological memory in response to immunization or infection.

As discussed in section 1.3, humoral immunity in chronic viral infections is often dysregulated with hypergammaglobulinemia is a hallmark of chronic viral infections such as HIV (Lane et al., 1983; Milito et al., 2004; Moir and Fauci,

2017), HCV (Bjøro et al., 1994; Cashman et al., 2014; Racanelli et al., 2006), a finding that is consistent with the mouse model of infection with the chronic strain of LCMV (Hunziker et al., 2003; Oldstone and Dixon, 1969). Despite increased total antibody titers, the antibodies made during chronic viral infections are poor

22 quality, displaying lower affinity maturation compared to acutely resolved infection and delayed development of virus-neutralizing function (Battegay et al.,

1993; Eschli et al., 2007; Gray et al., 2007; Logvinoff et al., 2004). In addition, vaccine-induced antibody responses are worse in individuals with pre-existing chronic viral infections compared to healthy control subjects (McKittrick et al.,

2013; Shive et al., 2018; Wiedmann et al., 2000). Together, these observations suggest that chronic viral infection may impair B cell differentiation and fate decisions, leading to dysregulated humoral immunity.

While dysregulated humoral immunity during chronic infection has long been appreciated, our understanding of the key steps in B cell differentiation and underlying mechanisms of the dysregulation remains limited. As discussed in sections 1.4, despite data in mice suggesting that chronic viral infection leads to loss of the pre-GC B cell, many lines of evidence in human and mice suggest that humoral immune responses during chronic infections involve considerable participation of GC B cell responses (Chen et al., 2018; Chou et al., 2016; Doria-

Rose et al., 2014; Kinchen et al., 2018; Wu et al., 2015). Yet, the serologic output of the GC remains altered (Battegay et al., 1993; Doria-Rose et al., 2014;

Kinchen et al., 2018; Wu et al., 2015). A better understanding of dysregulated humoral immunity and GC biology during chronic viral infections could provide valuable insights to improve vaccination and antibody responses during chronic infections.

The studies in this chapter interrogate the impact of chronic infection on the development and differentiation of GC B cell responses. Using novel 23 recombinant protein tools to track and characterize the virus-specific B cell response, we find that chronic infection skews GC B cell differentiation away from the GC B cell fate and towards the PC and MB fates. Skewing away from the GC B cell fate during chronic infection was associated with increased virus- specific antibody titers but decreased serum antibody avidity and impaired affinity maturation. Transcriptional profiling of virus-specific B cell responses identified enrichment of the PC and MB gene signatures associated with inflammatory signaling in virus-specific GC B cells during chronic infection. Taken together, these results indicate that chronic infection skews B cell differentiation away from the GC B cell fate and towards the PC and MB fates leading to delayed development of effective antibody responses.

2.2 Methods

2.2.1 Mice

6-8 week old C57BL/6 Ly5.2CR (CD45.1) or C57BL/6 (CD45.2) mice were purchased from NCI. Both male and female mice were used for these studies in accordance with NIH guidelines. Mice were maintained in a specific-pathogen- free facility at the University of Pennsylvania. All mice were used in accordance with Institutional Animal Care and Use Committee guidelines for the University of

Pennsylvania.

2.2.2 Viral infection

Mice were infected intraperitoneally (ip) with 2*105 plaque-forming units

(PFU) of LCMV-Armstrong (Arm) or intravenously (iv) with 4*106 PFU LCMV-

CL13 (CL13). LCMV strains were propagated and titers were determined as previously described (Odorizzi et al., 2015; Wherry et al., 2003b).

2.2.3 LCMV-NP protein production and purification

Expression plasmids encoding a 6x His tagged version of the LCMV nucleoprotein (NP) (nucleoprotein sequences from LCMV-Arm and LCMV-CL13 have 100% homology) were a kind gift from D. Pinschewer (University of Basel,

Basel, Switzerland) and originally developed by H. Pircher (University of

Freiberg, Germany) and was used as previously described (Fallet et al., 2016;

Schweier et al., 2019; Sommerstein et al., 2015). Briefly, plasmid was transformed into Rosetta™ 2(DE3) competent cells (Novagen, Millipore Sigma).

Protein expression was induced with .5M IPTG (Sigma) for 24 hours. Bacteria were harvested after induction, lysed and sonicated. The soluble fraction was purified over a HiTrap TALON crude column (GE Healthcare) on a GE

AKTAprime plus FPLC system (GE Healthcare). Purified protein was desalted using PD-10 columns (GE Healthcare) and concentrated prior to use.

2.2.4 Single cell suspension preparation and LCMV-specific B cell staining and enrichment.

Single cell suspensions were obtained from spleens by homogenizing tissues against a 70um cell strainer. Cells were washed through cell strainer and red blood cells were lysed in ACK lysis buffer (Thermo Fisher Scientific). Cells were resuspended in MACS buffer (PBS with 1mM EDTA and 2% fetal bovine serum) prior to cell staining or B cell enrichment. To identify LCMV-specific B cells, purified LCMV-NP was fluorescently labeled with Alexa Fluor 647 or Alexa

Fluor 488 using antibody labeling kits according to manufacturer’s instructions

(Thermo Fisher Scientific). Single cell suspensions were stained with LCMV-NP for 15 minutes at 37°C in the presence of Fc-block (anti-CD16/32, 24G2, BD) and washed twice with MACS buffer prior to staining with additional fluorescent antibody panels. For bulk RNAseq experiments, B cell enrichment (negative selection) was performed using mouse pan-B cell isolation kits (Stem Cell

Technologies) according to manufacturer’s instruction except incubation times were doubled. Following B cell enrichment, LCMV-NP staining was performed as described above. LCMV-NP-labeled cell suspensions were further enriched

(positive selection) for LCMV-specific B cells by incubating cells with anti-

Cy5/AF647 microbeads (Miltenyi Biotec) followed by purification over LS columns

(Miltenyi Biotec).

2.2.5 Flow cytometry and cell sorting

Single cell suspensions were prepared and stained with fluorescently labeled LCMV-NP as described above. Cell suspensions were stained at 4°C for

30 minutes for surface protein staining. Intracellular staining was conducted at room temperature for 1 hour. Antibodies were purchased from Biolegend:

CXCR4 (L276F12), CD83 (Michel-19), CD86 (GL-1), B220 (RA3-6B2), CD19

(1D3), IgD (11-26c.2a), CD138 (281-2), GL7 (GL7), CD38 (90), F4/80 (BM8),

NK1.1 (PK136), IgM (RMM-1), Ki67 (16A8), CD69 (H1.2F3), Steptavidin-BV786;

BD Biosciences: CXCR4 (2B11), CD83 (Michel-19), CD19, (1D3), CD138 (281-

2), CD38 (90), CD4 (GK1.5), Blimp1 (5E7), Bcl6 (K112-91), CD31 (MEC 13.3),

CCR6 (140706), Streptavidin BB780; eBioscience (Thermo Fisher Scientific):

GL7 (GL7), CD4 (RM4-5), CD8 (53-6.7), F4/80 (BM8), NK1.1 (PK136); or from

Southern Biotech: Ly6a (D7). Live cells were discriminated by staining with

Live/Dead Aqua (Thermo Fisher Scientific) or Zombie Yellow (Biolegend). For cell sorting for BCR repertoire analysis, samples were sorted directly into lysis buffer (RLT or RLT+, Qiagen) or into 10% FBS cRPMI supplemented with 25mM

HEPES. Flow cytometry data were acquired on a BD LSR II instrument or BD

Symphony A3 instrument. Cell sorting was performed on a BD FACSAria enclosed within a laminar flow hood. Data were analyzed using FlowJo software

(FlowJo, LLC).

2.2.6 Immunofluorescent microscopy

Mouse spleen samples were acquired from LCMV-Arm and LCMV-CL13 infected mice at day 15 post-infection. Spleens were embedded in Optimal

Cutting Temperature (OCT, Tissue-Tek, VWR) and snap frozen using a 2- methylbutane (Sigma) bath over dry ice. 7um sections on slides were stained as follows. Slides were desiccated at room temperature (RT) for 5-10 minutes prior to fixation in 4% freshly prepared paraformaldehyde for 10 minutes at RT.

Following fixation, samples were rehydrated in PBS for 5 minutes followed by blocking for 1 hour in blocking buffer (1% bovine serum albumin (Sigma) and

0.3% Triton X-100 (Sigma)). Fluorescently-conjugated primary antibodies (IgD, clone 11-26c.2a, Biolegend and GL7, clone GL7, Thermo Fisher Scientific) were added and incubated for 2 hours at RT. Following incubation with primary antibodies, samples were wash 3x in PBS followed by coverslip mounting with

ProLong Diamond with DAPI (Thermo Fisher Scientific).

Images were acquired on an inverted DMi8 widefield microscope (Leica) using a 40x objective (NA 0.95). At least 20 GCs were acquired per sample and

GCs were sampled randomly across each tissue. Analysis of GC area was done in using the FIJI distribution (Schindelin et al., 2012) of ImageJ (Rueden et al.,

2017) through manual annotation of GC boundaries followed by area determination.

2.2.7 Antibody quantification and avidity ELISA

LCMV-specific IgG serum antibody titers were determined using standard

ELISA protocols. Briefly, 96-well MaxiSorp plates (Nunc, Thermo Fisher

Scientific) were coated with 300ng purified recombinant LCMV-NP per well overnight. For measurement of anti-LCMV-NP serum antibody concentration, serum samples were assayed in 5-fold dilutions starting from 1:100 dilution and plated in quadruplicate on each plate. LCMV-NP-specific IgG was detected using biotin-conjugated goat-anti-mouse IgG(H+L) (Southern Biotech) followed by incubation with streptavidin-HRP (Southern Biotech) and developed with tetramethylbenzidine (Thermo Fisher Scientific). Amount of antibody in each serum sample was calculated by fitting a standard curve generated using an anti-

LCMV-NP monoclonal antibody (VL4, BioXcell).

Antibody avidity was measured by ELISA with modifications (Nawa, 1992;

Pullen et al., 1986). For serum antibody avidity measurements, the anti-LCMV-

NP ELISA was modified to include a wash step with either PBS, 1.5M sodium thiocyanate (NaSCN), 3M NaSCN, or 4.5M NaSCN for 15 minutes at room temperature after incubation of serum samples but before application of secondary goat-anti-mouse IgG(H+L) antibody. Avidity index was calculated as follows: Avidity index = (OD450NaSCN/OD450PBS)*100. Serum samples for serum avidity measurements were diluted to a concentration yielding OD450 of approximately 1 prior to incubation with NP-coated ELISA plates.

2.2.8 B cell receptor sequencing (BCRseq)

All LCMV-NP-specific B cells were first sorted for yield then for purity from individual mice. Between 3.8k and 28k cells were obtained per sample. Samples were sorted into buffer RLT (Qiagen) + supplemented with b-mercaptoethanol

(Sigma) and processed with an AllPrep DNA/RNA Micro Kit (Qiagen) according to manufacturer’s recommended protocol. Genomic DNA was sent to the Human

Immunology Core at the University of Pennsylvania for mouse antibody heavy chain library construction, sequencing and data analysis. Primers used were adapted from Wang et. al (Wang et al., 2000) and were modified to include adaptor sequences for the Illumina NexteraXT kit and are as follows:

MH1 primer

5’-

TCGCGAGTTAATGCAACGATCGTCGAAATTCGCSARGTNMAGCTGSAGSAG

TC-3’;

JH1/4

5’-

ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCTANTGAGGAGACGGTG

AC-3’;

JH2

5’ACACTCTTTCCCTACACGACGCTCTTCCGATCTTGAGGAGACTGTGAGAG

TGG-3’.

Amplicons were purified using the Agencourt AMPure XP beads system

(Beckman Coulter, Inc., Indianapolis, IN), and quantified by Qubit Fluorometric

Quantitation (Thermo Fisher Scientific, Grand Island, NY). Libraries were loaded onto an Illumina MiSeq and 2x300 bp paired end kits were used for all experiments (Illumina MiSeq Reagent Kit v3, 600 cycle, Illumina Inc., San Diego,

Cat. No. MS-102-3003).

Raw sequence data (FASTQ files) were processed through pRESTO v0.5.3 (Vander Heiden et al., 2014) and any sequence with a 10 bp section that had an average Phred quality score lower than 30 was removed, along with any sequence that was shorter than 100 bases. Germline V and J genes were identified by aligning pre-processed data to the murine IgH germline repertoire using the IMGT/HighV-QUEST tool hosted by the International Immunogenetics

Information Systems (IMGT, imgt.org) (Alamyar et al., 2012; Lefranc et al., 2015;

Li et al., 2013). Aligned reads were further processed with Change-O v0.4.5

(Gupta et al., 2015). The Change-O processing pipeline included filtering non- functional sequences and clustering sequences into clonal lineages. Data were then filtered by calculating the mean frequency (copy number) of all clones and selecting for clonal lineages above 50% of the mean copy number frequency.

The frequency of mutation from germline, replacement to silent ratios (R/S ratio), and selection pressure (BASELINE, (Yaari et al., 2012)) were calculated using

SHazaM v0.1.11 (Gupta et al., 2015).

2.2.9 Bulk RNAseq sample preparation

To assess the kinetic transcriptional differences between LCMV-specific B cell subsets responding to LCMV-Arm or LCMV-CL13 infection, LCMV-specific

GC B cells (CD38-GL7+), memory B cells (CD38+GL7-), and plasma cells

(CD138+) were sorted from mice infected with LCMV-Arm or LCMV-CL13 at days 7, 15, and 45 post-infection. All B cell subsets were first gated as LCMV-

LCMV-SPECIFICIgD-B220+Dump- prior to subset-specific gating. Dump gate included CD4, CD8, NK1.1, and F4/80. Naïve B cells were sorted from uninfected mice. Samples were sorted first for yield then for purity. 5-10 spleens were pooled for each of the 3 replicates sequenced at each timepoint. Prior to sorting, samples were enriched for LCMV-specific B cells and stained with surface antibodies as previously described. Sorted samples were stored in Buffer

RLT supplemented with b-mercaptoethanol until RNA isolation. RNA isolation was performed with a RNeasy Mico Kit (Qiagen) per the manufacturer’s instructions. Library preparation was performed using a SMART-Seq v4 Ultra

Low Input RNA kit (Takara) and Nextera XT DNA Library Preparation Kit

(Illumina) per the manufacturer’s instructions. Quality of extracted RNA and prepared libraries were assessed using a TapeStation 2200 instrument (Agilent).

Libraries were quantified using a KAPA Library Quantification Kit (Kapa

Biosystems) prior to sequencing on an Illumina NextSeq 550 instrument (150bp, paired-end, 300 cycle high output flowcells).

2.2.10 Bulk RNAseq data processing and analysis

FASTQ files were aligned using STAR 2.5.2a (Dobin et al., 2013)to the

GRCm38-p5 murine reference genome (GENCODE, (Frankish et al., 2019)).

Unique mapped reads were normalized using PORT v0.8.4-beta gene-based normalization (https://github.com/itmat/Normalization). Following PORT normalization, all downstream processing was conducted using R (R Core Team,

2019).Differential gene expression analysis was performed using Limma (Ritchie et al., 2015). Limma-voom was used to identify significantly different genes between groups using p value < 0.05. Ig genes were removed from each sample prior to differential gene expression analysis. Optimal K values for K-means clustering was calculated using the consensus K values calculated using the

Elbow, Silhouette, and Gap Statistic methods of the NbClust package (Charrad et al., 2015). Heatmaps depict all differentially expressed genes (p value < 0.05 and log2 fold change > 2) within LCMV-specific B cell subsets between samples isolated from LCMV-Arm and LCMV-CL13 infected mice at days 7, 15, and 45 post-infection. Z-score values for genes in each cluster were calculated using the pheatmap package v1.0.12 (Kolde, 2019), averaged for each timepoint and infection, then plotted over time. Cluster Gene Ontology (GO) Biological Process term enrichment was performed using Metascape v3.0 (Zhou et al., 2019) with a minimum overlap of 3, p value cutoff of 0.01, and a minimum enrichment of 1.5.

Gene set enrichment analysis (GSEA) was performed for gene sets of interest

(Liberzon et al., 2015; Subramanian et al., 2005). Hallmark gene sets were

33 obtained from MSigDB (Liberzon et al., 2015). GC B cell signature and spleen plasma cell signatures in were created using data deposited in the Gene

Expression Omnibus (GEO, GSE60927, (Shi et al., 2015)). Briefly, gene signatures were generated by determining the top 200 differentially expressed genes between follicular B cells and subsets of interest as described above.

Memory B cell signatures (Bhattacharya et al., 2007) and LZ and DZ GC B cell signatures (Victora et al., 2010) were previously published. Plots were made using ggplot2 (Wickham et al., 2019), the viridis package (Garnier, 2018), and the tidyverse package (Wickham, 2017).

2.3 Results

2.3.1 Chronic infection promotes terminal B cell differentiation

Humoral immunity in chronic viral infections is dysregulated and often associated with poor immunity. These defects in B cell responses are characterized by hypergammaglobulinemia, altered B cell differentiation, and the delayed development of neutralizing antibodies (Battegay et al., 1993; Bjøro et al., 1994; Cashman et al., 2014; Eschli et al., 2007; Gray et al., 2007; Hunziker et al., 2003; Lane et al., 1983; Logvinoff et al., 2004; Milito et al., 2004; Moir and

Fauci, 2017; Oldstone and Dixon, 1969; Racanelli et al., 2006). Despite these observations, it has remained challenging to quantify the virus-specific B cell response to chronic viral infections and track numerical changes and

34 differentiation over time. Previous studies in the mouse model of chronic LCMV infection identified a major detrimental effect of excessive IFN-I signaling causing profound B cell attrition at early time points (days 0-3) that led to defects in subsequent humoral immunity (Fallet et al., 2016; Moseman et al., 2016;

Sammicheli et al., 2016). Despite these early B cell defects, germinal centers

(GC) still form during chronic infections (Barnett et al., 2016; Clingan and

Matloubian, 2013; Daugan et al., 2016; Fukazawa et al., 2015) and the biology of

GC responses during these persisting infections remains poorly understood. To begin to dissect these questions, we interrogated the LCMV-specific B cell response to acutely resolved (LCMV Armstrong; Arm) versus chronic (LCMV clone 13) LCMV infection. We identified LCMV-NP-specific B cells using fluorescently labeled bacterially-produced LCMV nucleoprotein (NP) as previously described (Figures 2.1A and 2.2A) (Fallet et al., 2016; Schweier et al., 2019). Although LCMV-NP-specific antibodies are not neutralizing, NP is the dominant target of the LCMV-specific antibody responses (Battegay et al., 1993).

NP protein sequences are identical between the acute and chronic strains of

LCMV allowing control for strain-specific differences in other LCMV viral proteins

(Matloubian et al., 1990, 1993; Sullivan et al., 2011). Acute or chronic LCMV infection resulted in similar frequencies of LCMV-specific B cells during the first

7-10 days of infection (Figure 2.1B). However, by day 15 post-infection (pi)

LCMV-specific B cells increased in frequency and total number following LCMV

Arm infection but remained lower in magnitude during chronic LCMV infection

(Figure 2.3A). These data indicated that chronic infection resulted in decreased 35 frequency of virus-specific B cells during the timepoints when the GC and memory B (MB) cell responses are developing.

Tight regulation of B cell differentiation into GC B cells, plasma cells (PC) and memory B cells (MB), is critical for achieving optimal antibody avidity, antibody quantity, and immunological memory, respectively. Thus, to define the

LCMV-specific B cell response in more detail, we used markers to distinguish PC

(CD138+), GC B cells (CD38-GL7+) and MB (CD38+GL7-) at day 15 pi (Figures

2.1C and 2.2A). In acute infection, approximately 80% of the NP-specific B cells were GC phenotype at this time point (Figure 2.1C-D). In contrast, during chronic infection there were four-fold fewer NP-specific GC B cells whereas PC and MB frequencies were increased 10-fold and two-fold, respectively (Figure 2.1C-D).

The reduced frequency of GC B cells was not due to an absence of GCs during chronic infection, and the GCs that were present were similar in size to those observed following acute infection (Figure 2.3B). Thus, at the peak of the virus- specific B cell response, the distribution of NP-specific B cell responses during chronic infection was skewed towards the PC and MB fates and away from the

GC B cell fate.

Longitudinal analysis of these NP-specific B cell responses during acute or chronic infection revealed differences in the magnitude and kinetics of PC, GC

B cell and MB (Figure 2.1E-G and 2.3C-E). Whereas PC and MB responses were increased during chronic infection, especially during the second and third week of infection, GC B cell frequencies were substantially lower during this time frame. These data reveal a sustained quantitative defect in the overall virus- 36 specific B cell response during chronic infection, but also identify differences in

PC, MB and GC responses over time. Collectively, these data suggest that chronic infection promotes early induction of PC and MB responses at the expense of the GC response.

Figure 2.1

Figure 2.1: Chronic infection promotes terminal B cell differentiation. (A) Representative FACS plots of splenic LCMV-NP-specific B cell staining in naïve, LCMV-Arm (Acute), and LCMV-CL13 (Chronic) infected mice at d15 post infection. Cells were previously gated as IgD-B220+Dump-. Dump gate includes CD4, CD8, F4/80, and NK1.1. (B) Frequency of splenic LCMV-NP-specific B cells in LCMV-Arm and LCMV-CL13 infected mice at days 7, 10, 15, 21, 35, and 110 post-infection. (C) Representative FACS plots of splenic LCMV-specific plasma cells, germinal center B cells, and memory B cells in LCMV-Arm and LCMV-CL13 infected mice at d15 post infection. Cells were previously gated on LCMV-specific B cells as described in A. (D) Frequency of splenic plasma cells, germinal center B cells, and memory B cells in LCMV-Arm and LCMV-CL13 infected mice at d15 post-infection. (E-G) Kinetics of LCMV-specific splenic plasma cell (E), germinal center B cell (F), and memory B cell (G) frequencies in LCMV-Arm and LCMV-CL13 at days 7, 10, 15, 21, 35, and 110 post-infection. 38

Each data point in B, E-G and H-J shows mean ± SEM with 3-5 mice per time point from two independent experiments. Data points in D show mean ± SEM from 10 mice pooled from two independent experiments. *p<0.05, **p<0.01, ***p<0.001, Student’s t-test

Figure 2.2

Figure 2.2: Gating and sort strategies for analysis of LCMV-specific B cells (A) Gating strategy for identifying LCMV-specific B cells and LCMV-specific GC B cells, MB, and PC. Cells are first gated for live lymphocyte singlets before B cell gating. (B) Purification and FACS sorting strategy used for Figures 2.4, 2.5, (BCR sequencing, bulk RNAseq, and scRNA-seq). Strategy is as follows: 1) Bulk splenocytes are first negatively selected for pan-B cells using magnetic depletion, 2) Enriched B cells are stained with AF647-conjugated LCMV-NP and magnetic positive selection is done using anti-AF647 microbeads, 3) Enriched LCMV- specific B cells are then stained for FACS sorting and sorted for yield, 4) Yield- 40 sorted LCMV-specific B cells are finally sorted for purity and populations of interest. Representative FACS plots depict B cell and LCMV-specific B cell frequencies following each step in the purification process.

Figure 2.3

Figure 2.3: LCMV-specific B cell responses are diminished during chronic infection. Related to Figure 2.1. (A) Enumeration of LCMV-specific B cell response in LCMV-Arm and LCMV- CL13 at days 7, 10, 15, 21, 35, and 110 post-infection. (B) 40x images of GCs in LCMV-Arm and LCMV-CL13 infected mice at day 15 post-infection. GC size was quantitated for two individual spleens per infection by staining IgD (red) and GL7 (blue). At least 20 GCs per sample were enumerated. Size was manually determined by marking border between IgD+ cells in the follicular mantle zone and GL7+ cells within the GC as depicted by yellow dashed line in representative images. (C-E) Enumeration of LCMV-specific splenic plasma cell (C), germinal center B cell (D), and memory B cell (E) in LCMV-Arm and LCMV-CL13 at days 7, 10, 15, 21, 35, and 110 post-infection. Each data point in A and C-D shows mean ± SEM with 3-5 mice per time point from two independent experiments. Data points in D show mean ± SEM from 10 mice pooled from two independent experiments. *p<0.05, **p<0.01, ***p<0.001, Student’s t-test

2.3.2 Affinity maturation is impaired during chronic infection.

Germinal centers are the primary location of affinity maturation, a cyclic process by which B cells mutate their immunoglobulin genes, compete for antigen and T cell help, and undergo clonal expansion following positive selection (Cyster and Allen, 2019; Mesin et al., 2016). Previous studies have shown that accumulation of affinity-enhancing mutations is critical for eventual control of viremia during chronic LCMV infection (Chen et al., 2018; Chou et al.,

2016). However, the delayed production of neutralizing antibody during chronic

LCMV infection suggests the process of affinity maturation is dysregulated, an alteration that may also occur during other chronic infections (Battegay et al.,

1993; Doria-Rose et al., 2014; Eschli et al., 2007; Kinchen et al., 2018;

Minamitani et al., 2015; Wu et al., 2015). Based on the data above showing reduced LCMV-specific GC B cells during chronic infection but increased LCMV- specific PC, we hypothesized that chronic infection might increase the overall quantity of LCMV-specific antibodies but impair affinity maturation and therefore serum antibody avidity. We first measured serum LCMV-specific antibody titers using an LCMV-NP-specific ELISA. Consistent with our previous longitudinal flow cytometry analysis of LCMV-specific PC, LCMV-NP-specific IgG titers were increased during chronic infection at day 7 pi (prior to initiation of the GC response) and at all timepoints measured after d21 pi (Figure 2.4A). We next sought to determine whether LCMV-specific serum antibody avidity was impacted by chronic infection. Thus, we measured LCMV-specific serum

43 antibody avidity using an ELISA combined with a chaotropic sodium thiocyanate

(NaSCN) wash to remove low avidity antibody (Nawa, 1992; Pullen et al., 1986).

In this assay, antibody avidity (Avidity Index) was measured by comparing the fraction of serum antibody sensitive to removal by NaSCN treatment to total antibody levels (Figure 2.5A). Following acute infection, the avidity index increased with time post infection indicating affinity maturation and development of high avidity antibody (Figures 2.4B and C). Serum antibody avidity increased following the formation of the LCMV-specific GC response following acute infection, starting at day 10 and plateaued at day 45 pi. In contrast to acute infection, the development of avid antibodies was delayed in chronic infection.

The avidity indices of the LCMV-specific antibody response were similar between acute and chronic infection at early time points pi but began to diverge at day 35 pi with lower avidity antibodies present during chronic infection. Thus, chronic infection impairs and/or delays the development of highly avid antibodies.

Given that accumulation of mutations in immunoglobulin genes is critical for driving affinity maturation (Mesin et al., 2016; Muramatsu et al., 2002;

Odegard and Schatz, 2006), we measured heavy chain B cell receptor (BCR) repertoire usage and mutation levels by sequencing BCR genes from LCMV- specific B cells at day 45 pi. VH gene family usage was similar between LCMV- specific B cells in acute and chronic infection, in both cases favoring the use of the VH1 gene family (60% of all clones), in particular VH1-82 (Figures 2.4D and

2.5B). While VH gene usage was similar, the accumulation of mutations during acute and chronic infection was distinct. LCMV-specific B cells in acute infection 44 were approximately 2.5% mutated from their germline VH gene sequence, whereas LCMV-specific B cells in chronic infection were only 1.5% mutated from germline (Figure 2.4E and 2.5D). In chronic infection decreased BCR mutations occurred preferentially in the complementarity determining regions (CDR) 1 and

2, but not in the framework regions (Figure 2.5E). Additionally, mutations resulting in amino acid replacements were also decreased in the CDR regions during chronic infection (Figure 2.5F). We next applied the BASELINe algorithm

(Yaari et al., 2012) to quantify selection pressure in the CDR and framework regions. BCR sequences from LCMV-specific B cells in chronic infection had less positive selection in the CDRs than in acute infection (Figure 2.4E). Collectively, these data suggest that chronic infection impairs antibody affinity maturation by decreasing the accumulation of affinity-enhancing mutations during somatic hypermutation in the GC.

Figure 2.4

Figure 2.4: Affinity maturation is impaired during chronic infection. (A) Kinetic measurement of serum LCMV-NP-specific IgG titers in LCMV-Arm and LCMV-CL13 infected mice at days 7, 10, 15, 21, 35, and 110 post-infection. (B) Anti-LCMV-NP serum avidity index measurements in LCMV-Arm and LCMV CL13 infected mice at days 15 and 45 post-infection using chaotropic NaSCN ELISA. Dose response to increasing amounts of the chaotropic agent NaSCN is depicted. (C) Anti-LCMV-NP serum avidity index in response to 1.5M NaSCN in LCMV-Arm and LCMV-CL13 infected mice at days 7, 10, 15, 21, 35, and 110 post-infection. (D) Heavy chain VH family usage in LCMV-specific B cell BCR repertoires from LCMV-Arm and LCMV-CL13 infection mice at d45 post-infection. Each point represents one mouse. Data representative of two independent experiments. (E) Frequency of mutation in the heavy chain VH genes in LCMV- Arm and LCMV-CL13 infected mice at d45 post-infection. Data represent VH germline mutation frequency of one representative sequence from each clonotype. Data are pooled from IgH BCR repertoires of 5 individual mice per infection. ****p<0.0001, Mann-Whitney U-test of medians. (F) Selection pressure (S) in IgH CDR and FWR measured by BASELINE algorithm. Data are pooled 46 from IgH BCR repertoires of 5 individual mice per infection. ***p<0.001, Student’s t-test.

Figure 2.5

Figure 2.5: Chronic infection impairs affinity maturation but BCR repertoire usage is largely unaffected. Related to Figure 2.4 (A) Anti-LCMV-NP serum avidity index in response to NaSCN treatment in LCMV-Arm and LCMV-CL13 infected mice at days 7, 10, 15, 21, 35, and 110 post-infection. (B) Heavy chain VH1 gene usage in LCMV-specific B cell BCR repertoires from LCMV-Arm and LCMV-CL13 infection mice at d45 post-infection. Each point represents one mouse. Data representative of two independent experiments. (C) Frequency of heavy chain BCR repertoire represented by the 20 most abundant clonal families from LCMV-Arm and LCMV-CL13 infection mice at d45 post-infection. (D) Plot of BCR family clonal abundance by average mutation frequency. (E) Frequency of heavy chain mutations in broken down by heavy chain region. Complementarity determining region (CDR) and framework (FWR). (F) Frequency of replacement and silent mutations in BCR repertoires analyzed. Data are pooled from IgH BCR repertoires of 5 individual mice per infection. ***p<0.001, Student’s t-test.

2.3.3 Distinct molecular profiles of LCMV-specific GC, PC and MB during chronic infection.

Chronic infection promotes unique transcriptional programs and differentiation states in CD8+ and CD4+ T cells (Crawford et al., 2014; Doering et al., 2012), but it is unclear whether chronic infection impacts transcriptional programs of antigen-specific B cells and if so, how . To interrogate this question, we sorted LCMV-specific GC B cells, PC, and MB at days 7, 15, and 45 pi during acute and chronic infection (Figure 2.2B) and performed RNA-seq. In general, each LCMV-specific B cell population (GC B cells, PC, and MB) clustered by day pi in principle component space (Figure 2.6A). For all three cell types, transcriptional profiles were most similar at day 7 pi. For PC and MB, transcriptional programs diverged over time with the greatest difference in RNA- seq apparent at d15 or 45 p.i (Figure 2.7A-C). However, the largest transcriptional divergence between acute and chronic infection occurred for GC B cells at day 15 pi (Figure 2.6B). Indeed, differential gene expression analysis identified 346 genes differentially expressed between LCMV-specific GC B cells at day 15. 311 of the 346 differentially expressed genes at day 15 pi were upregulated during chronic infection. In LCMV-specific GC B cells at day 15 pi, acute infection enriched for genes related to B cell homeostasis (Bcl6, Lyn, and

Tnfrsf13c), cellular response to DNA damage stimulus (Aicda, Foxo1, H2afx, and

Gadd45b), and histone H3 acetylation (Brd1 and Hdac4) whereas chronic infection enriched for genes related to response to virus (C1qbp and Tlr7), 49 response to interferon-b and interferon-a (Irf1, Stat1, Bst2, and Ifi204), and response to biotic stimulus (Prdm1, Ptpn22, and Mif) (Figure 2.6C, 2.6D and

2.7D-F). We observed approximately 12-fold increased expression of the PC master regulatory transcription factor Prdm1 and 1.6-fold repression of Bcl6, the key transcriptional regulator of GC B cell fate, during chronic infection suggesting that chronic infection may promote excessive LCMV-specific GC B cell differentiation towards the PC fate (Figure 2.6E). Together these data suggested that LCMV-specific GC B cells have divergent transcriptional programs at day 15, and expression of key transcription factors pointed to potentially altered GC B cell fate decisions.

Given that the greatest difference in transcriptional profiles occurred in GC

B cells at day 15 pi, we next investigated the impact of chronic infection on the program of these cells in more detail. The GC is comprised of two anatomical zones where GC B cells undergo selection (light zone; LZ) and accumulate mutations through proliferation (dark zone; DZ). Given the data above on altered accumulation of affinity-enhancing mutations in the BCR during chronic infection, we hypothesized that chronic infection may disrupt the LZ to DZ balance for virus-specific B cells. Using Gene Set Enrichment Analysis (GSEA)

(Subramanian et al., 2005), we compared LCMV-specific GC B cells from acute and chronic infection to known antigen-specific GC B cell LZ and DZ gene signatures (Victora et al., 2010). In chronic infection, LCMV-specific GC B cells were enriched for the LZ gene signature whereas LCMV-specific GC B cells in

50 acute infection were enriched for the DZ (Figure 2.6F and 2.6G) suggesting an altered GC distribution of virus-specific B cells in chronic infection. Selection into the DZ is driven by Myc and the mTORC1 complex following positive selection in the LZ (Calado et al., 2012; Dominguez-Sola et al., 2012; Ersching et al., 2017;

Finkin et al., 2019; Luo et al., 2018). The mTORC1 signaling gene signature was enriched in chronic infection however, the Myc target gene signature was unchanged between acute and chronic infection (Supplemental Table 1). Given the critical role for mTORC1 signaling in plasma cell differentiation (Benhamron et al., 2015; Jones et al., 2016) and increased expression of Prdm1 and decreased expression of Bcl6 noted in Fig 2.6E during chronic infection, we next compared LCMV-specific GC B cells from acute and chronic infection to known antigen-specific GC B cell, PC and MB gene signatures using GSEA

(Bhattacharya et al., 2007; Shi et al., 2015). In agreement with the flow cytometric data in Fig 2.1D and transcription factor data in Fig 2.6E, LCMV- specific GC B cells from acute infection enriched for the GC B cell signature but not the PC or MB signatures (Figure 2.6H-J). In contrast, LCMV-specific GC B cells in chronic infection enriched for the PC and MB cell signatures, but not the

GC B cell signature. Collectively, these data indicate that chronic infection promotes a PC and MB cell transcriptional signature within GC B cells perhaps indicating early terminal differentiation within the GC in chronic infection.

Figure 2.6

Figure 2.6: Distinct molecular profiles of LCMV-specific GC, PC and MB during chronic infection. (A) PCA plot of LCMV-specific GC B cells, PC, and MB at day 7, 15, and 45 post-infection. (B) Number of significant differentially expressed genes (p<0.05 and >2 log fold-change) between LCMV-specific GC B cells, PC, and MB from LCMV-Arm and LCMV-CL13 infected mice at the indicated timepoints post- infection. (C) Heatmap depicting top 35 and bottom 50 differentially expressed (p<0.05 and >2 log fold-change) genes between LCMV-specific GC B cells from LCMV-Arm and LCMV-CL13 infected mice at day 15 post-infection. (D) Top 10 GO terms enriched in LCMV-specific GC B cells from LCMV-Arm and LCMV- CL13 infected mice at day 15 post-infection. Color of points represents p-value significance. Size of point represents percent of cluster genes enriched in each respective GO term. (E) mRNA normalized counts of Bcl6 and Prdm1 in LCMV- specific GC B cells from LCMV-Arm and LCMV-CL13 at day 15 pi. (F) GSEA plot of light zone gene signature in LCMV-specific GC B cells from LCMV-Arm and LCMV-CL13 infected mice at day 15 post-infection. (G) GSEA plot of dark zone gene signature in LCMV-specific GC B cells from LCMV-Arm and LCMV-CL13 infected mice at day 15 post-infection. (H) GSEA plot of GC B cell gene signature 52 in LCMV-specific GC B cells from LCMV-Arm and LCMV-CL13 infected mice at day 15 post-infection. (I) GSEA plot of PC gene signature in LCMV-specific GC B cells from LCMV-Arm and LCMV-CL13 infected mice at day 15 post-infection. (J) GSEA plot of MB gene signature in LCMV-specific GC B cells from LCMV- Arm and LCMV-CL13 infected mice at day 15 post-infection.

Figure 2.7

Figure 2.7: Transcriptional regulation of LCMV-specific GC B cells, PC, and MB is temporally dynamic (A) Heatmap depicting all differentially expressed (p<0.05 and >2 log fold- change) genes between LCMV-specific GC B cells from LCMV-Arm and LCMV- 54

CL13 infected mice at days 7, 15, or 45 post-infection. Selected genes are shown. Heatmaps were k-means clustered and average z-scores for each cluster were plotted over time for each infection. (B) Heatmap depicting all differentially expressed (p<0.05 and >2 log fold-change) genes between LCMV-specific plasma cells from LCMV-Arm and LCMV-CL13 infected mice at days 7, 15, or 45 post-infection. Selected genes are shown. Heatmaps were k-means clustered and average z-scores for each cluster were plotted over time for each infection. (C) Heatmap depicting all differentially expressed (p<0.05 and >2 log fold- change) genes between LCMV-specific memory B cells from LCMV-Arm and LCMV-CL13 infected mice at days 7, 15, or 45 post-infection. Selected genes are shown. Heatmaps were k-means clustered and average z-scores for each cluster were plotted over time for each infection. (D) Top 5 GO terms enriched in each GC B cell cluster identified in C. Color of points represents p-value significance. Size of point represents percent of cluster genes enriched in each respective GO term. (E) Top 5 GO terms enriched in each plasma cell cluster identified in C. Color of points represents p-value significance. Size of point represents percent of cluster genes enriched in each respective GO term. (F) Top 5 GO terms enriched in each memory B cell cluster identified in C. Color of points represents p-value significance. Size of point represents percent of cluster genes enriched in each respective GO term.

2.4 Discussion

In the studies outlined in this chapter, we used novel recombinant protein probes to interrogate the virus-specific B cell response in the LCMV mouse model of chronic viral infection (Sammicheli et al., 2016; Schweier et al., 2019;

Sommerstein et al., 2015). We found that chronic viral infection skews GC B cell differentiation towards the PC and MB fates and away from the GC B cell fate.

Skewing away from continued participation in the GC reaction was associated with decreased serum antibody avidity and impaired affinity maturation. Previous studies have identified altered B cell differentiation during chronic viral infection prior to the establishment of the GC (Fallet et al., 2016; Moseman et al., 2016;

Sammicheli et al., 2016). Our results identify a novel role for chronic infection in altering B cell differentiation after the establishment of the GC response and may explain why impaired affinity maturation is often observed during chronic viral infection.

Transcriptional profiling of virus-specific B cell responses during acute and chronic viral infection identified upregulation of the PC and MB gene programs in

GC B cells during chronic infection. Upregulation of these gene signatures was correlated with strong evidence of inflammatory signaling and suggested that inflammation may play a role in skewing B cell differentiation away from the GC B cell fate and towards the PC and MB fates. Inflammatory signaling is known to impact B cell differentiation prior to the GC (Fallet et al., 2016; Moseman et al.,

2016; Sammicheli et al., 2016) but it is thought that GC B cell responses are less

56 sensitive to inflammatory signaling than naïve B cells due to their decreased expression of receptors for many mediators of inflammation (immgen.org, (Heng et al., 2008)). While our studies point to type I and type II interferon signaling as key regulators of skewed GC B cell differentiation, we cannot point to one inflammatory cytokine that was causal for the observed phenotype. Persistent inflammation during chronic infection is complex and multivariable (Stelekati and

Wherry, 2012; Zuniga et al., 2015). Further interrogation of the inflammatory profile of chronic viral infection will be necessary to fully tease this apart.

Chronic viral infections present a unique challenge to the humoral immune response due to the presence of both persistent inflammation and chronic antigen (Virgin et al., 2009). While we observe an impact of inflammation on shaping virus-specific GC responses, we cannot discount the impact of chronic antigen exposure. B cell differentiation decisions are regulated in part by BCR affinity for antigen which involves both a signaling component and an antigen capture component (Gitlin et al., 2014; Ise and Kurosaki, 2019; Luo et al., 2018;

Shlomchik et al., 2019). It is likely that low-affinity but frequent signaling through the BCR may appear to a GC B cells as a “high-affinity signal”. High affinity B cells are preferentially recruited into the PC fate (Phan et al., 2006) and we observed increased PC differentiation during chronic viral infection consistent with this model. Whether inflammatory signaling or signals from antigen play a more important role in shaping B cell differentiation warrant future study.

Taken together, the studies from this chapter suggest that promotion of B cell terminal differentiation rather than the GC fate may explain why humoral 57 immune responses to chronic infection display hypergammaglobulinemia and delayed development of neutralizing antibodies.

CHAPTER 3: Transcriptional regulation of germinal center

responses during acute and chronic viral infection

3.1 Introduction

Balancing the decision to remain within the GC and continue participating in affinity maturation or to exit the GC as a plasma cell (PC) or memory B cell

(MB) is critical for achieving optimal antibody avidity, antibody quantity, and establishing immunological memory in response to immunization or infection.

Chapter 2 demonstrated that chronic infection alters this balance in favor of the

MB and PC fates. However, the exact signals that regulate this GC B cell differentiation checkpoint are currently unclear. Given the link between BCR signal strength and GC B cell differentiation (Phan et al., 2006), it is likely that chronic antigen during chronic viral infection promotes GC B cell differentiation into PC and GC exit. Previous studies have identified a critical role for BCR affinity in regulating this differentiation decision with GC B cells bearing high affinity receptors undergoing PC differentiation whereas those with low affinity receptors favor the MB fate. GC B cells with “just-right” BCR affinity undergo positive selection and DZ reentry (Ise et al., 2018; Kräutler et al., 2017; Laidlaw et al., 2017; Suan et al., 2017a).

Given the link between BCR signal strength and GC B cell differentiation discussed in Chapter 1.2.3, it is possible that chronic antigen alone promotes GC

B cell differentiation into PC and GC exit. However, our RNA Sequencing work in

GC B cells during acute and chronic infection demonstrated a marked inflammatory signal in day 15 GC B cells responding to chronic virus (Figure

2.6). The relative role of chronic antigen vs chronic inflammation on the GC B cell response remains unknown.

The studies in this chapter interrogate the impact of chronic infection on

GC B cell fate decisions. Using scRNAseq of virus-specific GC B cells, we identify a key GC B cell checkpoint that is dysregulated during chronic infection.

Specifically, during chronic viral infection, GC B cells inefficiently re-enter the dark zone during the cyclic GC reaction and instead GC B cell fate is skewed towards premature exit as MB or terminal PC. This short-circuiting of the cyclic process of GC B cell development occurred at a single light zone step in GC B cell differentiation. Single cell profiling pointed to attenuated BCR signaling and

CD40 signaling, but amplified inflammatory signals were also associated with this altered GC developmental pattern. Manipulation of CD40 signals, antigen, and inflammation identified a dominant role for persistent inflammation over antigen in skewing GC B cell differentiation during chronic infection. Thus, persistent inflammation during chronic infection promotes GC B cell differentiation and GC exit leading to impaired humoral immunity.

3.2 Methods

3.2.1 Mice

6-8 week old C57BL/6 Ly5.2CR (CD45.1) or C57BL/6 (CD45.2) mice were purchased from NCI. Both male and female mice were used for these studies.

Ifnarflox/flox (Prigge et al., 2015) and CD19-cre+/+ (Rickert et al., 1997) were purchased from Jackson Laboratory (strain numbers 028256 and 006785, respectively). Ifnarflox/floxCD19-cre+/wt mice were generated by breeding Ifnarflox/flox and CD19-cre+/+ mice and were maintained in our colony. Mice were maintained in a specific-pathogen-free facility at the University of Pennsylvania. All mice were used in accordance with Institutional Animal Care and Use Committee guidelines for the University of Pennsylvania.

3.2.2 Infections and antibody treatments.

Mice were infected intraperitoneally (ip) with 2*105 plaque-forming units

(PFU) of LCMV-Armstrong (Arm) or intravenously (iv) with 4*106 PFU LCMV-

CL13 (CL13). LCMV strains were propagated and titers were determined as previously described (Odorizzi et al., 2015; Wherry et al., 2003b). For IFNAR blockade experiments, mice were treated with anti-mouse IFNAR-1 (MAR1-5A3,

BioXcell) as previously described starting at d13 (Wilson et al., 2013a). For CD40 blockade experiments, mice were treated with 200ug ip anti-mouse CD40L

(MR1, Bioxcell) or 25ug ip anti-mouse CD40 (FGK45, Bioxcell). For BCR

61 inhibition experiments, ibrutinib (MedChem Express) was dissolved in Captex355

(Abitec) at 1.25 mg/ml and mice were treated with 250ug of ibrutnib per dose.

3.2.3 LCMV-NP protein production and purification

Basel, Switzerland) and originally developed by H. Pircher (University of

Freiberg, Germany) and was used as previously described (Fallet et al., 2016;

Schweier et al., 2019; Sommerstein et al., 2015). Briefly, plasmid was transformed into Rosetta™ 2(DE3) competent cells (Novagen, Millipore Sigma).

AKTAprime plus FPLC system (GE Healthcare). Purified protein was desalted using PD-10 columns (GE Healthcare) and concentrated prior to use.

3.2.4 Single cell suspension preparation and LCMV-specific B cell staining and enrichment.

Single cell suspensions were obtained from spleens by homogenizing tissues against a 70um cell strainer. Cells were washed through cell strainer and red blood cells were lysed in ACK lysis buffer (Thermo Fisher Scientific). Cells were resuspended in MACS buffer (PBS with 1mM EDTA and 2% fetal bovine 62 serum) prior to cell staining or B cell enrichment. To identify LCMV-specific B cells, purified LCMV-NP was fluorescently labeled with Alexa Fluor 647 or Alexa

Fluor 488 using antibody labeling kits according to manufacturer’s instructions

(Thermo Fisher Scientific). Single cell suspensions were stained with LCMV-NP for 15 minutes at 37°C in the presence of Fc-block (anti-CD16/32, 24G2, BD) and washed twice with MACS buffer prior to staining with additional fluorescent antibody panels. For bulk and single cell RNAseq experiments, B cell enrichment

(negative selection) was performed using mouse pan-B cell isolation kits (Stem

Cell Technologies) according to manufacturer’s instruction except incubation times were doubled. Following B cell enrichment, LCMV-NP staining was performed as described above. LCMV-NP-labeled cell suspensions were further enriched (positive selection) for LCMV-specific B cells by incubating cells with anti-Cy5/AF647 microbeads (Miltenyi Biotec) followed by purification over LS columns (Miltenyi Biotec).

3.2.5 Flow cytometry and cell sorting

Single cell suspensions were prepared and stained with fluorescently labeled LCMV-NP as described above. Cell suspensions were stained at 4°C for

30 minutes for surface protein staining. Intracellular staining was conducted at room temperature for 1 hour. Antibodies were purchased from Biolegend:

CXCR4 (L276F12), CD83 (Michel-19), CD86 (GL-1), B220 (RA3-6B2), CD19

(1D3), IgD (11-26c.2a), CD138 (281-2), GL7 (GL7), CD38 (90), F4/80 (BM8),

NK1.1 (PK136), IgM (RMM-1), Ki67 (16A8), CD69 (H1.2F3), Steptavidin-BV786; 63

BD Biosciences: CXCR4 (2B11), CD83 (Michel-19), CD19, (1D3), CD138 (281-

2), CD38 (90), CD4 (GK1.5), Blimp1 (5E7), Bcl6 (K112-91), CD31 (MEC 13.3),

CCR6 (140706), Streptavidin BB780; eBioscience (Thermo Fisher Scientific):

GL7 (GL7), CD4 (RM4-5), CD8 (53-6.7), F4/80 (BM8), NK1.1 (PK136); or from

Southern Biotech: Ly6a (D7). Live cells were discriminated by staining with

Live/Dead Aqua (Thermo Fisher Scientific) or Zombie Yellow (Biolegend). Cell proliferation was measured by injecting mice ip with 3mg of 5-bromo-2’- deoxyuridine (BrdU) 2 hours prior to analysis. Fixation of cells for measurement of BrdU incorporation was done using a FIX & PERM Cell Permeabilization Kit with adjusted timings and BrdU was detected using an anti-BrdU antibody

(MoBU-1, Thermo Fisher Scientifc). For cell sorting, samples were sorted directly into lysis buffer (RLT or RLT+, Qiagen) or into 10% FBS cRPMI supplemented with 25mM HEPES. Flow cytometry data were acquired on a BD LSR II instrument or BD Symphony A3 instrument. Cell sorting was performed on a BD

FACSAria enclosed within a laminar flow hood. Data were analyzed using

FlowJo software (FlowJo, LLC).

3.2.8 Single cell RNAseq sample preparation

LCMV-specific GC B cells (CD38-GL7+) were sorted from mice infected with LCMV-Arm or LCMV-CL13 at day 15 post-infection. Splenocytes from 5 mice per infection were pooled, enriched for LCMV-specific B cell and stained with surface antibodies as previously described. Samples were gated as LCMV-

LCMV-SPECIFICIgD-B220+Dump- prior to GC B cell gates. Dump stain included 64

CD4, CD8, NK1.1, and F4/80. Cells were sorted for yield then for purity. Samples were counted using a Countess II Automated Cell Counter (Thermo Fisher

Scientific) prior to library preparation. For library preparation, 5000 cells from each sample were loaded into a Chromium Controller (10x Genomics), emulsions were created, and libraries were constructed according to the manufacturer’s recommend protocols for the Chromium Single Cell 3’ v2 kit (10x

Genomics). Samples were pooled and quantified using a KAPA Library

Quantification Kit (Kapa Biosystems) prior to sequencing on an Illumina NextSeq

550 instrument (150 cycle, high output flow cell).

3.2.9 Single cell RNAseq data processing and analysis

BCL files from sequences samples were demultiplexed, aligned to the mouse mm10 genome, filtered, and UMI counted using Cell Ranger Software v2.1 (10x Genomics). 3,270 cells were recovered from LCMV-Arm library with

88,576 mean reads per cell and 2,803 median genes detected per cell. 3,374 cells were recovered from LCMV-CL13 library with 83,259 mean reads per cell and 2,328 median genes detected per cell. Downstream analysis was performed using Seurat v2.3.1 (Butler et al., 2018). Complete script reproducing all analyses from raw data will be available in a github repository upon publication of this data.

Briefly, individual library outputs from the Cell Ranger pipeline were loaded, filtered to remove cells with > 5% mitochondrial reads and > 6000 unique genes detected. Filtered data from individual libraries were imputed using SAVER v0.3.0 (Huang et al., 2018) with default settings. Imputed count matrices were 65 used to subsequent analysis in Seurat. Variable genes for each library were determined by calculating gene-gene correlations using Spearman distance, scoring gene-gene correlations, and taking the top 1500 highest correlated genes as described previously (https://hemberg-lab.github.io/scRNA.seq.course, section 8.3.2 Correlated Expression). Regression was performed to remove variation due to the number of genes detected or the percent of mitochondrial reads. Common correlation analysis (CCA) was performed to identify common sources of variation LCMV-Arm and LCMV-CL13 prior to dataset merging using the first 12 CCA dimensions. Data was then clustered using the first 12 CCA dimensions and a resolution parameter of 0.6. Differential gene expression was conducted using the Wilcoxon rank sum test with a threshold of 0.25.

Single-cell enrichment scores for Figure 3.1 were calculated using the

AddModuleScore function in the Seurat package. Previously published gene sets were used to calculate LZ and DZ (Victora et al., 2010), PC, MB vs. GC, and MB vs PC single-cell enrichment scores (Bhattacharya et al., 2007; Shi et al., 2015).

Cluster averaged single cell enrichment scores for each gene signature of interest were used to define GC zone and differentiation state. Cell cycle scores were calculated using the CellCycleScoring function in the Seurat package using previously defined gene signatures for each cell cycle stage (Tirosh et al., 2016).

GO term enrichment was conducted using Metascape v3.0 (Zhou et al., 2019) with a minimum overlap of 3, p value cutoff of 0.01, and a minimum enrichment of 1.5..

Pseudotime analysis of LZ GC B cells was conducted using Monocle 2 v2.10.1 (Qiu et al., 2017b, 2017a). CCA dimensions were imported from Seurat and used for pseudotime calculations. Pseudotime trajectory start was inferred using cell cycle staging and cluster function. Trajectories were rooted to start in cluster 4. Single cell enrichment scores for gene sets of interest were calculated using GSVA (Hänzelmann et al., 2013). Preprocessed, filtered, normalized, and log-transformed scRNAseq expression data was used as input to GSVA treating each cell as a unique sample. A Gaussian distribution was used for the GSVA calculation as input data was previously log-transformed. Single-cell GSVA scores were plotted as a function of pseudotime to obtain the graphs in Figure

3.2.

3.3 Results

3.3.1 scRNAseq reveals germinal center organization.

The bulk RNA sequencing data presented in chapter 2 suggested divergent fate decisions in LCMV-specific GC B cells responding to acute and chronic infection including an enrichment in signatures of PC and MB cells in the

GC population in chronic infection. To examine the GC reaction with more precision and to begin to dissect the specific steps in the GC impacted by chronic infection, we sorted LCMV-specific GC B cells at day 15 pi from acute and chronic infection and performed droplet-based single cell RNA sequencing

(scRNAseq). In total, we analyzed 3270 and 3374 LCMV-specific GC B cells from GCs of acute and chronic infection, respectively, and clustered initially without consideration of infection type. Unsupervised clustering identified seven populations of LCMV-specific GC B cells that clustered according to unique transcriptional and functional profiles (Figure 3.1A and 3.2A-B). We next sought to place the clusters identified within the known organization of the GC, in particular the LZ and DZ. To identify the LZ and DZ, single-cell enrichment scores were calculated for each cell using previously published genesets (Victora et al., 2010) for the LZ and DZ GC B cell program (Figure 3.1B and 3.2C). Using this approach, we identified enrichment of the LZ gene program in clusters 2, 4, 5 and 6 and enrichment of the DZ gene program in clusters 0, 1, and 3. We next sought to determine if we could identify the output of the GC reaction; PC and

MB. Calculation of single-cell enrichment scores for previously published PC and

MB genesets (Bhattacharya et al., 2007; Shi et al., 2015) identified enrichment of the MB gene program in cluster 5 and of the PC gene program in cluster 6

(Figure 3.1C and 3.2D). Enrichment of the PC and MB transcriptional programs occurred within clusters enriched for the LZ program supporting previously published observations (Ise et al., 2018; Kräutler et al., 2017; Laidlaw et al.,

2017; Suan et al., 2017a) that the decision to undergo terminal differentiation and exit the GC occurs within the LZ.

Given the importance of cell proliferation in the germinal center reaction, we next sought to understand the role of the cell cycle in driving clustering of

LCMV-specific GC B cells (Gitlin et al., 2014, 2015; Mcheyzer-Williams et al., 68

2015). Consistent with cell cycle biology as a major driver of the GC reaction we observed restriction of active proliferation and enrichment for the S phase and

G2/M phase of the cell cycle to clusters belonging to the DZ (clusters 0, 1, and 3)

(Figure 3.1D). In fact, the majority of the heterogeneity observed within the DZ clusters was driven by differences in cell cycle stage with cluster 0 enriched for cells in the G2/M phase, cluster 1 enriched for cells in the S phase, and cluster 3 containing a mixture of cells from all three phases of the cell cycle (3.2E). In contrast, the majority of cells in clusters 4, 5, and 6 were in G1, indicating that these cells had exited from active division. These data were consistent with previous studies (Dominguez-Sola et al., 2012; Finkin et al., 2019; Gitlin et al.,

2015) identifying regulation of cell cycle entry as critical for the GC reaction.

Consistent with the calculated single-cell enrichment scores in Figure 3.1B-D, we observed cluster- and zone-restricted expression of individual genes central to the LZ (Cd83, Cd86) or the DZ programs (Mki67, Cxcr4) (Figure 3.1E)

(Victora et al., 2010, 2012). Notably, despite being in cell cycle, clusters 2 and 3 enriched for the LZ and DZ signatures, respectively. These data combined with the cell cycle analysis in Figure 3.1D suggested that cells in cluster 2 were leaving the LZ, initiating cell cycle and entering the DZ whereas cells in cluster 3 were leaving cell cycle and exiting the DZ back to the LZ. Cells in cluster 4 enriched for genes related to antigen processing and presentation via MHCII

(Cd74 and H2-DMa) and B cell survival and signaling genes (Cd79b and

Tnfrsf13c) suggesting that cells in this cluster are interacting with Tfh and/or

FDCs. In addition to enriching for antigen processing and presentation via MHCII, 69 cluster 5 enriched for genes involved in T cell trafficking (Figure 3.2B). Notably, cluster 6 was defined by the high expression of the PC immunoglobulin secretory factor Jchain as well as genes related to protein folding, protein localization to the

ER, and response to ER stress. Other key GC B cell genes such as Gpr183,

Foxo1, Il21r, S1pr1, and S1pr2 were expressed in the expected LZ and/or DZ clusters (Figure 3.1E). Thus, single cell analysis identified LCMV-specific GC B cells within and transitioning between the LZ and DZ phases of the GC reaction and also identified clusters with transcriptional signatures consistent with antigen presentation where GC B cells may interact with APC and also with other known

GC biology such as commitment to PC differentiation (Figure 3.1F).

Figure 3.1

Figure 3.1: scRNAseq reveals germinal center organization. (A) tSNE plot of unbiased clustering of LCMV-specific GC B cells at day 15 pi. (B) tSNE plot depicting single-cell light zone or dark zone geneset enrichment. (C) tSNE plot depicting single-cell MB or PC geneset enrichment. (D) tSNE plot depicting cell cycle stage enrichment scores for each cell. (E) Plots showing expression of selected genes important for GC B cell biology. (F) tSNE plot depicting cluster identification and cluster annotation of LCMV-specific GC B cells determined using unbiased clustering.

Figure 3.2

Figure 3.2: Unsupervised clustering of LCMV-specific GC B cells identifies clusters with unique biology and known GC B cell function. (A) Heatmap of LCMV-specific GC B cell cluster-defining genes. Heatmap depicts top 10 most differentially expressed genes per cluster. (B) Top enriched GO terms for each identified cluster in LCMV-specific GC B cells scRNAseq data. (C) Averaged single-cell enrichment scores for the LZ and DZ gene signatures in each cluster. (D) Averaged single-cell enrichment scores for genes upregulated in PC, genes upregulated in MB versus GC B cells, and gene upregulated in MB vs PC for each cluster. (E) Frequency of cells in each cluster in the indicated cell cycle stage.

3.3.2 Chronic infection promotes terminal B cell differentiation and

GC exit.

The single cell analysis of LCMV-specific GC B cells above identified cells in the LZ and DZ phases, as well as cells in transition between these two phases, and GC B cells undergoing differentiation towards PC or MB. These data provide increased resolution into the single cell transcriptional coordination of GC B cell responses. To understand how chronic infection alters GC B cell fate decisions, we next investigated how LCMV-specific GC B cells in acute and chronic LCMV infection differ with respect to the clusters of GC B cell biology defined in the scRNA-seq data. To this end, we first determined the frequency of LCMV-specific

GC B cells from acute and chronic infection within each cluster. Both acute and chronic infection had similar frequencies of LCMV-specific GC B cells in the DZ clusters (Figure 3.3A). However, chronic infection resulted in increased frequency of LCMV-specific GC B cells in the pre-PC (cluster 6) and pre-MB

(cluster 5) clusters by 6.6-fold and 1.8-fold, respectively. Conversely, chronic infection diminished the frequency of cells in transition from LZ to DZ (cluster 2) by 1.4-fold. Thus, chronic infection promoted accumulation of virus-specific GC B cells in the pre-PC and pre-MB fates and depletion of cells re-entering the cell cycle and the DZ.

GC B cell differentiation reflects a stepwise, circular, differentiation scheme with distinct biology including rapid proliferation and somatic hypermutation in the DZ, exit from cell cycle and testing of BCR, and then key 73 decision points in the LZ to initiate either DZ reentry or begin a PC or MB differentiation path. To begin to understand how chronic infection might alter these biological events and decision points in the GC, we next applied pseudotime analysis to infer differentiation trajectories and cluster interconnectedness (Qiu et al., 2017b, 2017a). Pseudotime analysis of LZ clusters 2, 4, 5, and 6, revealed two major branch points (Figures 3.3B and

3.4A). The first branch point was between cluster 4 and 2 where GC B cells are interacting with FDC and Tfh (cluster 4), or GC B cells are returning to the DZ

(cluster 2), and clusters 5 and 6 where cells are undergoing differentiation into

MB and PC, respectively. LCMV-specific GC B cells in chronic infection favored the branch toward the pre-MB and pre-PC fates whereas cells in acute infection favored the DZ re-entry branch (Figure 3.3B). This pseudotime analysis is consistent with the strong shift in cluster frequency seen in Figure 3.3A with

LCMV-specific GC B cells from chronic infection highly biased to the pre-PC and pre-MB cell fates. The second branch point occurred in the DZ re-entry cluster also separated cells by infection. These data suggest that chronic infection not only drives terminal differentiation of GC B cells within the LZ when GC B cells are interacting with Tfh and FDC, but that chronic infection also alters LZ GC B cells re-entry to the DZ.

To test these predictions, we treated mice infected with acute LCMV or chronic LCMV with 5-bromo-2’-deoxyuridine (BrdU) 2 hours prior to analysis

(Figure 3.3C). We reasoned that because BrdU is only incorporated into actively dividing cells, the short BrdU labeling prior to analysis would provide a defined 74 window into the cellular output of the GC reaction at day 15 pi. BrdU incorporation was 2-fold lower in LCMV-NP-specific B cells in chronic infection

(Figure 3.3C) indicating less proliferation overall. Of the LCMV-NP-specific B cells that proliferated during the 2hr labeling window (BrdU+), GC B cell frequencies were decreased whereas MB and PC frequencies were increased in chronic infection compared to acute infection (Figure 3.3D). Although some of this MB and PC BrdU incorporation could be occurring outside the GC in this analysis, these data when combined with our scRNAseq pseudotime analysis further support the notion that chronic infection skews LCMV-NP-specific GC B cells towards terminal differentiation and GC exit rather than DZ reentry.

GC B cell differentiation and positive selection involves integration of multiple extracellular signals (Mesin et al., 2016; Shlomchik et al., 2019). Of the pathways that can influence GC B cell fate decisions, e.g. antigen, inflammatory cues, and signals from T follicular helper cells (Tfh), several were implicated in the scRNA-seq and bulk RNA-seq above. Therefore, we used pseudotime projection of transcriptional signals across the phases of GC progression in acute versus chronic infection to examine CD40 (i.e. from Tfh), inflammatory signaling,

BCR and activation of c-Myc and MTOR which occurs after a BCR positive selection signal is received (Dominguez-Sola et al., 2012; Ersching et al., 2017;

Finkin et al., 2019; Luo et al., 2018).

We first analyzed signals derived from Tfh cell help given that chronic

LCMV infection leads to an abundance of Tfh (Crawford et al., 2014; Fahey et al.,

2011; Greczmiel et al., 2017; Vella et al., 2017). Tfh cell help to GC B cells 75 comes not only through pMHCII:TCR interaction (Crotty, 2015; Schwickert et al.,

2011) but also through critical costimulatory interactions via CD40-CD40L (Ise et al., 2018; Kawabe et al., 1994; Watanabe et al., 2017) that drive GC B cell survival and entry into the DZ. Both acute and chronic infection upregulated expression of genes related to the response to CD40 signaling within LZ GC B cells of cluster 4 and along the branch leading to PC and MB differentiation

(Figure 3.3E). This result is consistent with previous observations that Tfh cells are abundant during both acute and chronic infection (Crawford et al., 2014;

Fahey et al., 2011; Greczmiel et al., 2017; Vella et al., 2017) and that CD40 agonism can drive terminal PC differentiation in vitro (Erickson et al., 2002; Foy* et al., 1996). However, although the enrichment was similar between infections, the pattern of this CD40 “help signal” was distinct. Enrichment of CD40 signaling genes in acute infection increased early in the pseudotime trajectory. In chronic infection, this CD40 signature was instead predominantly found in the latter parts of the pseudotime trajectory where the pre-MB and pre-PC features where present. Thus, despite the abundance of Tfh cells in chronic infection, these data suggest that Tfh help might not be available during the same steps in GC B cell differentiation. Whether these differences are due to B cell intrinsic features, alterations in Tfh homing and migration, GC architectural differences, or help signal delivery will require future examination.

Following interaction with Tfh cells and positive selection in the GC, B cells upregulate Myc targets, activate MTOR, and reenter cell cycle (Dominguez-

Sola et al., 2012; Ersching et al., 2017; Finkin et al., 2019; Luo et al., 2018). 76

Because LCMV-specific GC B cells in acute versus chronic infection appeared to differ at the DZ re-entry step (Figure 3.3B) and because CD40 signals were diminished in the T cell and FDC interaction cluster during chronic infection, we hypothesized that induction of MTOR or Myc target genes would be reduced in chronic infection. However, Myc target genes and MTORC1 signaling were enriched in LZ GC B cells returning to the DZ during both acute and chronic infection, and neither pathway was differently associated with the observed branching between the two infections (Figure 3.3F and 3.4B). These observations suggest that once an effective BCR and help signal was received, engagement of the downstream metabolic transcriptional program proceeded similarly in acute versus chronic infection. It remains possible that non- transcriptional control of metabolism checkpoints has a role, but the lack of differential enrichment of the Myc and MTORC1 signaling genesets suggested that transcriptional control of these pathways was not altered, despite differential

DZ reentry.

We next assessed transcriptional evidence of signaling BCR during chronic infection. Given the abundance and chronicity of antigen during chronic infection, we hypothesized that stronger BCR signaling in GC B cells would perturb GC B cell cluster dynamics. In acute infection, signatures of BCR signaling were enriched in GC B cells re-entering the DZ but not as cells exited the GC reaction (Figure 3.3G) supporting a role for BCR signaling in efficient DZ re-entry. However, despite substantially more antigen in chronic infection, the transcriptional signature of BCR signaling during chronic infection was actually 77 reduced in LCMV-specific GC B cells compared to cells responding to acute infection (Figure 3.3F). Although the LCMV-specific GC B cells that did re-enter the DZ during chronic infection did display modest enrichment for BCR signaling, the enrichment of this signature was both lower in magnitude and temporally delayed in pseudotime space compared to acute infection. These data suggested that BCR signaling in LCMV-specific GC B cells during chronic infection is attenuated, perhaps early in the process of testing BCR after emerging from the

DZ and entering cluster 4. Expression of cell surface inhibitory receptors or intracellular signaling attenuators did not provide an obvious explanation for attenuated BCR signaling during chronic infection (Figure 3.5A). Instead, we observed downregulation of many BCR pathway genes in LCMV-specific GC B cells within cluster 4 during chronic infection (Figure 3.5B). These results suggest that BCR signaling rather than its endocytic functions (i.e. because the ability to elicit Tfh help and receive CD40 signals were intact) were impaired in

LCMV-specific B cells during chronic infection.

Chronic infections can not only impact antibody responses to the persisting infection itself, but also bystander antigen such as vaccines suggesting a role for chronic inflammation in addition to persisting antigen (McKittrick et al.,

2013; Shive et al., 2018; Stelekati and Wherry, 2012; Wiedmann et al., 2000)

Thus, we next sought to determine the role of persistent inflammation in altering

GC B cell differentiation. Genes related to inflammatory responses, interferon-a,

78 and interferon-g were strongly enriched in LCMV-specific GC B cells undergoing differentiation into PC and MB (Figure 3.3F and 3.4D-E). Moreover, even in the pseudotime branch of LZ GC B cells returning to the DZ that was biased to chronic infection, inflammatory signaling gene transcription was also enriched.

Thus, persistent inflammatory signaling during chronic infection was strongly associated with the skewed differentiation of LCMV-specific GC B cell towards the PC and MB and early GC exit. Together, these data suggest that altered sensing of antigen and persistent inflammation are associated with altered B cell differentiation during chronic infection promoting early GC exit and altered reentry into the DZ. In particular, these data implicate a key step in the transition from DZ to LZ (i.e. cluster 4) where BCR signaling attenuation may be important and a role for signals from chronic inflammation throughout the cluster 4 à cluster 2 reentry or cluster 4 à GC exit program, but especially as GC B cells commit to the PC or MB fate.

Figure 3.3

Figure 3.3: Chronic infection promotes terminal B cell differentiation and GC exit.

(A) tSNE and bar plots depicting cluster occupancy of LCMV-specific GC B cells from LCMV-Arm (Acute) and LCMV-CL13 (Chronic) infected mice at day 15 post- infection. (B) Pseudotime analysis (Monocle 2) of light zone GC B cell populations identified using scRNAseq. Pseudotime constructed using all light zone GC B cell populations and visualized separated by infection. Pseudotime is rooted in cluster 4: T cell and FDC interactions. (C) Single-cell GSVA enrichment scores for Basso CD40 signaling genes UP geneset plotted as a function of pseudotime. (D) Single-cell GSVA enrichment scores for Hallmark MYC targets V2 geneset plotted as a function of pseudotime. (E) Single-cell GSVA enrichment scores for Reactome BCR signaling geneset plotted as a function of pseudotime. (F) Single-cell GSVA enrichment scores for Hallmark Inflammatory geneset plotted as a function of pseudotime.

Figure 3.4

Figure 3.4: Pseudotime analysis identifies GC B cell fate decision points within the LZ and correlated gene expression signatures. (A) Pseudotime analysis (Monocle 2) of light zone GC B cell populations identified using scRNAseq. Pseudotime is rooted in cluster 4: T cell and FDC interactions. Plots depict pseudotime, cell cycle stage, cluster identity, and distinct pseudotime states of LCMV-specific LZ GC B cells. (B) Single-cell GSVA enrichment scores for Hallmark MTORC1 signaling geneset plotted as a function of pseudotime. (C) Single-cell GSVA enrichment scores for Hallmark PI3K- MTOR-AKT signaling geneset plotted as a function of pseudotime. (D) Single-cell GSVA enrichment scores for Hallmark Response to Interferon Alpha geneset plotted as a function of pseudotime. (E) Single-cell GSVA enrichment scores for Hallmark Response to Interferon Gamma geneset plotted as a function of pseudotime.

Figure 3.5:

Figure 3.5: BCR signaling pathways are altered in LZ GC B cells during chronic infection. (A) Expression of B cell inhibitory receptors and phosphatases in LCMV-specific B cells in cluster 4: T cell FDC interactions. (B) Expression of genes downstream of the BCR in LCMV-specific B cells in cluster 4: T cell FDC interactions. ***p<0.001 Wilcoxon Rank Sum Test

3.3.3 Chronic inflammation and altered antigen sensing in GC responses during chronic viral infection.

Pseudotime analysis combined with GSVA identified decreased CD40 signaling, impaired BCR signaling, and increased inflammatory signaling associated with terminal differentiation and GC exit of GC B cells during chronic infection. Given the known role for CD40 in regulating GC B cell DZ reentry and positive selection (Gitlin et al., 2014; Shulman et al., 2013; Victora et al., 2010) and that we observed diminished CD40 signals in GC B cells interacting with Tfh or FDC in the LZ during chronic infection, we hypothesized that agonism of the

CD40 pathway during chronic infection would prevent PC differentiation and promote continued participation in the GC reaction. To address this question, we treated mice infected with either acute or chronic LCMV with the blocking antibody anti-CD40L (MR1) or the agonizing antibody anti-CD40 (FGK45) at day

13 pi and analyzed the response at day 16 pi. Agonism of the CD40 pathway during acute infection led to GC B cell expansion whereas CD40L blockade led to dissolution of the GC response and a corresponding increase in PC frequency

(Figure 3.6A). In contrast to acute infection, and in disagreement with our hypothesis, agonism of the CD40 pathway during chronic infection had no effect on LCMV-specific GC B cell, PC, or MB responses (Figure 3.6B). However, we did observe a trend towards MB differentiation during CD40L blockade. Impaired response to CD40 pathway manipulation during chronic infection was unexpected given the transcriptional evidence of diminished but present CD40 84 signals in our scRNAseq analysis. Whether the lack of response to CD40 pathway modulation in chronic infection was due to suboptimal dosing (Beatty et al., 2011), impaired antibody Fc-mediated functions during chronic infection

(Wieland et al., 2015), or functional inhibition of this signaling axis will require further investigation. Together, these data suggest that chronic infection impairs

GC B cell responsiveness to CD40 signals.

In addition to diminished CD40 signaling, we also observed transcriptional evidence of impaired BCR signaling in GC B cells during chronic infection.

Therefore, we next interrogated the impact of BCR signals on GC B cell differentiation. BCR signaling is critical for the maintenance of the GC response and blockade of BCR signaling using the BTK inhibitor has been previously used to inhibit GC B cell responses during acute LCMV infection (Mueller et al., 2015).

Because we observed enrichment of BCR signaling in GC B cells undergoing DZ reentry during acute infection (Figure 3.3G) we hypothesized that inhibiting BCR signaling during acute infection would lead to GC B cell exit and differentiation. In contrast, because the BCR signaling transcriptional network was diminished in chronic infection, we predicted that BCR inhibition would have a smaller impact in

LCMV-CL13 infection. To test these predictions, we administered the BCR- inhibiting drug ibrutinib to mice infected with acute or chronic LCMV starting at d3 as previously described (Figure 3.7A) (Mueller et al., 2015; Yang et al., 2016).

Indeed, BCR inhibition during acute infection led to decreased LCMV-specific GC

B cell responses and increased MB frequencies, consistent with the known role of BCR signaling in GC re-entry (Figure 3.7B). In contrast and as predicted, 85 inhibition of BCR during chronic infection had no impact of LCMV-specific B cell differentiation. These results were consistent with our scRNAseq analysis in

Figure 3.3G and suggest that skewing of GC B cell differentiation toward PC and

MB differentiation and away from DZ reentry during chronic infection is associated with BCR signaling so attenuated that the negative effects on the GC response cannot be furthered by molecular BCR inhibition with ibrutinib.

The data above suggested that chronic infection disrupts BCR signaling and the ability to respond to T cell help. Because IFN-I has been implicated in chronic infection as the driver of pre-GC disruption of B cell responses (Fallet et al., 2016; Moseman et al., 2016; Sammicheli et al., 2016), we next sought to understand whether chronic inflammation is a driver of altered fate decisions after the GC has formed. In support of this, one of the strongest differential signatures in our bulk RNAseq in chapter 2 (Figure 2.7D) was response to interferon and inflammatory signaling. We then assessed these pathways in our scRNA-seq data and also identified response to interferon and inflammatory signals as one of the strongest differential signals in GC B cells responding to acute versus chronic infection (Figure 3.3H). Of note, it has been previously understood that GC B cells are nonresponsive to IFN-I signals owing to their overall low expression of Ifnar1 and Ifnar2 (immgen.org, (Heng et al., 2008)).

However, when individual GC clusters were examined, we observed specific, focal expression of Ifnar1 and Ifnar2 in the testing cluster 4, where GC B cells are interacting with Tfh and FDC (Figure 3.8A). This observation suggested that GC

B cells may still be sensitive to IFN signaling, specifically at a key step in the GC where signals are being integrated for GC B cell fate decisions.

Given these scRNA-seq data and previous studies on the effects of IFN-I on pre-GC B cell responses, we directly interrogated the role of IFN-I on GC B cell fate decisions. To do this, we first tested whether disruption of IFN signaling only in B cells during chronic infection would alter the LCMV-NP-specific GC B cell response. Genetic ablation of IFN signaling only in B cells using

CD19CrexIFNARflox/flox mice during chronic infection resulted in a 1.6-fold reduction in the frequency of LCMV-specific PC but did not alter the GC B cell or MB compartments (Figure 3.8B). Because this reduction in PC frequencies could be due to the impact of type I interferon before the GC response, we next performed systemic antibody blockade of type I interferon signaling in acute and chronic infection as described (Wilson et al., 2013a) starting at d13 pi when GC formation had already been initiated. Systemic blockade of type I interferon signaling beginning at d13 pi did not impact the overall magnitude of the LCMV-NP- specific response (Figure 3.8C) nor did it change LCMV-NP-specific PC, GC B cell or MB frequencies during either acute or chronic infection (Figure 3.8D).

These results support the idea that although Ifnar1 and Ifnar2 are expressed in

LCMV-NP-specific GC B cells, the major impact of type I interferon during chronic infection is to promote PC differentiation prior to the formation of the GC response. Moreover, these data suggested that the IFN and inflammatory signature observed in GC B cells is due to a residual transcriptional effect of IFN-

I signaling prior to GC formation, signals from IFN-g, and/or signals from other inflammatory pathways.

We next wanted to test the inflammatory environment separately from the effects of chronic antigen. To do this, we infected mice with LCMV Arm or Cl13 and then injected the mice with one dose of recombinant phycoerythrin (1x PE;

Figure 3.8D). If inflammation was playing a driving role in skewed GC B cell fate decisions during chronic infection, we predicted that the PE responses in LCMV-

CL13-infected mice would be similarly skewed. Although the magnitude of the

PE-specific B cell response was similar between Arm and clone 13 mice (Figure

3.8F), the differentiation of PE-specific B cell responses in mice with chronic infection was, indeed, altered (Figure 3.8G-H). Specifically, PE-specific PC were increased during chronic infection and the GC B cell response was reduced. The ability of chronic infection to skew responses to a one-time dose of PE supported a dominant role for environment and/or inflammation in the skewing of B cell responses during chronic infection.

We next wondered whether the altered B cell responses to PE during

CL13 infection indicated that inflammation alone causes altered GC B cell fate decisions or whether some alterations could also be caused by chronic antigen as well. We therefore designed a variation of the PE immunization experiment but, instead of a single PE dose, we gave repeated immunizations to expose B cells to antigen “chronically” for 2 weeks (Figure 3.8D). In this model, if chronic antigen played an additive role, we would expect that chronic PE administration

88 in CL13 infection would further skew the PE-specific response to look like those observed against LCMV-NP. Repeated administration of PE increased the magnitude of the total PE-specific B cell response in both acute and chronic

LCMV infection (Figure 3.8E). However, repeated immunization with PE did not impact PC frequencies in either acute or chronic infection suggesting that antigen dose (at least within the range tested) did not play a major role in determining the

PC fate (Figure 3.8G-H). Repeated immunization did reduce the MB compartment in both Arm and clone 13 infection, although the effect was less dramatic in chronic infection (Figure 3.8H). In contrast to PC and MB fates, repeated immunization had a dramatically different impact on GC B cells in acute versus chronic infection. In acute LCMV, repeated immunization doubled the magnitude of the PE-specific GC response (Figure 3.8G-H). In contrast, repeated antigen exposure in the setting of chronic infection resulted in a non- significant increase in the PE-specific GC response (Figure 3.8G-H). Taken together, these data suggest that sustained antigen is better able to induce DZ re-entry in acute infection than in chronic infection and are consistent with the scRNA-seq data on compromised BCR signaling in GC B cells during chronic infection. However, with an increase in total PE-specific B cells and very little change in the PC, MB, and GC fates when PE was given 5 times in CL13 infection, it should be noted that the increase was driven instead by a CD38-GL7-

CD138- B cell response, an uncharacterized subset that may contain age- associated B cells (Rubtsova et al., 2015) or other B cell populations. Taken together, these data suggest that persistent inflammation and/or environmental 89 changes, rather than chronic antigen, is a dominant driver of increased terminal differentiation of GC B cells observed in chronic infection.

Figure 3.6

Figure 3.6: LCMV-specific B cell differentiation is not sensitive to CD40 pathway modulation during chronic infection. CD40 pathway was manipulated in mice with either acute or chronic LCMV by treatment of mice with anti-CD40L (MR1, 200ug/mouse ip) or anti-CD40 (FGK45, 25ug/mouse ip) at d14 post infection followed by analysis at day 15. (A) Frequency of LCMV-specific PC, GC B cells, and MB in LCMV-Arm (acute) infected mice at d16 pi. (B) Frequency of LCMV-specific PC, GC B cells, and MB in LCMV-CL13 (chronic) infected mice at d16 pi. Data points in A and B show mean ± SEM. *p<0.05, **p<0.01, Student’s t-test. 91

Figure 3.7

Figure 3.7: Inhibition of BCR signaling does not impact LCMV-specific B cell responses. (A) Experimental timeline. Mice were infected with either acute or chronic LCMV. Starting at d3 pi mice were treated either with vehicle or vehicle + 250ug Ibrutinib twice daily on days 3, 6, 9, and 12 pi before analysis on day 15 pi. (B) Frequency of LCMV-specific PC, GC B cells, and MB in LCMV-CL13 (chronic) infected mice at d15 pi as treated in A. Data points in B show mean ± SEM. *p<0.05, **p<0.01, Student’s t-test.

Figure 3.8

Figure 3.8: Chronic antigen and inflammation impair the germinal center response during chronic viral infection. (A) tSNE plot depicting expression of Ifnar1 and Ifnar2 in day 15 pi LCMV- specific GC B cell scRNAseq data. (B) Frequency of splenic LCMV-specific plasma cells, germinal center B cells, and memory B cells in wildtype or IFNARf/f infected mice at d15 post-infection. (C) Frequency of LCMV-specific plasma cells, germinal center B cells, and memory B cells in groups depicted in A. (D) Experimental setup for testing the role of inflammation in panels E and F. Mice were split into four groups and infected with either LCMV-Arm or LCMV-CL13. Within these groups, mice either received one 200ug ip immunization of recombinant PE or 5x immunizations with 200ug ip every three days post- infection (0, 3, 6, 9, 12). Mice were analyzed at d15 post infection. (E) Frequency of PE-specific B cells in mice in each group depicted in D. (F) Frequency of PE- specific plasma cells, germinal center B cells, and memory B cells in groups depicted in D. Each plot depicts mean ± SEM with 3-5 mice per group from one of two representative experiments. *p<0.05, **p<0.01, ***p<0.001, Student’s t- test.

3.4 Discussion

Balancing the decision to remain within the GC and continue undergoing affinity maturation or to exit the GC as an antibody-secreting PC or MB is critical for the development of effective humoral immune responses (Cyster and Allen,

2019; Mesin et al., 2016; Shlomchik et al., 2019). Regulation of this checkpoint involves translating BCR signals, antigen-presentation, and Tfh cell help into either PC or MB differentiation or DZ re-entry as discussed in chapter 1.2.4. The studies completed in chapter 2 of this thesis found that chronic viral infection alters virus-specific B cell responses by skewing B cell differentiation away from the GC fate and towards the PC and MB fates. Yet, despite skewed B cell differentiation during chronic infection where and how chronic infection impacts the GC B cell response are unclear. Given that persistent antigen is a feature of chronic viral infection, it is likely that chronic infection may disrupt GC B cell decision making.

The studies outline in this chapter address these questions. Interrogation of the GC B cell response to acute and chronic viral infection at single-cell resolution identified a key GC B cell checkpoint in the LZ that is dysregulated during chronic viral infection. Rather than undergoing positive selection and DZ reentry, GC B cells during chronic infection favor PC and MB differentiation and

GC exit. Inefficient DZ reentry during chronic infection was associated with diminished CD40 signaling, attenuated BCR signaling, and increased inflammatory gene signatures.

The finding that BCR signaling was attenuated in chronic infection was surprising. Instead, we had hypothesized that chronic antigen would promote strong BCR signals leading to PC differentiation consistent with the two-signal

BCR affinity model of differentiation outline in chapter 1. However, this result is consistent with other settings of B cell chronic antigen exposure such as autoimmunity where B cells undergo a state of functional unresponsiveness known as B cell anergy (Yarkoni et al., 2010). The distinction between anergy and BCR attenuation during chronic viral infection (like the distinction between anergy and exhaustion in CD8 T cells) (Schietinger and Greenberg, 2014) is not made clearer by these data.

To dissect the relative influence of chronic antigen and chronic inflammation, we modulated antigen loads, CD40 signals, and inflammation in the LCMV model system. Our studies identified a dominant role for persistent inflammation over chronic antigen exposure in shaping altered GC B cell differentiation during chronic infection. Interrogation of type I interferon signaling did not seem to impact virus-specific GC B cell responses following the formation of the GC, suggesting that the impact of type I interferon signaling on humoral immunity is in shaping the extrafollicular phase of the humoral immune response as has been previously shown but that these pathways do not robustly modulate the GC response. Taken together, our findings have important implications for vaccination. First, it may explain why individuals with chronic inflammation such as the aged or those with pre-existing chronic infections have poor response to vaccination (McKittrick et al., 2013; Shive et al., 2018; Wiedmann et al., 2000). 96

Second, while our studies did not identify a causal soluble inflammatory mediator as a driver of altered GC B cell differentiation, such a molecule could provide a therapeutic target for improved humoral immunity during chronic viral infection or other settings of chronic inflammation.

CHAPTER 4: Concluding Remarks and Future Directions

4.1 Overview of results

Effective humoral immune responses to acute infection and immunization are defined by the induction of high-affinity antibodies generated as the result of

B cell activation and differentiation (Cyster and Allen, 2019; Mesin et al., 2016;

Plotkin, 2008). In contrast to acute infection or immunization, humoral immune responses during chronic infections are often dysregulated. Chronic infections, like HIV (Gray et al., 2007; Lane et al., 1983; Milito et al., 2004; Moir and Fauci,

2017), HCV (Bjøro et al., 1994; Cashman et al., 2014; Logvinoff et al., 2004;

Racanelli et al., 2006), and chronic LCMV (Battegay et al., 1993; Hunziker et al.,

2003; Oldstone and Dixon, 1969), are characterized by increased antibody quantity but decreased antibody quality. Additionally, individuals with pre-existing chronic infections have poorer responses to vaccination (McKittrick et al., 2013;

Shive et al., 2018; Wiedmann et al., 2000). Given the critical role for the GC in regulating both antibody quantity and quality, these defects suggest a central role for alterations in GC biology during chronic infection. Yet, the exact impact of chronic infection on the formation and maintenance of the GC response remains poorly understood. Understanding how chronic viral infections alter the normal humoral immune response can therefore (i) define targets to modulate defective humoral immunity and (ii) give mechanistic insight into the signals that govern the

B cell response. The studies completed in this thesis identify a key LZ checkpoint

98 and underlying cellular and molecular mechanisms that lead to dysregulated humoral immunity in chronic infection.

In chapter 2 of this thesis, the impact of chronic infection on the development of antigen-specific humoral immunity was investigated. Through the use of novel recombinant protein reagents (Fallet et al., 2016; Schweier et al.,

2019; Sommerstein et al., 2015), virus-specific B cell responses were found to be diminished in magnitude and altered in phenotype when compared to B cells specific for the same viral protein responding to acute infection. Chronic infection skewed virus-specific GC B cell responses towards terminal PC and MB differentiation and away from continued participation in the GC response.

Skewing toward the PC and MB fates was associated with decreased serum antibody avidity and fewer accumulated BCR mutations. Transcriptional profiling of LCMV-specific B cells during acute and chronic infection identified early dysregulation of the LCMV-specific GC B cell response resulting in induction of the PC and MB gene signatures during chronic infection. Thus, chronic infection promotes PC and MB differentiation at the expense of the GC fate and this skewing correlates with delayed development of high-affinity antibodies.

In chapter 3 of this thesis, the mechanisms underlying skewed GC B cell differentiation during chronic infection were investigated. Interrogation of the

LCMV-specific GC B cell response using scRNAseq identified a key LZ checkpoint that is dysregulated during chronic infection. Rather than undergoing

DZ re-entry and continued participation in affinity maturation, GC B cells instead were biased towards GC exit and differentiation into the PC and MB fates. Early 99

GC B cell exit during chronic infection occurred in the LZ when GC B cells are interacting with Tfh or FDC. GC B cell exit and differentiation was associated with intact but diminished CD40 signaling, attenuated BCR signaling, and increased inflammatory signaling but not altered induction of the Myc or MTOR gene signatures. Modulation of the CD40 pathway and BCR signaling during chronic infection did not restore altered GC B cell differentiation. Type I interferon was found to have a role in promoting altered B cell differentiation during chronic infection but likely acts before the initiation of the GC response. Thus, the broad inflammatory milieu of chronic infection was found to have a dominate role in stunting GC B cell responses suggesting that chronic infection induces a broad state of immunosuppression.

Taken together, chapters 2 and 3 of this thesis not only extend our understanding of how GC B cell responses are coordinated during chronic infection, but also provide deeper clarity into molecular programs, including specific genes and pathways, that are altered by persisting infection. These studies found that persistent inflammation during chronic infection disrupts a key

DZ re-entry checkpoint in the GC causing skewing of GC B cell differentiation towards PC and MB differentiation and away from continued participation in the

GC reaction. This skewing away from the GC was associated with decreased antibody affinity maturation and gives at least one mechanism by which chronic infection delays development of neutralizing antibody.

As will be discussed below, this work adds to our knowledge of the signals and contexts driving B cell fate decisions by directly perturbing aspects of the 100 system and then interrogating the effects on the GC output. Our observations suggest that the two-signal model of GC B cell differentiation (antigen and T cell help) (Shlomchik et al., 2019) can be affected by inflammation and that inflammation can promote early GC exit even in the setting of abundant antigen and abundant Tfh. The findings outlined in chapters 2 and 3 also suggest persistent inflammatory signals, as is observed in individuals with other chronic infections, may serve as a mechanism for poor vaccine responses in other contexts (McKittrick et al., 2013; Shive et al., 2018; Wiedmann et al., 2000).

4.2 The role of chronic antigen in the regulation of the humoral immune responses to chronic infection

The studies in chapters 2 and 3 identified the problem – that chronic infection led GC B cells in the LZ to preferentially exit as a terminal PC or MB cells rather than re-enter the DZ for further affinity mature. The regulation of this

GC B cell checkpoint has been extensively studied in mouse models of hapten immunization with two model antigens used as immunogens: hen egg lysozyme and the (4-hydroxy-3-nitrophenyl)acetyl (NP) hapten (Gitlin et al., 2014; Ise et al.,

2018; Phan et al., 2006). The primary decision model derived from these immunization experiments is the BCR-based affinity model discussed in chapter

1.2.3, where regulation of the decision to undergo positive selection and DZ reentry or differentiation within the GC depends critically on translating BCR

101 affinity into 1) BCR signals and 2) Tfh help (Cyster and Allen, 2019; Mesin et al.,

2016; Shlomchik et al., 2019). This immunization model defined a “Goldilocks” threshold of “just right” signals that promote positive selection and DZ re-entry, where affinity maturation can continue. The decision to instead exit the GC reaction as a PC or MB typically occurs when the signal is slightly above the

Goldilocks range (PC differentiation) or slightly below (MB differentiation).

In contrast to immunization, the kinetics of antigen exposure during infection is not a bolus. In acute infection, antigen logarithmically rises to an eventual peak followed by a decline, and in chronic infection, antigen also rises logarithmically to a peak but then stabilizes at a high level of persisting antigen available to stimulate B cells. The way that the kinetics of antigen quantity regulates GC B cell fate decision differently than immunization is not clear. In acute infection, the BCR-based affinity model of GC fate decisions would predict that early in infection— as antigen is just beginning to rise but remains below the

DZ re-entry threshold for a given clone—a MB fate for the less competitive clone would be preferred (Laidlaw et al., 2017; Suan et al., 2017a; Weisel et al., 2016).

Then, as antigen becomes more abundant, the affinity requirements for antigen encounter and capture may be less competitive, allowing for more GC B cells to achieve sufficient signaling that exceeds the DZ-reentry threshold and leading to differentiation into the PC fate (Ise et al., 2018; Kräutler et al., 2017; Weisel et al.,

2016). Then, as virus is controlled and antigen decreases, the MB fate may again be preferred. Our data during LCMV-Arm acute infection were consistent with this model and showed that MB cell differentiation is favored early (d10) during 102 the GC response but also later as GC B cell responses are resolved. These results are consistent with recent studies outlining that the GC response to immunization favors output of MB early but PC later during the response (Weisel et al., 2016). At this coarse level, it appears that the differences in antigen kinetics between immunization and acute infection do not alter the relative GC differentiation into the PC or MB cell fates, possibly due to the rapid expansion of available antigen in viral infection that may mimic a bolus delivery of antigen.

In both immunization and acute infection, antigen moves from abundant to limiting. In contrast, during chronic infections antigen levels can often be persistently high and therefore potentially non-limiting for the duration of infection. What happens to GC B cell fate decisions in this setting? A fascinating feature of B cell biology is that the BCR has two mutually exclusive functions, signaling and endocytosis of antigen, that translate BCR affinity into downstream fate decisions (Hou et al., 2006; Shlomchik et al., 2019). In the context of abundant antigen, both BCR signals and capture of antigen would likely be increased for all B cells potentially minimizing differences in BCR affinity between

B cell clones. Therefore, based on the BCR-based affinity model it could be predicted that lower affinity B cells would be more apt to receive sufficient signals to exit as PC when in limiting antigen environments they would only have sufficient affinity to be positively selected for more affinity maturation. Likewise, persistent antigen (Wherry et al., 2003a) and abundant Tfh (Greczmiel et al.,

2017; Vella et al., 2017), as observed during chronic infection, may allow cells to

103 be shunted towards a MB fate when an environment of limiting antigen might otherwise place the B cell below the survival threshold.

It can then be predicted using the BCR-affinity differentiation model that high and sustained antigen loads during chronic infection would induce strong

BCR signals and favor PC differentiation over the GC B cell or MB fate. Indeed, this is what was observed in chapter 2. PC differentiation was favored prior to formation of the GC and through the initial 3 weeks of the immune response.

Only when viral loads begin to be controlled within the spleen (d10-d21) do GC B cells begin to expand during chronic infection (Wherry et al., 2003a). One particularly interesting finding is that although, chronic infection promotes PC and

MB differentiation early during the GC response, it isn’t until GC B cell numbers begin to increase that differences in serum antibody avidity between acute and chronic infection begin to emerge. Why this occurs is still unclear but it does suggest that some critical mass of GC B cells has to be achieved for GC B cell competition and affinity maturation to occur in an efficient manner.

We then used experimental modulation of Tfh signals and BCR signaling to test how these two signals and the BCR affinity model work in the context of acute and chronic infection. Consistent with previous studies, agonism of the

CD40 pathway during acute infection—mimicking Tfh help— promoted the GC B cell fate while diminishing the MB fate (PC responses were unchanged)

(Ersching et al., 2017; Victora et al., 2010). As expected, blocking BCR signaling had the opposite effect; BCR inhibition diminished the GC response and led to an increase in MB. For all intents and purposes, virus-specific B cell responses 104 during acute infection acted like previously described anti-hapten responses.

However, responses during chronic infection, where Tfh are abundant and antigen is plentiful (Greczmiel et al., 2017; Vella et al., 2017; Wherry et al.,

2003a), did not behave as expected. It would be predicted based on the BCR- affinity model that increasing CD40 signals would further promote PC differentiation whereas limiting BCR signaling would promote the GC B cell fate.

In fact, when experimentally modulated, neither inhibition of BCR signaling nor manipulation of the CD40 pathway impacted GC B cell fate during chronic infection.

Further interrogation of these pathways using scRNAseq revealed diminished CD40 signaling and BCR pathway inhibition in GC B cells during chronic infection. While transcriptional evidence of CD40 signaling was seen in

GC B cells undergoing positive selection and GC exit, transcriptional evidence of

BCR signaling was seen only in cells undergoing positive selection. This suggests that strong T cell help through CD40 signals in the absence of strong

BCR signals may promote GC B cell differentiation and exit during chronic infection. However, why CD40 agonism did not promote further PC differentiation during chronic infection is unclear. Whether the lack of response to CD40 pathway modulation in chronic infection was due to suboptimal dosing (Beatty et al., 2011), impaired antibody Fc-mediated functions during chronic infection

(Wieland et al., 2015), or functional inhibition of this signaling axis warrants further study. Perhaps more intriguingly, these results suggest that there may be

105 an additional signal, like persistent inflammation, that influences how CD40 and

BCR signals are integrated in GC B cells during chronic infection.

4.3 Inflammation: Dominant role of persistent inflammation over persistent antigen as a driver of B cell dysregulation in chronic infection

In Chapter 3, we used single cell sequencing at day 15 to determine the molecular pathways associated with divergence in the terminal PC and MB fates.

As would be predicted by the experiments with CD40 agonism and BCR inhibition, Transcriptional signatures of CD40 signaling in conjunction with BCR signaling favored GC B cell DZ reentry whereas CD40 signals in the absence of

BCR signaling favored GC exit in acute infection. In contrast, virus-specific GC B cell undergoing GC differentiation and GC exit during chronic infection were enriched and had distinct transcriptional signatures consistent with diminished

CD40 signals, attenuated BCR signaling, and high inflammatory signaling.

Agonism of the CD40 pathway and inhibition of BCR signaling did not impact virus-specific B cell responses during chronic infection. These results show that although virus-specific B cell responses to chronic infection did act as predicted by the BCR-affinity model, with GC B cells favoring differentiation over DZ reentry, the transcriptional mechanism of this choice is distinct.

106

However, a third feature was prominent in the pseudotime analyses of B cell fate decisions in chronic versus acute infection: inflammatory signaling.

Inflammation during chronic infections is sustained and complex and involves signals from type I and type II interferons and TNF family members (Stelekati and

Wherry, 2012; Zuniga et al., 2015). Inflammation, particularly type I interferon, has been shown to promote PC differentiation at the extrafollicular checkpoint prior to GC B cell differentiation and GC formation yet the impact of inflammation on GC B cell fate decisions following the formation of the GC was previously unclear (Fallet et al., 2016; Moseman et al., 2016; Sammicheli et al., 2016). Our studies in chapter 3 found that type I interferon does impact B cell responses during chronic viral infection but that the impact of this signaling pathway on B cell differentiation is likely prior to the formation of the GC consistent with other studies. Furthermore, development of an experimental system capable of discriminating the impact of chronic antigen versus persistent inflammation in vivo identified persistent inflammation rather than chronic antigen exposure as the mechanism underlying GC B cell skewing during chronic infection.

How attenuated BCR signaling, diminished CD40 signals, and increased inflammatory signaling leads to PC differentiation during chronic infection is still unclear. Integration of BCR and Tfh help signals in GC B cells is rewired such that GC B cells require both signals for positive selection and differentiation (Luo et al., 2018). As previously discussed, the balance between these two signals is critical for influencing GC B cell differentiation and DZ re-entry. Importantly,

CD40 signals in GC B cells lead to induction of the canonical NfKB pathway 107 which, in addition to activating c-Myc, can promote both PC differentiation and

GC B cell survival (Luo et al., 2018; Milanovic et al., 2017; Roy et al., 2019; De

Silva et al., 2016). While our studies did not identify a single inflammatory mediator that was causal many inflammatory signaling pathways, like those downstream of the type I interferon and TNF-a pathways, coalesce on NfKB activation (Hayden and Ghosh, 2014; Vallabhapurapu and Karin, 2009; Yang et al., 2001). Aberrant regulation of NfKB can lead to GC disruption (Heise et al.,

2014), PC differentiation, and, in some cases PC malignancies (Annunziata et al., 2007; Demchenko et al., 2010). Our results are consistent with a model in which sustained and elevated inflammatory signals drive NfKB activation leading to PC differentiation and GC exit. Interrogation of the NfKB pathway is an important future avenue of study. Further, it remains possible that inflammatory signaling pathways may converge and modify other signaling pathways in GC B cells like the PI3K-MTOR-Akt pathway possibly leading to altered GC B cell differentiation (Kaur et al., 2008; Luo et al., 2018, 2019).

4.4 B cell dysfunction in other settings of chronic antigen exposure – similarities and differences in lessons learned from chronic LCMV

Chronic LCMV infection is not the only situation in which chronic antigen exposure has been found to alter B cell responses. HIV and HCV lead to altered

108

B cell differentiation (Moir et al., 2008; Racanelli et al., 2006). Humans and mice infected with malaria-causing parasites also display altered humoral immunity, specifically altered MB differentiation (Kim et al., 2019; Obeng-Adjei et al., 2017;

Portugal et al., 2015; Thouvenel et al., 2016). Additionally, B cell anergy is a well described mechanism by which B cell recognizing self-antigen are rendered functionally inert in an effort to maintain immune tolerance (Yarkoni et al., 2010).

However, the context in which B cells see chronic self-antigen is different when compared to chronic viral infection indicating that additional cues beyond antigen loads, like inflammation or periodicity of exposure, are likely critical for regulating the outcome of the GC response.

It is striking to note the similarity between GC B cells in chronic infection and B cell anergy in autoimmunity. In the case of autoimmunity, the introduction of an inflammatory signal, such as infection or type I interferon, can overcome B cell anergy and lead to robust anti-self antibody responses (Domeier et al.,

2018). Anergic B cells are rendered functionally inert by chronic self-antigen exposure in a mechanism that involves inhibition of proximal BCR signaling through upregulation of PTEN and SHIP-1 (both inhibitors of proximal BCR signaling) (Yarkoni et al., 2010). However, despite the loss of BCR signaling capacity, anergic B cells are still able to process and present antigen and are responsive to CD40 signals (Eris et al., 2006). Virus-specific GC B cells in chronic infection display elevated levels of PDK1 and altered PI3K-Akt signaling, transcriptional evidence of diminished but present CD40 signal, and transcriptional signatures consistent with impaired BCR signaling. Whether 109 common mechanisms regulate B cell anergy and dysfunctional B cell responses to chronic infection warrant future study.

It is interesting to consider whether all adaptive immune lineages share a common transcriptional mechanism that renders them functionally diminished in settings of chronic antigen exposure and persistent inflammation. CD8 T cell exhaustion is a well characterized form of T cell dysfunction that occurs during chronic viral infection (McLane et al., 2019). T cell exhaustion is characterized by the sequential upregulation of inhibitory receptors that render these cells unable to undergo full activation and mediate cytotoxicity after undergoing TCR signaling. Interestingly, many of the same transcription factors and transcriptional networks seen in exhausted T cells are also seen in virus-specific B cells during chronic infection. For example, IRF4, which is important for driving PC responses is also a critical regulator of CD8 T cell fate (Man et al., 2017). Similar to the

BCR-affinity based model of selection, strong TCR signals increase the intracellular levels of IRF4 which favors the exhaustion fate whereas weak TCR signals favor the memory fate (Man et al., 2017). Similarly, the Blimp1-Bcl6 axis which controls PC vs GC B cell fates also influences exhausted T cell fates with

Bcl6+ CD8 T cells representing a more progenitor-like pool (Wu et al., 2016) and

Blimp1+ cells favoring the terminal effector population (Shin et al., 2009). Further analysis of our scRNAseq and bulk RNAseq data shows a dramatic overlap between the core T cell exhaustion transcriptional signature and B cells from chronic infection (unpublished data). This observation suggests that chronic infection may induce a shared program of immune dysfunction among all 110 adaptive immune lineages. Further study will be necessary to define these shared networks.

Why the adaptive immune system evolved mechanisms to limit T cell overactivation in the setting of chronic infection is clear as too vigorous a T cell response could lead to excessive immunopathology and mortality. However, why chronic infection would promote PC differentiation and delayed affinity maturation of the humoral immune response is an open question. Does this confer a survival advantage to the host?

4.5 Impact of these studies on approaches to vaccination

The studies outlined in chapters 2 and 3 of this thesis may have several important implications for new approaches to vaccination in individuals with pre- existing chronic inflammatory conditions. In particular, the finding that inflammatory signaling plays a dominant role over antigen loads in shaping the output of the humoral immune response. Chronic inflammation and the resulting skewing of B cell differentiation may explain why aged individuals or those with existing chronic infections respond poorly to immunization (McKittrick et al.,

2013; Shive et al., 2018; Wiedmann et al., 2000). Recently, it has been demonstrated that repeated antigen exposure, in the context of rational vaccine design for HIV, can greatly augment GC responses leading to increased affinity maturation and the production of broadly neutralizing antibodies (Cirelli et al.,

2019; Tam et al., 2016). Our studies support this approach and further suggest 111 that the choice of adjuvants that induce low but not high levels of inflammation will be important for optimally guiding affinity maturation of the GC response and balancing the output of the GC response.

Importantly, the work contained within this thesis extends our knowledge of the mechanisms underlying GC B cell positive selection and GC exit and therapeutic manipulation of the mechanisms described could allow finer control over the output of the humoral immune response. While the accumulation of affinity-increasing mutations is an important goal for vaccine responses to HIV

(Briney et al., 2016; Doria-Rose and Joyce, 2015; Kwong and Mascola, 2018), the optimal humoral immune response for other pathogens may be different. For example, seasonal re-exposure to influenza infection or vaccination leads to repeated recruitment of MB cells into the recall response (Linderman and

Hensley, 2016; Zost et al., 2019). One impact of this is that individuals have focusing of antibody responses towards epitopes the immune system has already seen. However, due to genetic alterations in the influenza viral proteins neuraminidase and hemagglutinin from year to year, a process known as antigenic drift, epitopes in these proteins that mediate neutralization can change and antibody responses that were previously effective can be rendered ineffective (Cobey and Hensley, 2017; Wilson et al., 2013b). Therefore, for vaccine responses against influenza, it may be desired to generate more diverse antibody responses against a variety of epitopes rather than high-affinity responses against one or a few epitopes. While our studies have focused on humoral immune responses that necessitate high-affinity antibodies to mediate 112 viral control, it is likely that by using similar approaches to track and transcriptionally profile virus-specific B cell responses in other mouse models will yield further mechanistic insights relevant to vaccine design.

4.6: Adding a fourth dimension - Spatial dysregulation at the tissue level during chronic infection is a big black box and may underly many aspects of adaptive immune dysfunction during chronic infection

One important and often overlooked aspect of many studies in immunology is the role that immune cell spatial location within tissues plays in regulating the immune response. While the studies completed in chapters 2 and

3 of this thesis provide increased mechanistic understanding of the GC B cell response during viral infection and identify a key GC B cell checkpoint that is dysregulated during chronic viral infection, our studies involved processing of tissues and loss of spatial information. The importance of GC structure is perhaps best illustrated by mice with lack the FOXO1 transcription factor in the B cell lineage. Loss of FOXO1 results in disorganization of the DZ GC B cell program and loss of GC polarization (Dominguez-Sola et al., 2015; Inoue et al.,

2017). GC disorganization in the setting of FOXO1 deficiency results in impaired affinity maturation (Dominguez-Sola et al., 2015). GC B cell movement between the LZ and DZ is critical as GC B cells without CXCR4 are unable to migrate to

113 the DZ efficiently and display impaired affinity maturation relative to CXCR4- sufficient GC B cells (Allen et al., 2004).

Many infections cause alterations to tissue homeostasis and tissue architecture that are resolved upon pathogen clearance. In contrast to acute infections, chronic infections can lead to immunopathology and disruption of tissue architecture. In HIV, disruption of the gut barrier can lead to bacterial translocation and inflammation (Brenchley et al., 2006; Klatt et al., 2013; Nazli et al., 2010). Chronic LCMV infection leads to disruption of splenic architecture as macrophages and stromal cells, like FDC, are infected and eliminated by the cytotoxic T cell response (Matter et al., 2011; Mueller et al., 2007; Wilson et al.,

2013a). Given the central role that FDC play in regulating the GC response (Allen and Cyster, 2008; Cyster and Allen, 2019; Heesters et al., 2014), it is likely that disruption of the splenic architecture does play a role in modulating virus-specific

GC B cell responses. Indeed, blockade of type I interferon can prevent splenic architecture disruption and leads to induction of the GC program in virus-specific

B cells during chronic infection (Wilson et al., 2013a). While the studies in this thesis mainly focused on phenotypic and transcriptomic characterization of virus- specific B cell responses during chronic infection, we did analyze the histological appearance of GCs. We found that histological evidence of GCs was present during acute and chronic infection and that GC size was the same regardless of infection. However, while GCs were the same size, the shape of GCs during chronic infection were distinct, displaying less circularity. Whether altered GC shape contributes to impaired affinity maturation and skewed GC B cell 114 differentiation during chronic infection remains unclear and will require further study.

It should be noted that we observed phenotypic and transcriptional polarization of GC B cells into the LZ and DZ compartments in both acute and chronic infection. Whether these observations translate into distinct spatial localization of these cells is an important future question. It will be especially interesting to see whether the clusters identified in our scRNAseq analysis of the

GC B cell response translate into unique spatial localization of these clusters.

However, one limitation to conducting these studies this is the current technical limitations of immunofluorescent microscopy. Traditional microscopy approaches have been limited to simultaneous analysis of at best 6 markers. Recent technical advances in our ability to assess a higher number of markers simultaneously using microscopy may advance our understanding of how spatial localization informs GC responses. Techniques such as co-detecting by indexing

(CODEX) (Goltsev et al., 2018) or multiplexed ion beam imaging (MIBI) (Angelo et al., 2014) allow the detection of 50+ markers on the same tissue sample. More widespread adoption of these technologies could allow finer dissection of the GC response and allow identification and validation of not only our scRNAseq data but other researcher’s scRNAseq findings. One particularly interesting possibility is the use of these techniques to understand how acute infection versus chronic infection perturbs tissue-level homeostasis and what the return to homeostasis looks like in these two settings.

115

4.7 Concluding thoughts

The work presented in this thesis provides novel insight into the molecular control of B cell responses to chronic viral infections. Prior to this work, it was unclear whether or how chronic viral infection impacted GC B cell responses and more specifically how chronic antigen and inflammation impact GC B cell differentiation. To address these questions, we interrogated the impact of chronic viral infection on B cell differentiation and fate decisions within the GC. Focusing specifically on virus-specific B cells, we employed a combination of bulk and single cell RNA sequencing (scRNA-seq), together with BCR repertoire profiling and in vivo cellular experiments. The studies in chapters 2 and 3 of this thesis revealed a key GC checkpoint dysregulated during chronic infection. Specifically, during chronic viral infection, GC B cells inefficiently re-enter the dark zone during the cyclic GC reaction and instead GC B cell fate is skewed towards premature exit as MB or terminal PC. This short-circuiting of the cyclic process of

GC B cell development occurred at a key light zone step in GC B cell differentiation and was associated with decreased affinity maturation and serum antibody avidity. Single cell profiling pointed to attenuated BCR signaling and

CD40 signaling, but amplified inflammatory signals associated with this altered

GC developmental pattern. Manipulation of the CD40 pathway, antigen chronicity, and inflammation identified a dominant role for inflammation in promoting skewed GC B cell differentiation during chronic infection. Our studies reveal the mechanistic roles of inflammation, CD40 signaling and antigen loads

116 on B cell differentiation during chronic viral infection and identify key pathways that can be targeted to modulate humoral immune responses and promote affinity maturation during anti-viral immune responses and immunization.

Additionally, these observations have important implications for generating optimal antibody responses during chronic infections.

117

BIBLIOGRAPHY

Ahmed, R. (1984). Selection of genetic variants of lymphocytic choriomeningitis virus in spleens of persistently infected mice. Role in suppression of cytotoxic T lymphocyte response and viral persistence. J. Exp. Med.

Alamyar, E., Duroux, P., Lefranc, M.-P., and Giudicelli, V. (2012). IMGT(®) tools for the nucleotide analysis of immunoglobulin (IG) and T cell receptor (TR) V-(D)- J repertoires, polymorphisms, and IG mutations: IMGT/V-QUEST and IMGT/HighV-QUEST for NGS. Methods Mol. Biol. 882, 569–604.

Allen, C.D.C., and Cyster, J.G. (2008). Follicular dendritic cell networks of primary follicles and germinal centers: Phenotype and function. Semin. Immunol. 20, 14–25.

Allen, C.D.C., Ansel, K.M., Low, C., Lesley, R., Tamamura, H., Fujii, N., and Cyster, J.G. (2004). Germinal center dark and light zone organization is mediated by CXCR4 and CXCR5. Nat. Immunol. 5, 943–952.

Allman, D., Wilmore, J.R., and Gaudette, B.T. (2019). The continuing story of T- cell independent antibodies. Immunol. Rev.

Angelo, M., Bendall, S.C., Finck, R., Hale, M.B., Hitzman, C., Borowsky, A.D., Levenson, R.M., Lowe, J.B., Liu, S.D., Zhao, S., et al. (2014). Multiplexed ion beam imaging of human breast tumors. Nat. Med.

Annunziata, C.M., Davis, R.E., Demchenko, Y., Bellamy, W., Gabrea, A., Zhan, F., Lenz, G., Hanamura, I., Wright, G., Xiao, W., et al. (2007). Frequent engagement of the classical and alternative NF-kappaB pathways by diverse genetic abnormalities in multiple myeloma. Cancer Cell 12, 115–130.

Ansel, K.M., Ngo, V.N., Hyman, P.L., Luther, S.A., Förster, R., Sedgwlck, J.D., Browning, J.L., Upp, M., and Cyster, J.G. (2000). A chemokine-driven positive feedback loop organizes lymphoid follicles. Nature.

Barnett, B.E., Staupe, R.P., Odorizzi, P.M., Palko, O., Tomov, V.T., Mahan, A.E., Gunn, B., Chen, D., Paley, M.A., Alter, G., et al. (2016). Cutting edge: B cell- 118 intrinsic T-bet expression is required to control chronic viral infection. J. Immunol. 197.

Barton, E., Mandal, P., and Speck, S.H. (2011). Pathogenesis and Host Control of Gammaherpesviruses: Lessons from the Mouse. Annu. Rev. Immunol.

Batista, F.D., and Harwood, N.E. (2009). The who, how and where of antigen presentation to B cells. Nat. Rev. Immunol. 9, nri2454.

Batista, F.D., and Neuberger, M.S. (2000). B cells extract and present immobilized antigen: implications for affinity discrimination. EMBO J. 19, 513– 520.

Battegay, M., Moskophidis, D., Waldner, H., Bründler, M.A., Fung-Leung, W.P., Mak, T.W., Hengartner, H., and Zinkernagel, R.M. (1993). Impairment and delay of neutralizing antiviral antibody responses by virus-specific cytotoxic T cells. J. Immunol. 151, 5408–5415.

Beatty, G.L., Chiorean, E.G., Fishman, M.P., Saboury, B., Teitelbaum, U.R., Sun, W., Huhn, R.D., Song, W., Li, D., Sharp, L.L., et al. (2011). CD40 agonists alter tumor stroma and show efficacy against pancreatic carcinoma in mice and humans. Science 331, 1612–1616.

Benhamron, S., Pattanayak, S.P., Berger, M., and Tirosh, B. (2015). mTOR Activation Promotes Plasma Cell Differentiation and Bypasses XBP-1 for Immunoglobulin Secretion. Mol. Cell. Biol.

Berek, C., Berger, A., and Apel, M. (1991). Maturation of the immune response in germinal centers. Cell 67, 1121–1129.

Bhattacharya, D., Cheah, M.T., Franco, C.B., Hosen, N., Pin, C.L., Sha, W.C., and Weissman, I.L. (2007). Transcriptional Profiling of Antigen-Dependent Murine B Cell Differentiation and Memory Formation. J. Immunol. 179, 6808– 6819.

Bjøro, K., Frøland, S.S., Yun, Z., Samdal, H.H., and Haaland, T. (1994). Hepatitis C infection in patients with primary hypogammaglobulinemia after treatment with contaminated immune globulin. N. Engl. J. Med. 331, 1607–1611. 119

Bonsignori, M., Liao, H.X., Gao, F., Williams, W.B., Alam, S.M., Montefiori, D.C., and Haynes, B.F. (2017). Antibody-virus co-evolution in HIV infection: paths for HIV vaccine development. Immunol. Rev.

Brenchley, J.M., Price, D.A., Schacker, T.W., Asher, T.E., Silvestri, G., Rao, S., Kazzaz, Z., Bornstein, E., Lambotte, O., Altmann, D., et al. (2006). Microbial translocation is a cause of systemic immune activation in chronic HIV infection. Nat. Med. 12, 1365–1371.

Briney, B., Sok, D., Jardine, J.G., Kulp, D.W., Skog, P., Menis, S., Jacak, R., Kalyuzhniy, O., de Val, N., Sesterhenn, F., et al. (2016). Tailored Immunogens Direct Affinity Maturation toward HIV Neutralizing Antibodies. Cell.

Butler, A., Hoffman, P., Smibert, P., Papalexi, E., and Satija, R. (2018). Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420.

Calado, D.P., Sasaki, Y., Godinho, S.A., Pellerin, A., Köchert, K., Sleckman, B.P., Alborán, I.M. de, Janz, M., Rodig, S., Rajewsky, K., et al. (2012). The cell- cycle regulator c-Myc is essential for the formation and maintenance of germinal centers. Nat. Immunol. 13, 1092.

Cao, W., Henry, M.D., Borrow, P., Yamada, H., Elder, J.H., Ravkov, E. V, Nichol, S.T., Compans, R.W., Campbell, K.P., and Oldstone, M.B. (1998). Identification of alpha-dystroglycan as a receptor for lymphocytic choriomeningitis virus and Lassa fever virus. Science 282, 2079–2081.

Cashman, S.B., Marsden, B.D., and Dustin, L.B. (2014). The humoral immune response to HCV: Understanding is key to vaccine development. Front. Immunol.

Charrad, M., Ghazzali, N., Boiteau, V., and Niknafs, A. (2015). NbClust : An R Package for Determining the Relevant Number of Clusters in a Data Set . J. Stat. Softw.

Chen, C., Zhai, S., Zhang, L., Chen, J., Long, X., Qin, J., Li, J., Huo, R., and Wang, X. (2018). Uhrf1 regulates germinal center B cell expansion and affinity maturation to control viral infection. J. Exp. Med. 215, jem.20171815.

120

Chevrier, S., Kratina, T., Emslie, D., Tarlinton, D.M., and Corcoran, L.M. (2017). IL4 and IL21 cooperate to induce the high Bcl6 protein level required for germinal center formation. Immunol. Cell Biol. 95, 925–932.

Chou, C., Verbaro, D.J., Tonc, E., Holmgren, M., Cella, M., Colonna, M., Bhattacharya, D., and Egawa, T. (2016). The Transcription Factor AP4 Mediates Resolution of Chronic Viral Infection through Amplification of Germinal Center B Cell Responses. Immunity 45, 570–582.

Cirelli, K.M., Carnathan, D.G., Nogal, B., Martin, J.T., Rodriguez, O.L., Upadhyay, A.A., Enemuo, C.A., Gebru, E.H., Choe, Y., Viviano, F., et al. (2019). Slow Delivery Immunization Enhances HIV Neutralizing Antibody and Germinal Center Responses via Modulation of Immunodominance. Cell 177, 1153- 1171.e28.

Clingan, J.M., and Matloubian, M. (2013). B Cell–Intrinsic TLR7 Signaling Is Required for Optimal B Cell Responses during Chronic Viral Infection. J. Immunol.

Cobey, S., and Hensley, S.E. (2017). Immune history and influenza virus susceptibility. Curr. Opin. Virol.

Crawford, A., Angelosanto, J.M., Kao, C., Doering, T.A., Odorizzi, P.M., Barnett, B.E., and Wherry, E.J. (2014). Molecular and transcriptional basis of CD4+ T cell dysfunction during chronic infection. Immunity 40, 289–302.

Crotty, S. (2015). A brief history of T cell help to B cells. Nat. Rev. Immunol. 15, 185–189.

Crotty, S. (2019). T Follicular Helper Cell Biology: A Decade of Discovery and Diseases. Immunity 50, 1132–1148.

Cyster, J.G., and Allen, C.D.C. (2019). B Cell Responses: Cell Interaction Dynamics and Decisions. Cell.

Cyster, J.G., Ansel, K.M., Reif, K., Ekland, E.H., Hyman, P.L., Tang, H.L., Luther, S.A., and Ngo, V.N. (2000). Follicular stromal cells and lymphocyte homing to follicles. Immunol. Rev. 176, 181–193. 121

Daugan, M., Murira, A., Mindt, B.C., Germain, A., Tarrab, E., Lapierre, P., Fritz, J.H., and Lamarre, A. (2016). Type I Interferon Impairs Specific Antibody Responses Early during Establishment of LCMV Infection. Front. Immunol. 7, 564.

Demchenko, Y.N., Glebov, O.K., Zingone, A., Keats, J.J., Bergsagel, P.L., and Kuehl, W.M. (2010). Classical and/or alternative NF-kappaB pathway activation in multiple myeloma. Blood 115, 3541–3552.

Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M., and Gingeras, T.R. (2013). STAR: ultrafast universal RNA-seq aligner - Supplementary Data. Bioinformatics.

Doering, T.A., Crawford, A., Angelosanto, J.M., Paley, M.A., Ziegler, C.G., and Wherry, E.J. (2012). Article Network Analysis Reveals Centrally Connected Genes and Pathways Involved in CD8 + T Cell Exhaustion versus Memory. 1–15.

Domeier, P.P., Chodisetti, S.B., Schell, S.L., Kawasawa, Y.I., Fasnacht, M.J., Soni, C., and Rahman, Z.S.M. (2018). B-Cell-Intrinsic Type 1 Interferon Signaling Is Crucial for Loss of Tolerance and the Development of Autoreactive B Cells. Cell Rep.

Dominguez-Sola, D., Victora, G.D., Ying, C.Y., Phan, R.T., Saito, M., Nussenzweig, M.C., and Dalla-Favera, R. (2012). The proto-oncogene MYC is required for selection in the germinal center and cyclic reentry. Nat. Immunol. 13, 1083.

Dominguez-Sola, D., Kung, J., Holmes, A.B., Wells, V.A., Mo, T., Basso, K., and Dalla-Favera, R. (2015). The FOXO1 Transcription Factor Instructs the Germinal Center Dark Zone Program. Immunity 43, 1064–1074.

Doria-Rose, N.A., and Joyce, M.G. (2015). Strategies to guide the antibody affinity maturation process. Curr. Opin. Virol.

Doria-Rose, N.A., Schramm, C.A., Gorman, J., Moore, P.L., Bhiman, J.N., DeKosky, B.J., Ernandes, M.J., Georgiev, I.S., Kim, H.J., Pancera, M., et al. (2014). Developmental pathway for potent V1V2-directed HIV-neutralizing antibodies. Nature 508.

122

Doria-Rose, N.A., Bhiman, J.N., Roark, R.S., Schramm, C.A., Gorman, J., Chuang, G.-Y., Pancera, M., Cale, E.M., Ernandes, M.J., Louder, M.K., et al. (2016). New member of the V1V2-directed CAP256-VRC26 lineage that shows increased breadth and exceptional potency. J. Virol. 90.

Dörner, T., and Radbruch, A. (2007). Antibodies and B Cell Memory in Viral Immunity. Immunity.

Erickson, L.D., Durell, B.G., Vogel, L.A., O’Connor, B.P., Cascalho, M., Yasui, T., Kikutani, H., and Noelle, R.J. (2002). Short-circuiting long-lived humoral immunity by the heightened engagement of CD40. J. Clin. Invest. 109, 613–620.

Eris, J.M., Basten, A., Brink, R., Doherty, K., Kehry, M.R., and Hodgkin, P.D. (2006). Anergic self-reactive B cells present self antigen and respond normally to CD40-dependent T-cell signals but are defective in antigen-receptor-mediated functions. Proc. Natl. Acad. Sci.

Ersching, J., Efeyan, A., Mesin, L., Jacobsen, J.T., Pasqual, G., Grabiner, B.C., Dominguez-Sola, D., Sabatini, D.M., and Victora, G.D. (2017). Germinal Center Selection and Affinity Maturation Require Dynamic Regulation of mTORC1 Kinase. Immunity.

Eschli, B., Zellweger, R.M., Wepf, A., Lang, K.S., Quirin, K., Weber, J., Zinkernagel, R.M., and Hengartner, H. (2007). Early antibodies specific for the neutralizing epitope on the receptor binding subunit of the lymphocytic choriomeningitis virus glycoprotein fail to neutralize the virus. J. Virol. 81, 11650– 11657.

Fahey, L.M., Wilson, E.B., Elsaesser, H., Fistonich, C.D., McGavern, D.B., and Brooks, D.G. (2011). Viral persistence redirects CD4 T cell differentiation toward T follicular helper cells. J. Exp. Med.

Fallet, B., Narr, K., Ertuna, Y.I., Remy, M., Sommerstein, R., Cornille, K., Kreutzfeldt, M., Page, N., Zimmer, G., Geier, F., et al. (2016). Interferon-driven deletion of antiviral B cells at the onset of chronic infection. Sci. Immunol. 1.

Finkin, S., Hartweger, H., Oliveira, T.Y., Kara, E.E., and Nussenzweig, M.C. (2019). Protein Amounts of the MYC Transcription Factor Determine Germinal Center B Cell Division Capacity. Immunity. 123

Foy*, T.M., Aruffo, A., Bajorath, J., Buhlmann*+, J.E., and Noelle*, R.J. (1996). IMMUNE REGULATION BY CD40 AND ITS LIGAND GP39. Immunology 14, 591–617.

Foy, T.M., Laman, J.D., Ledbetter, J.A., Aruffo, A., Claassen, E., and Noelle, R.J. (1994). gp39-CD40 interactions are essential for germinal center formation and the development of B cell memory. J. Exp. Med. 180, 157–163.

Frankish, A., Diekhans, M., Ferreira, A.-M., Johnson, R., Jungreis, I., Loveland, J., Mudge, J.M., Sisu, C., Wright, J., Armstrong, J., et al. (2019). GENCODE reference annotation for the human and mouse genomes. Nucleic Acids Res. 47, D766–D773.

Fukazawa, Y., Lum, R., Okoye, A.A., Park, H., Matsuda, K., Bae, J.Y., Hagen, S.I., Shoemaker, R., Deleage, C., Lucero, C., et al. (2015). B cell follicle sanctuary permits persistent productive simian immunodeficiency virus infection in elite controllers. Nat. Med.

Garnier, S. (2018). viridis: Default Color Maps from “matplotlib.”

Gitlin, A.D., Shulman, Z., and Nussenzweig, M.C. (2014). Clonal selection in the germinal centre by regulated proliferation and hypermutation. Nature 509, 637.

Gitlin, A.D., Mayer, C.T., Oliveira, T.Y., Shulman, Z., Jones, M.J.K., Koren, A., and Nussenzweig, M.C. (2015). HUMORAL IMMUNITY. T cell help controls the speed of the cell cycle in germinal center B cells. Science 349, 643–646.

Goltsev, Y., Samusik, N., Kennedy-Darling, J., Bhate, S., Hale, M., Vazquez, G., Black, S., and Nolan, G.P. (2018). Deep Profiling of Mouse Splenic Architecture with CODEX Multiplexed Imaging. Cell 174, 968-981.e15.

Gray, E.S., Moore, P.L., Choge, I.A., Decker, J.M., Bibollet-Ruche, F., Li, H., Leseka, N., Treurnicht, F., Mlisana, K., Shaw, G.M., et al. (2007). Neutralizing Antibody Responses in Acute Human Immunodeficiency Virus Type 1 Subtype C Infection▿. J. Virol. 81, 6187–6196.

Greczmiel, U., Kräutler, N.J., Pedrioli, A., Bartsch, I., Agnellini, P., Bedenikovic, G., Harker, J., Richter, K., and Oxenius, A. (2017). Sustained T follicular helper 124 cell response is essential for control of chronic viral infection. Sci. Immunol. 2.

Gupta, N.T., Vander Heiden, J.A., Uduman, M., Gadala-Maria, D., Yaari, G., and Kleinstein, S.H. (2015). Change-O: A toolkit for analyzing large-scale B cell immunoglobulin repertoire sequencing data. Bioinformatics.

Hangartner, L., Zellweger, R.M., Giobbi, M., Weber, J., Eschli, B., McCoy, K.D., Harris, N., Recher, M., Zinkernagel, R.M., and Hengartner, H. (2006). Nonneutralizing antibodies binding to the surface glycoprotein of lymphocytic choriomeningitis virus reduce early virus spread. J. Exp. Med.

Hänzelmann, S., Castelo, R., and Guinney, J. (2013). GSVA: Gene set variation analysis for microarray and RNA-Seq data. BMC Bioinformatics.

Hardy, R.R., and Hayakawa, K. (2001). B cell development pathways. Annu. Rev. Immunol.

Hayden, M.S., and Ghosh, S. (2014). Regulation of NF-κB by TNF family cytokines. Semin. Immunol.

Haynes, B.F., Gilbert, P.B., McElrath, M.J., Zolla-Pazner, S., Tomaras, G.D., Alam, S.M., Evans, D.T., Montefiori, D.C., Karnasuta, C., Sutthent, R., et al. (2012). Immune-Correlates Analysis of an HIV-1 Vaccine Efficacy Trial. N. Engl. J. Med.

Haynes, N.M., Allen, C.D.C., Lesley, R., Ansel, K.M., Killeen, N., and Cyster, J.G. (2007). Role of CXCR5 and CCR7 in Follicular Th Cell Positioning and Appearance of a Programmed Cell Death Gene-1High Germinal Center- Associated Subpopulation. J. Immunol. 179, 5099–5108.

Heesters, B.A., Myers, R.C., and Carroll, M.C. (2014). Follicular dendritic cells: dynamic antigen libraries. Nat. Rev. Immunol. 14, nri3689.

Heesters, B.A.A., Chatterjee, P., Kim, Y.-A.A., Gonzalez, S.F.F., Kuligowski, M.P.P., Kirchhausen, T., and Carroll, M.C.C. (2013). Endocytosis and Recycling of Immune Complexes by Follicular Dendritic Cells Enhances B Cell Antigen Binding and Activation. Immunity 38, 1164–1175.

125

Vander Heiden, J.A., Yaari, G., Uduman, M., Stern, J.N.H., O’Connor, K.C., Hafler, D.A., Vigneault, F., and Kleinstein, S.H. (2014). pRESTO: a toolkit for processing high-throughput sequencing raw reads of lymphocyte receptor repertoires. Bioinformatics 30, 1930–1932.

Heise, N., De Silva, N.S., Silva, K., Carette, A., Simonetti, G., Pasparakis, M., and Klein, U. (2014). Germinal center B cell maintenance and differentiation are controlled by distinct NF-κB transcription factor subunits. J. Exp. Med. 211, 2103–2118.

Heng, T.S.P., Painter, M.W., and Immunological Genome Project Consortium (2008). The Immunological Genome Project: networks of gene expression in immune cells. Nat. Immunol. 9, 1091–1094.

Horwitz, J.A., Bar-On, Y., Lu, C.L., Fera, D., Lockhart, A.A.K., Lorenzi, J.C.C., Nogueira, L., Golijanin, J., Scheid, J.F., Seaman, M.S., et al. (2017). Non- neutralizing Antibodies Alter the Course of HIV-1 Infection In Vivo. Cell.

Hou, P., Araujo, E., Zhao, T., Zhang, M., Massenburg, D., Veselits, M., Doyle, C., Dinner, A.R., and Clark, M.R. (2006). B Cell Antigen Receptor Signaling and Internalization Are Mutually Exclusive Events. PLoS Biol. 4, e200.

Huang, M., Wang, J., Torre, E., Dueck, H., Shaffer, S., Bonasio, R., Murray, J.I., Raj, A., Li, M., and Zhang, N.R. (2018). SAVER: Gene expression recovery for single-cell RNA sequencing. Nat. Methods.

Hunziker, L., Recher, M., Macpherson, A.J., Ciurea, A., Freigang, S., Hengartner, H., and Zinkernagel, R.M. (2003). Hypergammaglobulinemia and autoantibody induction mechanisms in viral infections. Nat. Immunol. 4, 343–349.

Inoue, T., Shinnakasu, R., Ise, W., Kawai, C., Egawa, T., and Kurosaki, T. (2017). The transcription factor Foxo1 controls germinal center B cell proliferation in response to T cell help. J. Exp. Med.

Ise, W., and Kurosaki, T. (2019). Plasma cell differentiation during the germinal center reaction. Immunol. Rev.

Ise, W., Fujii, K., Shiroguchi, K., Ito, A., Kometani, K., Takeda, K., Kawakami, E., 126

Yamashita, K., Suzuki, K., Okada, T., et al. (2018). T Follicular Helper Cell- Germinal Center B Cell Interaction Strength Regulates Entry into Plasma Cell or Recycling Germinal Center Cell Fate. Immunity 48, 702-715.e4-715.e4.

Jacob, J., Kelsoe, G., Rajewsky, K., and Weiss, U. (1991). Intraclonal generation of antibody mutants in germinal centres. Nature 354, 354389a0.

Jones, D.D., Gaudette, B.T., Wilmore, J.R., Chernova, I., Bortnick, A., Weiss, B.M., and Allman, D. (2016). MTOR has distinct functions in generating versus sustaining humoral immunity. J. Clin. Invest.

Junt, T., Scandella, E., and Ludewig, B. (2008). Form follows function: lymphoid tissue microarchitecture in antimicrobial immune defence. Nat. Rev. Immunol. 8, 764–775.

Kaur, S., Sassano, A., Dolniak, B., Joshi, S., Majchrzak-Kita, B., Baker, D.P., Hay, N., Fish, E.N., and Platanias, L.C. (2008). Role of the Akt pathway in mRNA translation of interferon-stimulated genes. Proc. Natl. Acad. Sci.

Kawabe, T., Naka, T., Yoshida, K., Tanaka, T., Fujiwara, H., Suematsu, S., Yoshida, N., Kishimoto, T., and Kikutani, H. (1994). The immune responses in CD40-deficient mice: impaired immunoglobulin class switching and germinal center formation. Immunity 1, 167–178.

Kerfoot, S.M., Yaari, G., Patel, J.R., Johnson, K.L., Gonzalez, D.G., Kleinstein, S.H., and Haberman, A.M. (2011). Germinal Center B Cell and T Follicular Helper Cell Development Initiates in the Interfollicular Zone. Immunity 34, 947– 960.

Kim, C.C., Baccarella, A.M., Bayat, A., Pepper, M., and Fontana, M.F. (2019). FCRL5+ Memory B Cells Exhibit Robust Recall Responses. Cell Rep. 27, 1446- 1460.e4.

Kinchen, V.J., Zahid, M.N., Flyak, A.I., Soliman, M.G., Learn, G.H., Wang, S., Davidson, E., Doranz, B.J., Ray, S.C., Cox, A.L., et al. (2018). Broadly Neutralizing Antibody Mediated Clearance of Human Hepatitis C Virus Infection. Cell Host Microbe 24, 717-730.e5.

127

Kitano, M., Moriyama, S., Ando, Y., Hikida, M., Mori, Y., Kurosaki, T., and Okada, T. (2011). Bcl6 Protein Expression Shapes Pre-Germinal Center B Cell Dynamics and Follicular Helper T Cell Heterogeneity. Immunity 34, 961–972.

Klatt, N.R., Funderburg, N.T., and Brenchley, J.M. (2013). Microbial translocation, immune activation, and HIV disease. Trends Microbiol. 21, 6–13.

Klein, J.S., and Bjorkman, P.J. (2010). Few and far between: How HIV may be evading antibody avidity. PLoS Pathog.

Klein, U., Casola, S., Cattoretti, G., Shen, Q., Lia, M., Mo, T., Ludwig, T., Rajewsky, K., and Dalla-Favera, R. (2006). Transcription factor IRF4 controls plasma cell differentiation and class-switch recombination. Nat. Immunol. 7, ni1357.

Kleinstein, S.H., Louzoun, Y., and Shlomchik, M.J. (2003). Estimating Hypermutation Rates from Clonal Tree Data. J. Immunol.

Kolde, R. (2019). pheatmap: Pretty Heatmaps.

Kräutler, N.J., Suan, D., Butt, D., Bourne, K., Hermes, J.R., Chan, T.D., Sundling, C., Kaplan, W., Schofield, P., Jackson, J., et al. (2017). Differentiation of germinal center B cells into plasma cells is initiated by high-affinity antigen and completed by Tfh cells. J. Exp. Med. 214, jem.20161533.

Kunz, S. (2009). Receptor binding and cell entry of Old World arenaviruses reveal novel aspects of virus-host interaction. Virology 387, 245–249.

Kurosaki, T., Kometani, K., and Ise, W. (2015). F O C U S O N 5 0 y e a r s of R B c I e s E l Memory B cells. Nat. Publ. Gr.

Kwak, K., Akkaya, M., and Pierce, S.K. (2019). B cell signaling in context. Nat. Immunol. 20, 963–969.

Kwong, P.D., and Mascola, J.R. (2018). HIV-1 Vaccines Based on Antibody Identification, B Cell Ontogeny, and Epitope Structure. Immunity.

128

Laidlaw, B.J., Schmidt, T.H., Green, J.A., Allen, C.D.C., Okada, T., and Cyster, J.G. (2017). The Eph-related tyrosine kinase ligand Ephrin-B1 marks germinal center and memory precursor B cells. J. Exp. Med.

Lane, H.C., Masur, H., Edgar, L.C., Whalen, G., Rook, A.H., and Fauci, A.S. (1983). Abnormalities of B-cell activation and immunoregulation in patients with the acquired immunodeficiency syndrome. N. Engl. J. Med. 309, 453–458.

Lefranc, M.P., Giudicelli, V., Duroux, P., Jabado-Michaloud, J., Folch, G., Aouinti, S., Carillon, E., Duvergey, H., Houles, A., Paysan-Lafosse, T., et al. (2015). IMGT R, the international ImMunoGeneTics information system R 25 years on. Nucleic Acids Res.

Li, S., Lefranc, M.-P., Miles, J.J., Alamyar, E., Giudicelli, V., Duroux, P., Freeman, J.D., Corbin, V.D.A., Scheerlinck, J.-P., Frohman, M.A., et al. (2013). IMGT/HighV QUEST paradigm for T cell receptor IMGT clonotype diversity and next generation repertoire immunoprofiling. Nat. Commun. 4, 2333.

Liberzon, A., Birger, C., Thorvaldsdóttir, H., Ghandi, M., Mesirov, J.P., and Tamayo, P. (2015). The Molecular Signatures Database Hallmark Gene Set Collection. Cell Syst.

Linderman, S.L., and Hensley, S.E. (2016). Antibodies with “Original Antigenic Sin” Properties Are Valuable Components of Secondary Immune Responses to Influenza Viruses. PLoS Pathog. 12, e1005806.

Logvinoff, C., Major, M.E., Oldach, D., Heyward, S., Talal, A., Balfe, P., Feinstone, S.M., Alter, H., Rice, C.M., and McKeating, J.A. (2004). Neutralizing antibody response during acute and chronic hepatitis C virus infection. Proc. Natl. Acad. Sci. U. S. A. 101, 10149–10154.

Luo, W., Weisel, F., and Shlomchik, M.J. (2018). B Cell Receptor and CD40 Signaling Are Rewired for Synergistic Induction of the c-Myc Transcription Factor in Germinal Center B Cells. Immunity 48, 313-326.e5-326.e5.

Luo, W., Hawse, W., Conter, L., Trivedi, N., Weisel, F., Wikenheiser, D., Cattley, R.T., and Shlomchik, M.J. (2019). The AKT kinase signaling network is rewired by PTEN to control proximal BCR signaling in germinal center B cells. Nat. Immunol. 20, 736–746. 129

MacLennan, I.C.M., Toellner, K.-M., Cunningham, A.F., Serre, K., Sze, D.M.-Y., Zuniga, E., Cook, M.C., and Vinuesa, C.G. (2003). Extrafollicular antibody responses. Immunol. Rev. 194, 8–18.

Man, K., Gabriel, S.S., Liao, Y., Gloury, R., Preston, S., Henstridge, D.C., Pellegrini, M., Zehn, D., Berberich-Siebelt, F., Febbraio, M.A., et al. (2017). Transcription Factor IRF4 Promotes CD8+ T Cell Exhaustion and Limits the Development of Memory-like T Cells during Chronic Infection. Immunity.

Matloubian, B.M., Somasundaram, T., Kolhekar, S.R., Selvakumar, R., and Ahmed, R. (1990). Genetic Basis of Viral Persistence : Single Amino Acid. 172.

Matloubian, M., Kolhekar, S.R., Somasundaram, T., and Ahmed, R. (1993). Molecular determinants of macrophage tropism and viral persistence: importance of single amino acid changes in the polymerase and glycoprotein of lymphocytic choriomeningitis virus. J. Virol.

Matloubian, M., Concepcion, R.J., and Ahmed, R. (1994). CD4+ T cells are required to sustain CD8+ cytotoxic T-cell responses during chronic viral infection. J. Virol.

Matter, M.S., Hilmenyuk, T., Claus, C., Marone, R., Schürch, C., Tinguely, M., Terracciano, L., Luther, S.A., and Ochsenbein, A.F. (2011). Destruction of lymphoid organ architecture and hepatitis caused by CD4 + T cells. PLoS One.

Mayer, C.T., Gazumyan, A., Kara, E.E., Gitlin, A.D., Golijanin, J., Viant, C., Pai, J., Oliveira, T.Y., Wang, Q., Escolano, A., et al. (2017). The microanatomic segregation of selection by apoptosis in the germinal center. Science (80-. ).

Mcheyzer-Williams, L.J., Milpied, P.J., Okitsu, S.L., and Mcheyzer-Williams, M.G. (2015). Class-switched memory B cells remodel BCRs within secondary germinal centers. Nat. Immunol.

McKittrick, N., Frank, I., Jacobson, J.M., White, C.J., Kim, D., Kappes, R., DiGiorgio, C., Kenney, T., Boyer, J., and Tebas, P. (2013). Improved immunogenicity with high-dose seasonal influenza vaccine in HIV-infected persons: a single-center, parallel, randomized trial. Ann. Intern. Med. 158, 19–26.

130

McLane, L.M., Abdel-Hakeem, M.S., and Wherry, E.J. (2019). CD8 T Cell Exhaustion During Chronic Viral Infection and Cancer. Annu. Rev. Immunol. 37, 457–495.

Mesin, L., Ersching, J., and Victora, G.D.D. (2016). Germinal Center B Cell Dynamics. Immunity 45, 471–482.

Milanovic, M., Heise, N., De Silva, N.S., Anderson, M.M., Silva, K., Carette, A., Orelli, F., Bhagat, G., and Klein, U. (2017). Differential requirements for the canonical NF-κB transcription factors c-REL and RELA during the generation and activation of mature B cells. Immunol. Cell Biol.

Milito, A. De, Nilsson, A., Titanji, K., Thorstensson, R., Reizenstein, E., Narita, M., Grutzmeier, S., Sönnerborg, A., and Chiodi, F. (2004). Mechanisms of hypergammaglobulinemia and impaired antigen-specific humoral immunity in HIV-1 infection. Blood 103, 2180–2186.

Minamitani, T., Yasui, T., Ma, Y., Zhou, H., Okuzaki, D., Tsai, C.-Y., Sakakibara, S., Gewurz, B.E., Kieff, E., and Kikutani, H. (2015). Evasion of affinity-based selection in germinal centers by Epstein–Barr virus LMP2A. Proc. Natl. Acad. Sci.

Moir, S., and Fauci, A.S. (2017). B-cell responses to HIV infection. Immunol. Rev.

Moir, S., Ho, J., Malaspina, A., Wang, W., DiPoto, A.C., O’Shea, M.A., Roby, G., Kottilil, S., Arthos, J., Proschan, M.A., et al. (2008). Evidence for HIV-associated B cell exhaustion in a dysfunctional memory B cell compartment in HIV-infected viremic individuals. J. Exp. Med. 205, 1797–1805.

Moseman, E.A., Wu, T., Torre, J.C. de la, Schwartzberg, P.L., and McGavern, D.B. (2016). Type I interferon suppresses virus-specific B cell responses by modulating CD8(+) T cell differentiation. Sci. Immunol. 1.

Mueller, J., Matloubian, M., and Zikherman, J. (2015). Cutting Edge: An In Vivo Reporter Reveals Active B Cell Receptor Signaling in the Germinal Center. J. Immunol. 194, 2993–2997.

131

Mueller, S.N., Matloubian, M., Clemens, D.M., Sharpe, A.H., Freeman, G.J., Gangappa, S., Larsen, C.P., and Ahmed, R. (2007). Viral targeting of fibroblastic reticular cells contributes to immunosuppression and persistence during chronic infection. Proc. Natl. Acad. Sci.

Muramatsu, M., Kinoshita, K., Fagarasan, S., Yamada, S., Shinkai, Y., and Honjo, T. (2000). Class Switch Recombination and Hypermutation Require Activation-Induced Cytidine Deaminase (AID), a Potential RNA Editing Enzyme. Cell 102, 553–563.

Muramatsu, M., Davidson, N.O., Sankaranand, V.S., Sugai, M., Honjo, T., Kinoshita, K., and Anant, S. (2002). Specific Expression of Activation-induced Cytidine Deaminase (AID), a Novel Member of the RNA-editing Deaminase Family in Germinal Center B Cells. J. Biol. Chem.

Nawa, M. (1992). An Enzyme-Linked Immunosorbent Assay Using a Chaotropic Agent (Sodium Thiocyanate) for Sero-type Specific Reaction between Crude Dengue Viral Antigen and Anti-Dengue Mouse Antibody.

Nazli, A., Chan, O., Dobson-Belaire, W.N., Ouellet, M., Tremblay, M.J., Gray- Owen, S.D., Arsenault, A.L., and Kaushic, C. (2010). Exposure to HIV-1 directly impairs mucosal epithelial barrier integrity allowing microbial translocation. PLoS Pathog. 6, e1000852.

Nutt, S.L., Hodgkin, P.D., Tarlinton, D.M., and Corcoran, L.M. (2015). The generation of antibody-secreting plasma cells. Nat. Rev. Immunol. 15, 160–171.

Obeng-Adjei, N., Portugal, S., Holla, P., Li, S., Sohn, H., Ambegaonkar, A., Skinner, J., Bowyer, G., Doumbo, O.K., Traore, B., et al. (2017). Malaria-induced interferon-γ drives the expansion of Tbethi atypical memory B cells. PLoS Pathog. 13, e1006576.

Ochiai, K., Maienschein-Cline, M., and Simonetti, G. (2013). Transcriptional Regulation of Germinal Center B and Plasma Cell Fates by Dynamical Control of IRF4. Immunity.

Odegard, V.H., and Schatz, D.G. (2006). Targeting of somatic hypermutation. Nat. Rev. Immunol. 6, 573–583.

132

Odorizzi, P.M., Pauken, K.E., Paley, M.A., Sharpe, A., and Wherry, E.J. (2015). Genetic absence of PD-1 promotes accumulation of terminally differentiated exhausted CD8+ T cells. J. Exp. Med. 212, 1125–1137.

Oldstone, M.B.A., and Dixon, F.J. (1969). PATHOGENESIS OF CHRONIC DISEASE ASSOCIATED WITH PERSISTENT LYMPHOCYTIC CHORIOMENINGITIS VIRAL INFECTION. J. Exp. Med. 129.

Oliviero, B., Mantovani, S., Ludovisi, S., Varchetta, S., Mele, D., Paolucci, S., Baldanti, F., and Mondelli, M.U. (2015). Skewed B cells in chronic hepatitis C virus infection maintain their ability to respond to virus-induced activation. J. Viral Hepat.

Pereira, J.P., Kelly, L.M., Cyster, J.G., and jason, J.G.C. (2010). Finding the right niche: B-cell migration in the early phases of T-dependent antibody responses. Int. Immunol. 22, 413–419.

Phan, T.G., Paus, D., Chan, T.D., Turner, M.L., Nutt, S.L., Basten, A., and Brink, R. (2006). High affinity germinal center B cells are actively selected into the plasma cell compartment. J. Exp. Med. 203, 2419–2424.

Pillai, S., Cariappa, A., and Moran, S.T. (2005). Marginal zone B cells. Annu. Rev. Immunol. 23, 161–196.

Plotkin, S.A. (2008). Correlates of Vaccine-Induced Immunity. Clin. Infect. Dis. 47, 401–409.

Portugal, S., Tipton, C.M., Sohn, H., Kone, Y., Wang, J., Li, S., Skinner, J., Virtaneva, K., Sturdevant, D.E., Porcella, S.F., et al. (2015). Malaria-associated atypical memory B cells exhibit markedly reduced B cell receptor signaling and effector function. Elife 4.

Prigge, J.R., Hoyt, T.R., Dobrinen, E., Capecchi, M.R., Schmidt, E.E., and Meissner, N. (2015). Type I IFNs Act upon Hematopoietic Progenitors To Protect and Maintain Hematopoiesis during Pneumocystis Lung Infection in Mice. J. Immunol.

Pullen, G.R., Fitzgerald, M.G., and Hosking, C.S. (1986). Antibody avidity 133 determination by ELISA using thiocyanate elution. J. Immunol. Methods 86, 83– 87.

Qiu, X., Hill, A., Packer, J., Lin, D., Ma, Y.A., and Trapnell, C. (2017a). Single-cell mRNA quantification and differential analysis with Census. Nat. Methods.

Qiu, X., Mao, Q., Tang, Y., Wang, L., Chawla, R., Pliner, H.A., and Trapnell, C. (2017b). Reversed graph embedding resolves complex single-cell trajectories. Nat. Methods 14, 979–982.

R Core Team (2019). R: A Language and Environment for Statistical Computing.

Racanelli, V., Frassanito, M.A., Leone, P., Galiano, M., Re, V. De, Silvestris, F., and Dammacco, F. (2006). Antibody Production and In Vitro Behavior of CD27- Defined B-Cell Subsets: Persistent Hepatitis C Virus Infection Changes the Rules. J. Virol. 80, 3923–3934.

Reif, K., Ekland, E.H., Ohl, L., Nakano, H., Lipp, M., Förster, R., and Cyster, J.G. (2002). Balanced responsiveness to chemoattractants from adjacent zones determines B-cell position. Nature 416, 94.

Reinhardt, R.L., Liang, H.-E., and Locksley, R.M. (2009). Cytokine-secreting follicular T cells shape the antibody repertoire. Nat. Immunol. 10, 385–393.

Richter, K., and Oxenius, A. (2013). Non-neutralizing antibodies protect from chronic LCMV infection independently of activating FcγR or complement. Eur. J. Immunol. 43, 2349–2360.

Rickert, R.C., Roes, J., and Rajewsky, K. (1997). B lymphocyte-specific, Cre- mediated mutagenesis in mice. Nucleic Acids Res. 25, 1317–1318.

Ritchie, M.E., Phipson, B., Wu, D., Hu, Y., Law, C.W., Shi, W., and Smyth, G.K. (2015). Linear Models for Microarray and RNA-Seq Data. Nucleic Acids Res.

Rodda, L.B., Bannard, O., Ludewig, B., Nagasawa, T., and Cyster, J.G. (2015). Phenotypic and Morphological Properties of Germinal Center Dark Zone Cxcl12- Expressing Reticular Cells. J. Immunol. 195, 4781–4791. 134

Rodda, L.B., Lu, E., Bennett, M.L., Sokol, C.L., Wang, X., Luther, S.A., Barres, B.A., Luster, A.D., Ye, C.J., and Cyster, J.G. (2018). Single-Cell RNA Sequencing of Lymph Node Stromal Cells Reveals Niche-Associated Heterogeneity. Immunity 48, 1014-1028.e6-1028.e6.

Roy, K., Mitchell, S., Liu, Y., Ohta, S., Lin, Y. sheng, Metzig, M.O., Nutt, S.L., and Hoffmann, A. (2019). A Regulatory Circuit Controlling the Dynamics of NFκB cRel Transitions B Cells from Proliferation to Plasma Cell Differentiation. Immunity.

Rubtsova, K., Rubtsov, A. V., Cancro, M.P., and Marrack, P. (2015). Age- Associated B Cells: A T-bet–Dependent Effector with Roles in Protective and Pathogenic Immunity. J. Immunol.

Rueden, C.T., Schindelin, J., Hiner, M.C., DeZonia, B.E., Walter, A.E., Arena, E.T., and Eliceiri, K.W. (2017). ImageJ2: ImageJ for the next generation of scientific image data. BMC Bioinformatics 18, 529.

Sammicheli, S., Kuka, M., Lucia, P. Di, Oya, N.J. de, Giovanni, M. De, Fioravanti, J., Cristofani, C., Maganuco, C.G., Fallet, B., Ganzer, L., et al. (2016). Inflammatory monocytes hinder antiviral B cell responses. Sci. Immunol. 1.

Sander, S., Chu, V.T., Yasuda, T., Franklin, A., Graf, R., Calado, D.P., Li, S., Imami, K., Selbach, M., Di Virgilio, M., et al. (2015). PI3 Kinase and FOXO1 Transcription Factor Activity Differentially Control B Cells in the Germinal Center Light and Dark Zones. Immunity 43, 1075–1086.

Schacker, T.W., Brenchley, J.M., Beilman, G.J., Reilly, C., Pambuccian, S.E., Taylor, J., Skarda, D., Larson, M., Douek, D.C., and Haase, A.T. (2006). Lymphatic tissue fibrosis is associated with reduced numbers of naïve CD4+ T cells in human immunodeficiency virus type 1 infection. Clin. Vaccine Immunol.

Schietinger, A., and Greenberg, P.D. (2014). Tolerance and exhaustion: Defining mechanisms of T cell dysfunction. Trends Immunol.

Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., et al. (2012). Fiji: An open-source platform for biological-image analysis. Nat. Methods.

135

Schweier, O., Aichele, U., Marx, A.-F., Straub, T., Verbeek, J.S., Pinschewer, D.D., and Pircher, H. (2019). Residual LCMV antigen in transiently CD4+ T cell- depleted mice induces high levels of virus-specific antibodies but only limited B- cell memory. Eur. J. Immunol. 49, 626–637.

Schwickert, T.A., Victora, G.D., Fooksman, D.R., Kamphorst, A.O., Mugnier, M.R., Gitlin, A.D., Dustin, M.L., and Nussenzweig, M.C. (2011). A dynamic T cell–limited checkpoint regulates affinity-dependent B cell entry into the germinal center. J. Exp. Med. 208, 1243–1252.

Sciammas, R., Shaffer, A.L., Schatz, J.H., Zhao, H., Staudt, L.M., and Singh, H. (2006). Graded Expression of Interferon Regulatory Factor-4 Coordinates Isotype Switching with Plasma Cell Differentiation. Immunity 25, 225–236.

Seiler, P., Kalinke, U., Rülicke, T., Bucher, E.M., Böse, C., Zinkernagel, R.M., and Hengartner, H. (1998). Enhanced virus clearance by early inducible lymphocytic choriomeningitis virus-neutralizing antibodies in immunoglobulin- transgenic mice. J. Virol. 72, 2253–2258.

Shaffer, A.L., Yu, X., He, Y., Boldrick, J., Chan, E.P., and Staudt, L.M. (2000). BCL-6 Represses Genes that Function in Lymphocyte Differentiation, Inflammation, and Cell Cycle Control. Immunity 13, 199–212.

Shi, W., Liao, Y., Willis, S.N., Taubenheim, N., Inouye, M., Tarlinton, D.M., Smyth, G.K., Hodgkin, P.D., Nutt, S.L., and Corcoran, L.M. (2015). Transcriptional profiling of mouse B cell terminal differentiation defines a signature for antibody-secreting plasma cells. Nat. Immunol. 16, 663–673.

Shin, H., Blackburn, S.D., Intlekofer, A.M., Kao, C., Angelosanto, J.M., Reiner, S.L., and Wherry, E.J. (2009). A Role for the Transcriptional Repressor Blimp-1 in CD8+ T Cell Exhaustion during Chronic Viral Infection. Immunity.

Shinnakasu, R., Inoue, T., Kometani, K., Moriyama, S., Adachi, Y., Nakayama, M., Takahashi, Y., Fukuyama, H., Okada, T., and Kurosaki, T. (2016). Regulated selection of germinal-center cells into the memory B cell compartment. Nat. Immunol. 17, 861–869.

Shive, C.L., Judge, C.J., Clagett, B., Kalayjian, R.C., Osborn, M., Sherman, K.E., Fichtenbaum, C., Gandhi, R.T., Kang, M., Popkin, D.L., et al. (2018). Pre-vaccine 136 plasma levels of soluble inflammatory indices negatively predict responses to HAV, HBV, and tetanus vaccines in HCV and HIV infection. Vaccine 36, 453– 460.

Shlomchik, M.J., and Weisel, F. (2012). Germinal center selection and the development of memory B and plasma cells. Immunol. Rev. 247, 52–63.

Shlomchik, M.J., Luo, W., and Weisel, F. (2019). Linking signaling and selection in the germinal center. Immunol. Rev. 288, 49–63.

Shulman, Z., Gitlin, A.D., Targ, S., Jankovic, M., Pasqual, G., Nussenzweig, M.C., and Victora, G.D. (2013). T follicular helper cell dynamics in germinal centers. Science 341, 673–677.

De Silva, N.S., and Klein, U. (2015). Dynamics of B cells in germinal centres. Nat. Rev. Immunol.

De Silva, N.S., Anderson, M.M., Carette, A., Silva, K., Heise, N., Bhagat, G., and Klein, U. (2016). Transcription factors of the alternative NF-κB pathway are required for germinal center B-cell development. Proc. Natl. Acad. Sci.

Slifka, M.K., Antia, R., Whitmire, J.K., and Ahmed, R. (1998). Humoral Immunity Due to Long-Lived Plasma Cells. Immunity 8, 363–372.

Sommerstein, R., Flatz, L., Remy, M.M., Malinge, P., Magistrelli, G., Fischer, N., Sahin, M., Bergthaler, A., Igonet, S., ter Meulen, J., et al. (2015). Arenavirus Glycan Shield Promotes Neutralizing Antibody Evasion and Protracted Infection. PLoS Pathog.

Stelekati, E., and Wherry, E.J. (2012). Chronic bystander infections and immunity to unrelated antigens. Cell Host Microbe.

Stewart, I., Radtke, D., Phillips, B., McGowan, S.J., and Bannard, O. (2018). Germinal Center B Cells Replace Their Antigen Receptors in Dark Zones and Fail Light Zone Entry when Immunoglobulin Gene Mutations are Damaging. Immunity.

137

Straub, T., Schweier, O., Bruns, M., Nimmerjahn, F., Waisman, A., and Pircher, H. (2013). Nucleoprotein-specific nonneutralizing antibodies speed up LCMV elimination independently of complement and FcγR. Eur. J. Immunol. 43, 2338– 2348.

Suan, D., Nguyen, A., Moran, I., Bourne, K., Hermes, J.R., Arshi, M., Hampton, H.R., Tomura, M., Miwa, Y., Kelleher, A.D., et al. (2015). T Follicular Helper Cells Have Distinct Modes of Migration and Molecular Signatures in Naive and Memory Immune Responses. Immunity 42, 704–718.

Suan, D., Kräutler, N.J., Maag, J.L.V. V, Butt, D., Bourne, K., Hermes, J.R., Avery, D.T., Young, C., Statham, A., Elliott, M., et al. (2017a). CCR6 Defines Memory B Cell Precursors in Mouse and Human Germinal Centers, Revealing Light-Zone Location and Predominant Low Antigen Affinity. Immunity 47, 1142- 1153.e4-1153.e4.

Suan, D., Sundling, C., and Brink, R. (2017b). Plasma cell and memory B cell differentiation from the germinal center. Curr. Opin. Immunol. 45, 97–102.

Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gillette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., et al. (2005). Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci.

Sullivan, B.M., Emonet, S.F., Welch, M.J., Lee, A.M., Campbell, K.P., de la Torre, J.C., Oldstone, M.B., Torre, J.C. de la, and Oldstone, M.B. (2011). Point mutation in the glycoprotein of lymphocytic choriomeningitis virus is necessary for receptor binding, dendritic cell infection, and long-term persistence. Proc. Natl. Acad. Sci. 108, 2969–2974.

Szczawinska-Poplonyk, A., Breborowicz, A., Samara, H., Ossowska, L., and Dworacki, G. (2015). Impaired antigen-specific immune response to vaccines in children with antibody production defects. Clin. Vaccine Immunol.

Takahashi, Y., Dutta, P.R., Cerasoli, D.M., and Kelsoe, G. (1998). In Situ Studies of the Primary Immune Response to (4-Hydroxy-3-Nitrophenyl)Acetyl. V. Affinity Maturation Develops in Two Stages of Clonal Selection. J. Exp. Med.

Tam, H.H., Melo, M.B., Kang, M., Pelet, J.M., Ruda, V.M., Foley, M.H., Hu, J.K., 138

Kumari, S., Crampton, J., Baldeon, A.D., et al. (2016). Sustained antigen availability during germinal center initiation enhances antibody responses to vaccination. Proc. Natl. Acad. Sci.

Thouvenel, C.D., Kim, K.S., Holder, A., Krishnamurty, A.T., Portugal, S., Crompton, P.D., Rawlings, D.J., Pepper, M., and Keitany, G.J. (2016). Somatically Hypermutated Plasmodium-Specific IgM+ Memory B Cells Are Rapid, Plastic, Early Responders upon Malaria Rechallenge. Immunity.

Tirosh, I., Izar, B., Prakadan, S.M., Wadsworth, M.H., Treacy, D., Trombetta, J.J., Rotem, A., Rodman, C., Lian, C., Murphy, G., et al. (2016). Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science (80-. ).

Urata, S., and Yasuda, J. (2012). Molecular mechanism of arenavirus assembly and budding. Viruses 4, 2049–2079.

Vallabhapurapu, S., and Karin, M. (2009). Regulation and Function of NF-κB Transcription Factors in the Immune System. Immunology 27, 693–733.

Vella, L.A., Herati, R.S., and Wherry, E.J. (2017). CD4+ T Cell Differentiation in Chronic Viral Infections: The Tfh Perspective. Trends Mol. Med. 23, 1072–1087.

Victora, G.D., Schwickert, T.A., Fooksman, D.R., Kamphorst, A.O., Meyer- Hermann, M., Dustin, M.L., and Nussenzweig, M.C. (2010). Germinal center dynamics revealed by multiphoton microscopy with a photoactivatable fluorescent reporter. Cell 143, 592–605.

Victora, G.D., Dominguez-Sola, D., Holmes, A.B., Deroubaix, S., Dalla-Favera, R., and Nussenzweig, M.C. (2012). Identification of human germinal center light and dark zone cells and their relationship to human B-cell lymphomas. Blood.

Virgin, H.W., Wherry, E.J., and Ahmed, R. (2009). Redefining chronic viral infection. Cell 138, 30–50.

Wang, Q., Kieffer-Kwon, K.-R., Oliveira, T.Y., Mayer, C.T., Yao, K., Pai, J., Cao, Z., Dose, M., Casellas, R., Jankovic, M., et al. (2017). The cell cycle restricts activation-induced cytidine deaminase activity to early G1. J. Exp. Med. 214, 49– 139

58.

Wang, X., Cho, B., Suzuki, K., Xu, Y., Green, J.A., An, J., and Cyster, J.G. (2011). Follicular dendritic cells help establish follicle identity and promote B cell retention in germinal centers. J. Exp. Med.

Wang, Z., Raifu, M., Howard, M., Smith, L., Hansen, D., Goldsby, R., and Ratner, D. (2000). Universal PCR amplification of mouse immunoglobulin gene variable regions: the design of degenerate primers and an assessment of the effect of DNA polymerase 3’ to 5’ exonuclease activity. J. Immunol. Methods 233, 167– 177.

Watanabe, M., Fujihara, C., Radtke, A.J., Chiang, Y.J., Bhatia, S., Germain, R.N., and Hodes, R.J. (2017). Co-stimulatory function in primary germinal center responses: CD40 and B7 are required on distinct antigen-presenting cells. J. Exp. Med.

Weinstein, J.S., Herman, E.I., Lainez, B., Licona-Limón, P., Esplugues, E., Flavell, R., and Craft, J. (2016). TFH cells progressively differentiate to regulate the germinal center response. Nat. Immunol. 17, ni.3554.

Weisel, F.J., Zuccarino-Catania, G.V., Chikina, M., and Shlomchik, M.J. (2016). A Temporal Switch in the Germinal Center Determines Differential Output of Memory B and Plasma Cells. Immunity 44, 116–130.

Wherry, E.J., Blattman, J.N., Murali-krishna, K., Most, R. Van Der, and Ahmed, R. (2003a). Viral Persistence Alters CD8 T-Cell Immunodominance and Tissue Distribution and Results in Distinct Stages of Functional Impairment Viral Persistence Alters CD8 T-Cell Immunodominance and Tissue Distribution and Results in Distinct Stages of Functional Im.

Wherry, E.J., Blattman, J.N., Murali-Krishna, K., van der Most, R., and Ahmed, R. (2003b). Viral Persistence Alters CD8 T-Cell Immunodominance and Tissue Distribution and Results in Distinct Stages of Functional Impairment. J. Virol.

Wickham, H. (2017). tidyverse: Easily Install and Load the “Tidyverse.”

Wickham, H., Chang, W., Henry, L., Pedersen, T.L., Takahashi, K., Wilke, C., 140 and Woo, K. (2019). ggplot2: Create Elegant Data Visualisations Using the Grammar of Graphics.

Wiedmann, M., Liebert, U.G., Oesen, U., Porst, H., Wiese, M., Schroeder, S., Halm, U., Mössner, J., and Berr, F. (2000). Decreased immunogenicity of recombinant hepatitis B vaccine in chronic hepatitis C. Hepatology 31, 230–234.

Wieland, A., Shashidharamurthy, R., Kamphorst, A.O., Han, J.H., Aubert, R.D., Choudhury, B.P., Stowell, S.R., Lee, J., Punkosdy, G.A., Shlomchik, M.J., et al. (2015). Antibody effector functions mediated by Fcγ-receptors are compromised during persistent viral infection. Immunity.

Wilhelm, K., Happel, K., Eelen, G., Schoors, S., Oellerich, M.F., Lim, R., Zimmermann, B., Aspalter, I.M., Franco, C.A., Boettger, T., et al. (2016). FOXO1 couples metabolic activity and growth state in the vascular endothelium. Nature 529, 216–220.

Wilson, E.B., Yamada, D.H., Elsaesser, H., Herskovitz, J., Deng, J., Cheng, G., Aronow, B.J., Karp, C.L., and Brooks, D.G. (2013a). Blockade of chronic type I interferon signaling to control persistent LCMV infection. Science 340, 202–207.

Wilson, P.C., Esposito, S., Hensley, S.E., Linderman, S.L., Madara, J., Wrammert, J., Carter, D.M., Principi, N., Liu, Q., Plotkin, J.B., et al. (2013b). Immune history shapes specificity of pandemic H1N1 influenza antibody responses. J. Exp. Med.

Wu, T., Ji, Y., Ashley Moseman, E., Xu, H.C., Manglani, M., Kirby, M., Anderson, S.M., Handon, R., Kenyon, E., Elkahloun, A., et al. (2016). The TCF1-Bcl6 axis counteracts type I interferon to repress exhaustion and maintain T cell stemness. Sci. Immunol.

Wu, X., Zhang, Z., Schramm, C.A., Joyce, M.G., Kwon, Y. Do, Zhou, T., Sheng, Z., Zhang, B., O’Dell, S., McKee, K., et al. (2015). Maturation and Diversity of the VRC01-Antibody Lineage over 15 Years of Chronic HIV-1 Infection. Cell 161, 470–485.

Xu, H., Chaudhri, V.K., Wu, Z., Biliouris, K., Dienger-Stambaugh, K., Rochman, Y., and Singh, H. (2015). Regulation of bifurcating B cell trajectories by mutual antagonism between transcription factors IRF4 and IRF8. Nat. Immunol. 141

Yaari, G., Uduman, M., and Kleinstein, S.H. (2012). Quantifying selection in high- throughput Immunoglobulin sequencing data sets. Nucleic Acids Res. 40, e134.

Yang, C.H., Murti, A., Pfeffer, S.R., Kim, J.G., Donner, D.B., and Pfeffer, L.M. (2001). Interferon α/β Promotes Cell Survival by Activating Nuclear Factor κB through Phosphatidylinositol 3-Kinase and Akt. J. Biol. Chem.

Yang, Z., Robinson, M.J., Chen, X., Smith, G.A., Taunton, J., Liu, W., and Allen, C.D.C. (2016). Regulation of B cell fate by chronic activity of the IgE B cell receptor. Elife.

Yarkoni, Y., Getahun, A., and Cambier, J.C. (2010). Molecular underpinning of B- cell anergy. Immunol. Rev.

Yusuf, I., Kageyama, R., Monticelli, L., Johnston, R.J., DiToro, D., Hansen, K., Barnett, B., and Crotty, S. (2010). Germinal Center T Follicular Helper Cell IL-4 Production Is Dependent on Signaling Lymphocytic Activation Molecule Receptor (CD150). J. Immunol. 185, 190–202.

Zhou, Y., Zhou, B., Pache, L., Chang, M., Khodabakhshi, A.H., Tanaseichuk, O., Benner, C., and Chanda, S.K. (2019). Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat. Commun. 10, 1523.

Zost, S.J., Wu, N.C., Hensley, S.E., and Wilson, I.A. (2019). Immunodominance and Antigenic Variation of Influenza Virus Hemagglutinin: Implications for Design of Universal Vaccine Immunogens. J. Infect. Dis. 219, S38–S45.

Zotos, D., Coquet, J.M., Zhang, Y., Light, A., D’Costa, K., Kallies, A., Corcoran, L.M., Godfrey, D.I., Toellner, K.-M., Smyth, M.J., et al. (2010). IL-21 regulates germinal center B cell differentiation and proliferation through a B cell–intrinsic mechanism. J. Exp. Med. 207, 365–378.

Zuniga, E.I., Macal, M., Lewis, G.M., and Harker, J.A. (2015). Innate and Adaptive Immune Regulation During Chronic Viral Infections. Annu. Rev. Virol.

142