UC San Diego UC San Diego Electronic Theses and Dissertations

Title Defining the Role of Nr4a Transcription Factors in CD8+ T Cell Exhaustion Using a Murine CAR T Cell Tumor Model

Permalink https://escholarship.org/uc/item/5mj1t7tp

Author Chen, Joyce

Publication Date 2019

Supplemental Material https://escholarship.org/uc/item/5mj1t7tp#supplemental

Peer reviewed|Thesis/dissertation

eScholarship.org Powered by the California Digital Library University of California UNIVERSITY OF CALIFORNIA SAN DIEGO

Defining the Role of Nr4a Transcription Factors in CD8+ T Cell Exhaustion Using a Murine CAR T Cell Tumor Model

A dissertation submitted in partial satisfaction of the requirements for the degree Doctor of Philosophy

Biomedical Sciences

Joyce Chen

Committee in charge:

Professor Anjana Rao, Chair Professor Victor Nizet, Co-Chair Professor Rafael Bejar Professor Christopher K. Glass Professor Ananda W. Goldrath

2019

Joyce Chen, 2019

ii The Dissertation of Joyce Chen is approved, and is acceptable in quality and form for publication on microfilm and electronically:

Co-Chair

Chair

University of California San Diego

2019

iii EPIGRAPH

The most difficult thing is the decision to act. The rest is merely tenacity. The fears are paper tigers. You can do anything you decide to do. You can act to change and control your life and the procedure. The process is its own reward.

Amelia Earhart

iv TABLE OF CONTENTS

Signature Page ...... iii

Epigraph ...... iv

Table of Contents ...... v

List of Abbreviations ...... vii

List of Supplemental Files ...... viii

List of Figures ...... ix

Acknowledgements ...... xi

Vita ...... xiv

Abstract of the Dissertation ...... xv

Chapter 1: Introduction ...... 1

1.1 Cancer Immunotherapy: Checkpoint Blockade Therapies...... 2 1.2 Cancer Immunotherapy: Chimeric Antigen Receptor T Cell Therapies ...... 3 1.3 T Cell Exhaustion ...... 7 1.4 Scope of this Work...... 12 1.5 Conclusion and Future Directions ...... 13 1.6 Figures ...... 14 1.7 References ...... 17

Chapter 2: CD8+ CAR Tumor-Infiltrating T Cells Exhibit Similarity to CD8+ Endogenous Tumor- Infiltrating T Cells ...... 22 Abstract ...... 23 2.1 Introduction ...... 24 2.2 Results and Discussion ...... 26 2.3 Conclusion ...... 31 2.4 Figures ...... 32 2.5 Experimental Section ...... 45 2.6 Acknowledgements...... 56 2.7 References ...... 57

Chapter 3: Nr4a Transcription Factors Limit CAR T Cell Function in Solid Tumors ...... 63

v Abstract ...... 64 3.1 Introduction ...... 65 3.2 Results and Discussion ...... 67 3.3 Conclusion ...... 73 3.4 Figures ...... 75 3.5 Experimental Section ...... 93 3.6 Acknowledgements...... 106 3.7 References ...... 107

vi LIST OF ABBREVIATIONS

ATAC-seq assay for transposase-accessible chromatin using sequencing

CAR chimeric antigen receptor hCD19 human CD19 antigen

HCV hepatitis C virus

HIV human immunodeficiency virus

IFNg interferon gamma

IL-2 interleukin 2

LCMV lymphocytic choriomeningitis virus

NFAT nuclear factor of activated T cells

Nr4a nuclear receptor subfamily 4 group A

PD-1 programmed cell death 1

RNA-seq RNA-sequencing

TCR t cell receptor

TIM3 T cell immunoglobulin and mucin-containing domain-3

TILs tumor-infiltrating lymphocytes

TNF tumor necrosis factor

SIV simian immunodeficiency virus

vii LIST OF SUPPLEMENTAL FILES

Supplementary Table 1 Comparisons of gene expression program in CAR vs endogenous CD8+ T cells infiltrating B16-OVA-hCD19 tumor

Supplementary Table 2 Comparisons of human CD8+ TILs infiltrating a human melanoma

Supplementary Table 3 Comparisons of cells ectopically expressing Nr4a in CD8+ T cells in vitro. a) RNA-seq DESeq2 results comparing empty vector pMIN vs Nr4a1; b) RNA-seq DESeq2 results comparing empty vector pMIN vs Nr4a2; c) RNA-seq DESeq2 results comparing empty vector pMIN vs Nr4a3.

Supplementary Table 4 Comparison of gene expression program in Nr4a TKO vs wild-type CD8+ CAR T cells infiltrating B16-OVA-hCD19 tumor

Supplementary Table 5 Comparison of gene expression program in antigen-specific cells from LCMV-infected mice treated with anti-PDL1 or IgG control

viii LIST OF FIGURES Chapter 1

Figure 1.1. Schematic of checkpoint blockade ...... 14

Figure 1.2. Schematic of chimeric antigen receptor design ...... 15

Figure 1.3. Schematic of T cell response ...... 16

Chapter 2

Figure 2.1. Functional assessment of a human CD19-reactive chimeric antigen Receptor (CAR) ...... 32

Figure 2.2. CAR, OT-I and endogenous CD8+ TILs isolated from B16-OVA-hCD19 tumors exhibit similar phenotypes ...... 34

Figure 2.3. Tumor-bearing mice adoptively transferred with CAR and OT-I T cells exhibit similar tumor growth rates ...... 35

Figure 2.4. CAR and endogenous CD8+ TILs exhibit similar gene expression and chromatin accessibility profiles ...... 36

Figure 2.5. Adoptively transferred CD8+ CAR T cells infiltrating B16-OVA-hCD19 tumors exhibit gene expression profiles similar to those of endogenous CD8+ TILs (extended data for Figure 2.4) ...... 37

Figure 2.6. Adoptively transferred CAR-expressing mouse CD8+ T cells infiltrating B16-OVA-hCD19 tumors exhibit chromatin accessibility profiles similar to those of endogenous CD8+ TILs (extended data for Figure 2.4) ...... 38

Figure 2.7. CAR and endogenous CD8+ PD-1highTIM3high TILs exhibit upregulation of Nr4a family members ...... 40

Figure 2.8. Mouse CD8+ TILs exhibit increased expression of Nr4a1, Nr4a2, Nr4a3 (extended data for Figure 2.7) ...... 41

Figure 2.9. Human CD8+ TILs exhibit increased Nr4a family members in PD-1hi TIM3hi TILS ...... 43

Figure 2.10. Human CD8+ PD-1hi TILs and antigen-specific T cells from HIV patients show similar chromatin accessibility to mouse TILs ...... 44

Chapter 3

Figure 3.1. Nr4a-deficient CAR TILs promote tumor regression and prolong survival...... 75

ix Figure 3.2. Preparation of wild-type and Nr4a TKO CAR T cells for adoptive transfer ...... 76

Figure 3.3. Prolonged survival of immunocompetent tumor-bearing mice adoptively transferred with CD8+ Nr4a TKO compared to WT CAR T cells ...... 78

Figure 3.5. Phenotypic features of mouse CD8+ T cells expressing Nr4a1, Nr4a2 or Nr4a3 ...... 81

Figure 3.6. Genomic features of mouse CD8+ T cells expressing Nr4a1, Nr4a2 or Nr4a3 ...... 83

Figure 3.7. Nr4a-deficient CAR TILs show rescued effector function ...... 84

Figure 3.8. CD8+ Nr4a TKO CAR TILs show increased effector function compared to WT CAR TILs… ...... 85

Figure 3.9. Gene expression profiles indicate increased effector function of Nr4a TKO compared to WT CAR TILs ...... 86

Figure 3.10. Gene set enrichment analyses ...... 87

Figure 3.11. Gene expression profiles indicate increased effector function of Nr4a TKO compared to WT CAR TILs ...... 88

Figure 3.12. Nr4a family members bind to predicted Nr4a binding motifs that are more accessible in the WT CAR TILs compared to the Nr4a TKO CAR TILs, and regions more accessible in WT compared to Nr4a TKO CAR TILs are more accessible in CA-RIT-NFAT1- and Nr4a1/2/3-transduced cells ...... 89

Figure 3.13. Nr4a family members show a moderate decrease in mRNA expression in antigen-specific cells from LCMV-infected mice treated with anti-PDL1 or IgG control...... 91

Figure 3.14. Proposed role of Nr4a in T cells chronically stimulated with antigen...... 92

x ACKNOWLEDGEMENTS

I am immensely grateful of those without whom this work would not be possible:

Dr. Anjana Rao, my mentor, whom I admire for refusing to believe in limitations and for believing in science in its purest form unaffected by politics and the mundane demands of everyday life. I thank her for her undying support, for the resources and scientific freedom that she has provided me, and for pushing me to become a better scientist every step of the way. I could not have asked for a better mentor for my PhD.

Isaac F. López-Moyado, bioinformatics extraordinaire, my collaborator and lab confidante, for his careful and thorough analysis of all the RNA-seq data generated in this work, as well as his analyses of publicly available datasets, but most importantly, for being an amazing friend. I truly would not have gotten through graduate school without his support and friendship.

Dr. James P. Scott-Browne, my collaborator and our former resident immunologist and expert on all things that derive from a hematopoietic stem cell, for his invaluable expertise that was instrumental in getting this project off the ground, and for his careful and thorough analysis of all of the ATAC-seq data generated in this work as well as his analyses of publicly available datasets.

Dr. Chan-Wang (Jerry) Lio, for being my go-to guy for questions on absolutely anything – ranging from how to do chromatin immunoprecipitation to how to get back up after life knocks you down. His words of wisdom both on things in the lab and out of the lab have been invaluable. I would also like to thank Jerry for always being armed with snacks. There were definitely a few late nights where I ran out of hot Cheetos and my dinner ended up being whatever snacks Jerry had on hand.

Dr. Takashi Sekiya and Dr. Akihiko Yoshimura, my collaborators in Japan, for their advice and for their generosity in sharing their Nr4a1 fl/fl Nr4a2 fl/fl Nr4a3-/- mice.

xi Dr. Hyungseok Seo, for lending a much-needed hand or both hands, rather, with our revision experiments and for carrying on the experiments that follow this body of work.

Laura J. Hempleman, for showing up to every single mouse harvest at 6:00 am with me and helped me take tumors out of mice, blend them, and strain the CD8+ tumor-infiltrating lymphocytes out of the mush. She always had great stories and they made the mornings go by so much faster.

Samantha Blake, for helping me with all things administrative, for her unyielding emotional support and for introducing me to the soothing effect of 3-wick candles. Sam got me through some of the really tough days in the final stretch of my graduate school career.

Dr. Anand Balasubramani and Dr. Yun (Nancy) Huang, for mentoring me during my rotation and for giving me my start in Anjana’s lab. My thesis work turned out to be very different from my rotation project, but it would not have been possible without them taking me in early on.

My thesis dissertation committee, Dr. Victor Nizet, Dr. Rafael Bejar, Dr. Christopher K.

Glass, and Dr. Ananda W. Goldrath, for their time, support, and scientific and career advice.

The LJI Flow Cytometry Core, Cheryl Kim, Denise Hinz, Robin Simmons, Lara Nosworthy,

Chris Dillingham, for all of the cell sorts. I would especially like to thank Denise for staying after hours when the cell sorts ran late.

The LJI Next-Generation Sequencing Core, Jeremy Day, Steven (Nick) Wlodychak, and

Christina Kim, for performing all of the next-generation sequencing presented in this work. I would especially like to thank Jeremy Day for helping me prepare my first RNA-seq library and for giving me advice on all things sequencing.

The Department of Laboratory Animal CAR (DLAC) and the animal facility, for all their efforts towards making this possible – Juan (Fernando) Vasquez, for snipping the tails to genotype hundreds of Nr4a TKO mice; Kristine Suchey and Mindy Hockaday for working with me to keep my mice healthy; Morag Mackay for running the best animal facility I have ever worked in.

xii Dr. Alejandro B. Balazs, whose support played a large role in getting me to where I am today, for believing in me when I didn’t believe in myself, and for reminding me that getting to do science is “living the dream.”

The UCSD Pharmacology Department, for the financial support through the T32 training grant and for giving me an additional forum where I can learn and grow as a graduate student; and the PhRMA Foundation for the financial support through the Paul Calabresi Medical Student

Research Fellowship.

Finally, I would like to thank my close friends and family without whom this journey would not be worthwhile. My parents, for humoring what seems to be my desire to be a student forever.

My brother, Kid, for his dry sense of humor and for indulging me in existential conversations. The following friends for their support, and especially for being a light during the darkest times of my

PhD: Christine Y. Park, for her unwavering emotional support from 8 or 9 time zones away and for bringing me perspective; Dr. Fides D. Lay, for being my voice of reason and for keeping me grounded; Dr. Linh Y. Truong, for her optimism and for always believing in me – I strive to be the person that she believes me to be; Eulanca Y. Liu, for weathering the storms of the PhD with me and telling me to add oil.

Chapters 2 and 3, in full, are reprints with modifications as it appears in NR4A transcription factors limit CAR T cell function in solid tumours, Nature (2019), published online Feb 27, 2019. doi: 10.1038/s41586-019-0985-x. The dissertation author was the primary investigator and author of this paper. Other authors include Isaac F. López-Moyado, Hyungseok Seo, Chan-Wang J. Lio,

Laura J. Hempleman, Takashi Sekiya, Akihiko Yoshimura, James P. Scott-Browne & Anjana Rao.

xiii VITA

2008 Bachelor of Science, University of California Los Angeles 2009-2012 Research Associate, California Institute of Technology 2012- Medical Scientist Training Program, University of California San Diego 2014-2019 Doctor of Philosophy, University of California San Diego

PUBLICATIONS

1. Chen, J.*, López-Moyado, I.F., Seo, H., Lio, C.-W. J., Hempleman, L.J., Sekiya, T., Yoshimura, A., Scott-Browne J.P. †*, Rao, A†*. NR4A transcription factors limit CAR T cell function in solid tumors. Nature (2019), published online Feb 27, 2019. doi: 10.1038/s41586-019-0985-x. † these authors contributed equally; * co-corresponding authors

2. Seo H., Chen J., Gonzalez-Avalos E., Samaniego-Castruita D., López-Moyado I.F., Georges R., Zhang W., Wu C.-J., Das A., Liu L.-F., Bhandoola A. and Rao A. NR4A and TOX transcription factors cooperate to impose CD8+ T cell exhaustion. Submitted, in review.

xiv

ABTRACT OF THE DISSERTATION

Defining the Role of Nr4a Transcription Factors in CD8+ T Cell Exhaustion Using a Murine CAR T Cell Tumor Model

Joyce Chen

Doctor of Philosophy in Biomedical Sciences

University of California San Diego, 2019

Professor Anjana Rao, Chair Professor Victor Nizet, Co-Chair

In cancer immunotherapy, T cells expressing chimeric antigen receptors (CAR T) targeting human CD19 (hCD19) antigen have exhibited impressive clinical efficacy against B cell leukemias and lymphomas. However, CAR T cells have been less effective against solid tumors, in part because they enter a hyporesponsive (“exhausted” or “dysfunctional”) state that is triggered by chronic antigen stimulation and characterized by upregulation of several inhibitory receptors and loss of effector function. To identify transcriptional and other regulators contributing to the diminished function of CAR T cells in solid tumors, we developed a mouse model in which recipient mice bearing murine melanoma tumors expressing hCD19 antigen were adoptively transferred with hCD19-reactive CAR T cells. We show that CD8+ CAR-expressing tumor-

xv infiltrating lymphocytes (TILs) and endogenous TILs with low effector function and high expression of inhibitory surface receptors PD-1 and TIM3 exhibit similar profiles of gene expression and chromatin accessibility, associated with secondary activation of the three Nr4a

(Nuclear Receptor Subfamily 4 Group A) transcription factors Nr4a1 (Nur77), Nr4a2 (Nurr1) and

Nr4a3 (Nor1) by the initiating transcription factor NFAT (nuclear factor of activated T cells).

Through a similar comparison of data from human CD8+ TILs and viral antigen specific CD8+ T cells from humans chronically infected with HIV, we observed high expression of Nr4a transcription factors and enrichment of Nr4a binding motifs in uniquely accessible chromatin regions. Treatment of tumor-bearing mice with CAR T cells lacking all three Nr4a transcription factors (Nr4a TKO) resulted in tumor regression and prolonged survival compared to mice adoptively transferred with Nr4a-sufficient (WT) CAR T cells. Nr4a TKO CAR TILs displayed phenotypes and gene expression profiles characteristic of CD8+ effector T cells, and chromatin regions uniquely accessible in Nr4a TKO CAR TILs compared to wild type were enriched for binding motifs for NFkB and AP-1, transcription factors classically involved in T cell activation.

Our data identify Nr4a transcription factors as major players in the cell-intrinsic program of T cell hyporesponsiveness and point to Nr4a inhibition as a promising strategy for cancer immunotherapy.

xvi Chapter 1:

Introduction

1 Clinical successes in cancer immunotherapy have advanced greatly due to two fields: checkpoint blockade therapies that modulate the immune response, and chimeric antigen receptor (CAR) T cell therapy in which the T cells are targeted to recognize specific tumor antigens. Checkpoint blockade will be discussed briefly in the first section below, and CAR T cells will be discussed in somewhat more detail in the second section, as they are integral to the model system I established for my thesis work. Checkpoint blockade is designed to counter a phenomenon known as CD8+ T cell exhaustion or dysfunction that is discussed in detail in the third section; it is pertinent to CAR T cells since these cells, like conventional antigen-reactive T cells, become hyporesponsive to stimulation by tumor antigens after prolonged residence in tumors.

1.1 Cancer Immunotherapy: Checkpoint Blockade Therapies

The immune system naturally has regulatory mechanisms built in to dampen unwanted immune responses, such as autoimmunity. PD-1 and CTLA-4 are two cell surface receptors that are expressed in activated T cells and act in a cell-intrinsic negative feedback pathway to modulate or inhibit the immune response (Ribas and Wolchok, 2018; Sharma and Allison, 2015;

Topalian et al., 2015; Wei et al., 2018). Briefly, CTLA-4 decreases T cell receptor (TCR) signaling by competing with co-stimulatory receptor CD28 for its ligands on the antigen- presenting cell, resulting in a negative co-stimulatory signal that attenuates T cell activation. PD-

1 binds directly to its ligands PD-L1 or PD-L2, usually on the tumor cell, and this results in negative downstream signaling to attenuate T cell activation via the SHP2 tyrosine phosphatase. Both these inhibitory surface receptors enforce a program by which T cells first become activated, then turn on negative signaling and transcriptional pathways that impose a program of hyporesponsiveness (also known as “exhaustion or “dysfunction”) and serve to attenuate and prevent a hyperactive immune response. Immune checkpoint blockade therapies target these inhibitory signaling pathways through the use of antibodies against CTLA-4 and

2 PD-1/PD-L1 (Figure 1.1), thus allowing tumor-reactive T cells to mount a more effective immune response in a certain subset of patients.

We have seen some initial successes with checkpoint blockade, but still, a large majority of patients do not respond to checkpoint blockade therapy, and the efficacy varies across tumor types. Ipilimumab (against CTLA-4) has only shown long-term survival in 20% of metastatic melanoma patients (Hodi et al., 2010). It was shown that in previously untreated melanoma patients, combination therapy against PD-1 and CTLA-4 allowed for longer progression-free survival than that of monotherapy with either PD-1 or CTLA-4, with roughly 36% of patients showing long-term survival after combination therapy (Larkin et al., 2015).

Studies investigating the effects of anti-CTLA-4 and PD-1 blockade on exhausted T cells have shown that these treatments do not exhibit a generalized effect, but rather impact certain subsets of T cells more than others. In the case of PD-1 blockade, several studies have shown that CXCR5+ cells expand after PD-1 blockade (He et al., 2016; Im et al., 2016; Leong et al.,

2016), and more recent studies further confirm TCF1 as a requirement for this proliferative ability (Im et al., 2016; Siddiqui et al., 2019); these studies will be further discussed in a later section. In another recent study, anti-CTLA-4 was described to cause an increase in the ICOS+

Th1-like CD4+ T cells (Wei et al., 2017). The clinical successes of combination therapies over that of monotherapies cannot be denied, and investigations of the mechanisms resulting in the improved efficacy of combination therapies are currently underway.

1.2 Cancer Immunotherapy: Chimeric Antigen Receptor T cells

Chimeric antigen receptor (CAR) T cells have exhibited impressive clinical efficiency in bolstering the effector or anti-tumor response (June and Sadelain, 2018; Sadelain, 2017).

Chimeric antigen receptors consist of an extracellular antigen-recognition domain (usually an antibody single-chain variable fragment) and intracellular signaling domains that mimic T cell receptor signaling, with a hinge and/or transmembrane domain linking the extracellular and

3 intracellular portions (Fesnak et al., 2016; Jackson et al., 2016). In the clinic, the patient’s T cells are isolated and then activated and genetically modified ex vivo to express the CAR. These

CAR T cells are then infused back into the patient.

Since its inception, the design of the signaling domains of the CAR has undergone several improvements, resulting in what we now refer to as first-generation, second-generation, and third-generation CARs (Jackson et al., 2016) (Figure 1.2). In all three generations of the

CAR, the design of the extracellular portion of the CAR remains the same; it allows an MHC- independent recognition of the desired antigen by the T cell. The first-generation CARs only contained the signaling domain to mimic “signal 1” for T cell activation, usually the zeta chain of

CD3 (CD3z). These cells showed limited efficacy in clinical trials (Jackson et al., 2016). In second-generation CARs, the intracellular signaling domains contain both the “signal 1” T cell receptor stimulation (CD3z) and “signal 2” co-stimulus (CD28, 4-1BB/CD137 for example), resulting in a simultaneous triggering of signals 1 and 2 that leads to the signaling cascade inducing T cell activation and the mount of an optimal effector response: proliferation, cytokine production, and lysis of target cells. Second-generation CARs have shown improved efficacy over that of first-generation CARs (Jackson et al., 2016). Studies have shown that second- generation CARs with co-stimulatory domain 4-1BB while less effective initially, allow for better persistence than those with co-stimulatory domain CD28, but may vary for different tumors.

Both FDA approved CAR T cell therapies are second-generation CARs; Kymriah

(Tisagenlecleucel), approved in August 2017 for the treatment of B-cell precursor acute lymphoblastic leukemia (ALL) in patients 25 years of age or younger, is a second-generation

CAR containing the intracellular signaling domains CD3z and 4-1BB (CD137). Yescarta

(Axicabtagene ciloleucel), approved a couple of months after, in October 2017, for the treatment of diffuse large B-cell lymphoma (DLBCL) in adults, is a second-generation CAR containing the intracellular signaling domains CD3z and CD28. Third-generation CARs consist of two co-

4 stimulatory domains, and have shown increased tumor efficacy in preclinical studies (Jackson et al., 2016). Currently, there is one clinical trial in the United States recruiting patients for the comparison of second-generation CAR with co-stimulatory domain CD28 and third-generation

CAR with both CD28 and 4-1BB co-stimulatory domains (Jackson et al., 2016). The study is designed such that patients receive a 1:1 ratio of second- to third- generation CARs.

As evidenced by the recent FDA approvals of CAR T cell therapies for leukemias and lymphomas and numerous clinical trials, CAR T cells have been remarkably effective in the treatment of B cell malignancies (Davila et al., 2014; Maude et al., 2014). However, CAR T cells have seen limited success in solid tumors, in part because of the difficulties faced in the availability and heterogeneity of solid tumor antigen targets, homing and infiltration of certain tumor types, the immunosuppression of the local tumor microenvironment, and cell-intrinsic T cell exhaustion (Moon et al., 2014; Yong et al., 2017). Despite these difficulties, studies in the past few years have begun to chip away at the daunting task of using CARs to target solid tumors. In patients with HER2-positive tumors that were either metastatic or refractory to standard therapies, HER2 CAR T cells did stabilize patients’ disease for 6 weeks, but did not result in a complete response (Ahmed et al., 2015); the median overall survival of all enrolled patients was 10.3 months. In patients with refractory and multifocal glioblastoma expressing

EGFR variant III (EGFRvIII), EGFRvIII-directed CAR T cells show no cross-reactivity with the wild-type EGFR. The EGFRvIII CAR T cells are able to cross the blood-brain barrier and stabilize the disease for a median overall survival of 251 days (O’Rourke et al., 2017). More recently, NEJM published the report of a multifocal glioblastoma patient with metastatic spine tumors that was treated with IL13R-alpha targeted CAR T cells (Brown et al., 2016). This patient showed a dramatic complete response that was sustained for 7.5 months before tumors surfaced in new locations.

As we and several other groups have shown, CAR T cells also undergo T cell exhaustion (Cherkassky et al., 2016; Long et al., 2015; Moon et al., 2014). Hence, it may be

5 expected that combination therapy using CAR T cells and checkpoint blockade therapies in solid tumors will show efficacy over the use of CAR T cells alone in solid tumors. Cherkassky and colleagues (Cherkassky et al., 2016) show that in an orthotopic mouse model of malignant pleural mesothelioma, that a low dose of mesothelin-targeted second-generation CAR T cells is insufficient to clear the tumor, but can be reinvigorated by anti-PD-1 treatment. Separately, using an orthotopic renal cell carcinoma model, Suarez and colleagues (Suarez et al., 2014) sought to combine checkpoint blockade and CAR T cells in a different fashion – designing a carbonic-anhydrase-targeting CAR to secrete anti-PDL1 antibodies at the tumor site by inserting the anti-PDL1 cassette after the internal ribosome entry site (IRES) following the CAR. This study shows that in comparison to control, the CAR tumor-infiltrating lymphocytes (TILs) secreting anti-PDL1 exhibited lower expression of PD-1, TIM3, LAG3, and higher expression of granzyme B. The CAR TILs secreting anti-PDL1 also exhibit significantly better anti-tumor effects. The combination of PD-1 blockade and CAR T cells is only one of the options to decrease CAR T cell exhaustion. These successes suggest that if we can identify and find regulators of T cell exhaustion, we can modify CAR T cells to become “less exhausted”.

Last year, there was an exciting report (Fraietta et al., 2018) of a patient being treated with CD19-targeted CAR for his chronic lymphocytic leukemia; the patient experienced a delayed therapeutic response and it was found that 94% of the CAR T cell population in his blood was comprised of a single clone in which the CAR had integrated into one allele of the

TET2 gene. The other TET2 allele already contained a mutation, and the integration of the CAR into the wild-type allele gave the CAR T cell the phenotype of a TET2 knockout cell – massive clonal expansion. As TET2 is a tumor suppressor gene, the patient continues to be carefully monitored. However, he did achieve a complete remission and after the peak of the therapeutic response, the clonal CAR T cell population did contract. The report serves as an example showing that genetic modifications of CAR T cells may be a promising approach to improving the efficacy and persistence of CAR T cells.

6 Finally, development of new strategies tackle the problem of T cell exhaustion caused by chronic stimulation by limiting the interaction of the CAR to its target. Eyquem and colleagues

(Eyquem et al., 2017) used gRNAs to knock the CAR into the TCR-alpha locus, subjecting the

CAR to the same internalization that an endogenous TCR is subject to; they show that having the regulated CAR results in less exhausted T cells. Studies from the Wendell Lim group

(Morsut et al., 2016; Roybal et al., 2016) show a synNotch system in which presentation of one antigen leads to expression of the receptor targeting a second antigen. This dual-antigen presentation system allows for more precise tumor targeting.

1.3 T Cell Exhaustion

The advances in cancer immunotherapy described above would not have been possible without the discoveries and continued work from the fields of T cell function and T cell exhaustion. Highlights and recent works from the T cell exhaustion field are considered below.

In a normal immune response, naïve CD8+ T cells are stimulated by antigen, costimulation, and inflammation, and they expand clonally and differentiate into effector T cells.

Following antigen clearance, the majority of the effector T cells die by apoptosis in the contraction phase, and a small remaining fraction mature into memory T cells. Memory T cells are able to mount robust recall responses upon secondary exposure to antigen (Kaech and Cui,

2012). In cases of cancer or chronic infection, the antigen persists beyond the effector phase, and the T cells undergo chronic stimulation (Figure 1.3, left). Over time, T cells present with “T cell exhaustion,” a hyporesponsive state (Wherry, 2011; Wherry and Kurachi, 2015) characterized by sustained upregulation of multiple inhibitory surface receptors, decrease in effector function (loss of cytokine production and cytolytic function), altered use and/or expression of key transcription factors, metabolic derangement, and ultimately, an inability to form functional memory T cells (McLane et al., 2019; Wherry, 2011; Wherry and Kurachi, 2015).

T cell exhaustion was first described in CD8+ T cells in mice with chronic LCMV

7 infection, as virus-specific CD8+ T cells that persisted but were unable to exact effector function and hence could not control the viral infection (Moskophidis et al., 1993; Zajac et al., 1998).

CD8+ T cells appeared to lose function in a hierarchical manner (Wherry et al., 2003) (Figure

1.3, right); in the early stages of exhaustion, CD8+ T cells first lose the ability to produce IL-2 and exert cytotoxic effects, and then progressively lose the ability to produce TNF, and finally,

IFNg. Co-expression of other multiple inhibitory receptors is characteristic of exhausted CD8+ T cells; in addition to PD-1, these include TIM3, LAG3, CTLA-4, CD244, TIGIT, amongst others.

Gene expression profiles of exhausted CD8+ T cells were first identified through microarrays

(Wherry et al., 2007), showing that exhausted T cells acquire a distinct transcriptional state separate from that of memory or effector T cells.

Prior to the undertaking of this thesis work, one of the big questions in the field of T cell exhaustion is whether the exhausted T cell is a terminally differentiated state primarily resulting from the failure to achieve memory (as a linear progression model), or a reversible state that still retains the ability to become an effector or memory T cell (a divergent model). Studies published during the course of this thesis work seem to point towards a merging of the two theories, that the term “exhaustion” describes a heterogenous population of dysfunctional cells ranging from cells that can indeed be reversed to regain function and cells that seemingly cannot. Exhausted

T cells are potentially reversible to effector or memory T cells at earlier stages of chronic antigen stimulation, but not in later stages of dysfunction. Schietinger and colleagues (Schietinger et al.,

2016) show that anti-PD1 treatment can only recover function in T cells that have been exposed to tumor antigen for 8 days rather than for 35 days, distinguishing an “early dysfunctional stage” which can be recovered, compared to a “late dysfunctional state” from which T cell functions cannot be recovered.

Many transcription factors have been shown to have important roles in CD8+ T cell differentiation during infection (Chang et al., 2014), and logically, many in the field have attempted to identify the transcription factors responsible for the exhausted state, or

8 transcription factors that are required to help the cell “move past” the exhausted state after therapeutic intervention. The advent of the assay for transposase-accessible chromatin using sequencing (ATAC-seq) (Buenrostro et al., 2013) has allowed us to look at changes in the chromatin level and investigate the idea that it is the condition of the epigenetic landscape permissive for T cell exhaustion rather than a single lineage-determining transcription factor.

Both approaches have been enlightening and not mutually exclusive, and it is likely a combination of transcriptional and epigenetic events that lead to T cell exhaustion.

No lineage-specific transcription factor for exhausted T cells has been identified; instead, several transcription factors that have important functions in physiological contexts or in other T cell subsets have been shown to exhibit exhaustion-specific functions. In some cases, this is due to an altered transcriptional program due to a change in transcription factor binding partners; Martinez and colleagues (Martinez et al., 2015) show that NFAT (nuclear factor of activated T cells), the initiating transcription factor for T cell activation, results in a hyporesponsive state in absence of its binding partner, AP-1. In other cases, it may be due to the concentration/availability of a transcription factor in the exhausted state. In the case of T-bet, it was shown that that T-bet repressed expression of PD-1 through weak binding to the upstream -23kb Pdcd1 enhancer, but antigen-specific cells in chronic LCMV infection exhibit lower levels of T-bet, leading the authors to hypothesize that T-bet binding in the weaker binding locations are not preserved when lower levels of T-bet are present (Kao et al., 2011).

The recent studies highlighting TCF1 as a requirement for response to checkpoint blockade is an excellent example of identifying transcription factors that are required to help move the cell out of the exhaustion state. TCF1 is required for the formation of central memory

T cells and required for the secondary expansion of the central memory T cells (Zhou et al.,

2010). Rafi Ahmed’s group showed that in LCMV-infected mice, after PD-1 blockade, the population of PD-1+ cells that expands expresses CXCR5 (a chemokine receptor that homes B and T cells to B cell follicles), and that in the absence of TCF1, the CXCR5+ population does not

9 develop (Im et al., 2016). By separate measures, both Leong and colleagues (Leong et al.,

2016) and He and colleagues (He et al., 2016) also showed that CXCR5+ LCMV-antigen- specific cells are less exhausted than CXCR5- cells, as determined by decreased expression of multiple inhibitory surface receptors. These studies (He et al., 2016; Im et al., 2016; Leong et al., 2016) suggest that PD-1+ antigen-specific T cells have a memory-precursor population that are more responsive to reinvigoration after PD-1 blockade. Recently published work (Siddiqui et al., 2019) establish the similar requirement of TCF1+ for antigen-specific (PD-1+) tumor infiltrating T cells to respond to checkpoint blockade therapy; they showed that selective ablation of Tcf7 in antigen-specific cells resulted in poorer tumor growth control compared to

Tcf7+/+ antigen-specific cells when both groups were treated with a combination of anti-PD-1 and anti-CTLA4. Additionally, mice transferred with cells lacking TCF1 had significantly fewer tumor-infiltrating cells of both TCF1+ or TCF1- expression, showing that TCF1+ cells are required for differentiation to TCF1- cells.

Prior to the advance of chromatin accessibility assays, a study (Youngblood et al., 2011) published in 2011 gave us a glimpse of how epigenetic changes may contribute to the regulation of T cell exhaustion. Youngblood and colleagues (Youngblood et al., 2011) showed that regulatory regions upstream of PD-1 that were conserved in both mouse and human were completely demethylated in exhausted T cells and were not remethylated even after a decrease in viral titer. In acute infections, these regions are demethylated in effector T cells and then remethylated in memory T cells. The study posited that lack of methylation at the Pdcd1 locus leaves it primed and ready for the expression of PD-1. More recent papers have explored the global epigenetic landscape of T cell exhaustion (Pauken et al., 2016; Philip et al., 2017a;

Scharer et al., 2017; Scott-Browne et al., 2016; Sen et al., 2016) through the use of ATAC-seq

(Buenrostro et al., 2013). Two papers (Scharer et al., 2017; Scott-Browne et al., 2016) compared chromatin accessibilities of naïve, effector, memory, and exhausted CD8+ T cells in

LCMV-infected mice and showed that compared to naïve, effector, and memory T cells,

10 exhausted T cells are most similar to effector T cells and they have a distinct profile of accessible regions that are enriched for certain transcription factor binding motifs. Philip and colleagues (Philip et al., 2017b) performed a temporal examination of chromatin states of tumor- infiltrating T cells and showed that successive chromatin remodeling occurs in waves at defined points throughout the T cells’ continued response to antigen-presenting tumor. These studies revealed that early on in the T cell response to cancer, the chromatin changes are more plastic and can be reversed, whereas in later stages the chromatin remodeling is more permanent.

Of note, these papers and others show an enhancer region roughly 23kb upstream of the PD-1 locus that is uniquely accessible to PD-1high cells. While not conserved in humans, the accessibility of this region in PD-1-expressing cells is consistent across all mouse models of dysfunction or exhaustion investigated so far. Sen and colleagues (Sen et al., 2016) show that deletion of this enhancer region results in only a mild decrease in PD-1 expression. Separately, two groups (Mognol et al., 2017; Pauken et al., 2016) showed that anti-PD-1 blockade does not change the chromatin accessibility profile of exhausted cells. Pauken and colleagues (Pauken et al., 2016) show that while PD-1 blockade in LCMV-infected mice seemed to revive function initially, the reinvigoration was not reflected at the chromatin level, and did not help to control viral load. Mognol and colleagues (Mognol et al., 2017) also reveal very little chromatin changes due to PD-1 blockade in tumor-bearing mice.

Finally, a study published recently identifying Runx3 as a key regulator of tissue-resident

+ memory (TRM) CD8 T cells (Milner et al., 2017) took a different approach; the authors noted previous studies identifying TILs that had tissue-resident memory characteristics correlated with better prognosis, and investigated the role of Runx3 in TILs. The authors showed that antigen- specific T cells lacking Runx3 showed decreased accumulation in tumors, whereas antigen- specific T cells overexpressing Runx3 accumulated in tumors and delayed tumor growth. Rather than attempting to reverse exhaustion, the authors posit that modifying the transcriptional program of TILs towards a tissue residency program would allow the TILs to take on relevant

11 and useful TRM characteristics (such as retention in tumors) to provide an enhanced TIL response.

1.4 Scope of this Work

We have taken the approach of identifying candidate transcription factors that may be involved in T cell exhaustion from genes that are found to be upregulated in exhausted T cells and binding motifs found to be uniquely accessible to exhausted T cells. We and others have found Nr4a to be upregulated in exhausted T cells and enrichment of Nr4a and NFAT binding motifs in regions uniquely accessible to exhausted T cells. We set up a CAR T cell model in mice and used it to explore the role of Nr4a transcription factors in CD8+ T cell exhaustion. We hypothesized that Nr4a was contributing to the exhaustion phenotype, and that the absence of

Nr4a transcription factors would result in a delay or reversal of the exhaustion phenotype.

The work presented in this thesis is centrally focused on the establishment of a murine

CAR T cell tumor model and the use of this model to define and functionally prove the contribution of the Nr4a transcription factors to CD8+ T cell exhaustion. In Chapter 2, we show that in this model, CD8+ CAR T cells exhibit similar gene expression and chromatin accessibility profiles to that of CD8+ endogenous T cells, and we identify Nr4a as a family of transcription factors upregulated in both endogenous exhausted T cells and exhausted CAR T cells. In

Chapter 3, we used the murine CAR T cell system to show functionally that deletion of Nr4a in

CD8+ CAR T cells allows for tumor regression in mice and prolonged survival. We found that the lack of Nr4a not only improves effector functions, but provides an epigenetic landscape more permissible to T cell activation by opening up binding sites accessible to activation-related transcription factors. This work suggests that Nr4a may be a promising target for cancer immunotherapy, as it would be expected to mimic the behavior of multiple checkpoint blockade inhibition.

12 1.5 Conclusion and Future Directions

In this study, we established a murine CAR T cell tumor model that has the benefit of introducing a guaranteed receptor-to-tumor-antigen interaction via expression of the CAR that bypasses the need to breed TCR-transgenic mouse strains to floxed or knockout strains. We showed that in this model, Nr4a transcription factors are upregulated in CAR T cells and endogenous T cells undergoing chronic antigen stimulation, and adoptive transfer of CAR T cells lacking all three Nr4a transcription factors into tumor-bearing mice results in a striking tumor regression and prolonged survival.

In the course of our studies on Nr4a, we found that Nr4a affects other transcription factors that may play a role in T cell exhaustion. Another transcription factor that was upregulated in exhausted CAR T cells was Tox, and interestingly, in CAR tumor-infiltrating lymphocytes lacking all three Nr4a transcription factors, Tox expression is also decreased.

Unpublished data show that Tox deficiency in CAR T cells results in a phenotype similar to that of Nr4a deficiency, suggesting some sort of regulatory loop between the two transcription factors. Notably, the chromatin accessibility profiles showing enrichment of bZIP binding motifs opens up the large family of bZIP transcription factors to investigation for their roles in exhaustion. Answers to these questions will further elucidate the effects of Nr4a, and the overall transcriptional network regulating the T cell exhaustion phenotype.

Understanding how Nr4a contributes to T cell exhaustion and determining the role of

Nr4a in vivo are critical questions, which could have an important impact on developing new strategies to improve current cancer immunotherapies. Checkpoint blockade therapy and CAR

T cell therapy have both seen success in select groups of patients. With current research showing the beneficial effects of combination therapies or gene editing in adoptive T cell therapy, abrogation of Nr4a expression either through inhibition or deletion of Nr4a has great promise in cancer immunotherapy.

13 1.6 Figures

(Nivolumab)

Figure 1.1. Schematic of inhibitory surface receptors PD-1/PD-L1 and CTLA-4, and the antibodies blocking them. (a) View of the interaction between a T cell and an antigen- presenting cell (APC). CTLA-4 is a negative co-stimulatory receptor; its engagement can be blocked by anti-CTLA-4 antibodies. (b) View of the interaction between an activated T cell and a tumor cell. Once the T cells recognize the MHC-antigen complex on the tumor cells, they will express the negative regulatory receptor PD-1, which can be bound by the PD-L1 on the tumor cells. This PD-1/PD-L1 interaction can be blocked by anti-PD-1 or anti-PD-L1 antibodies.

Figure 1.2. Schematic of first-, second-, and third-generation CARs. In all three generations of the CAR, the design of the extracellular portion of the CAR remains the same, usually consisting of the antibody single-chain variable fragment (the heavy chain VH connected to the light chain VL by a linker); it allows an MHC-independent recognition of a specific antigen by the T cell. From left to right, the first-generation CARs only contained the signaling domain to mimic “signal 1” for T cell activation, usually the zeta chain of CD3 (CD3z). In second-generation CARs, the intracellular signaling domains contain both the “signal 1” T cell receptor stimulation (CD3z) and “signal 2” co-stimulus (CD28, 4-1BB/CD137 for example; CD28 is shown here). Third-generation CARs consist of two co-stimulatory domains; CD28 and 4-1BB are shown here.

Figure 1.3. Schematic of immune response over time and the hierarchical loss of T cell function in T cell exhaustion. In a normal immune response, naïve CD8+ T cells encounter antigen, expand clonally to become effector T cells, and then contract after antigen clearance, leaving a small subset of T cells to develop into memory T cells. In cases of cancer or chronic infection, the antigen persists beyond the effector phase, and the T cells undergo chronic stimulation, leading to T cell exhaustion. CD8+ T cells appear to lose function in a hierarchical manner; in the early stages of exhaustion, CD8+ T cells first lose the ability to produce IL-2 and exert cytotoxic effects, and then progressively lose the ability to produce TNF, and finally, IFNg.

16 1.7 References

Ahmed, N., Brawley, V.S., Hegde, M., Robertson, C., Ghazi, A., Gerken, C., Liu, E., Dakhova, O., Ashoori, A., Corder, A., Gray, T., Wu, M.-F., Liu, H., Hicks, J., Rainusso, N., Dotti, G., Mei, Z., Grilley, B., Gee, A., Rooney, C.M., Brenner, M.K., Heslop, H.E., Wels, W.S., Wang, L.L., Anderson, P., Gottschalk, S., 2015. Human Epidermal Growth Factor Receptor 2 (HER2) –Specific Chimeric Antigen Receptor–Modified T Cells for the Immunotherapy of HER2-Positive Sarcoma. J. Clin. Oncol. 33, 1688–1696. https://doi.org/10.1200/JCO.2014.58.0225

Brown, C.E., Ph, D., Alizadeh, D., Ph, D., Starr, R., Ostberg, J.R., Ph, D., Blanchard, M.S., Ph, D., Kilpatrick, J., Simpson, J., Kurien, A., Priceman, S.J., Ph, D., Wang, X., Ph, D., Harshbarger, T.L., Apuzzo, M.D., Ressler, J.A., Jensen, M.C., Barish, M.E., Ph, D., Chen, M., Ph, D., Portnow, J., Forman, S.J., Badie, B., 2016. Regression of Glioblastoma after Chimeric Antigen Receptor T-Cell Therapy. N. Engl. J. Med. https://doi.org/10.1056/NEJMoa1610497

Buenrostro, J.D., Giresi, P.G., Zaba, L.C., Chang, H.Y., Greenleaf, W.J., 2013. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA- binding proteins and nucleosome position. Nat. Methods 10. https://doi.org/10.1038/nmeth.2688

Chang, J.T., Wherry, E.J., Goldrath, A.W., 2014. Molecular regulation of effector and memory T cell differentiation. Nat. Immunol. https://doi.org/10.1038/ni.3031

Cherkassky, L., Morello, A., Villena-Vargas, J., Feng, Y., Dimitrov, D.S., Jones, D.R., Sadelain, M., Adusumilli, P.S., 2016. Human CAR T cells with cell-intrinsic PD-1 checkpoint blockade resist tumor-mediated inhibition. J. Clin. Invest. 126, 3130–3144. https://doi.org/10.1172/JCI83092

Davila, M.L., Riviere, I., Wang, X., Bartido, S., Park, J., Curran, K., Chung, S.S., Stefanski, J., Borquez-Ojeda, O., Olszewska, M., Qu, J., Wasielewska, T., He, Q., Fink, M., Shinglot, H., Youssif, M., Satter, M., Wang, Y., Hosey, J., Quintanilla, H., Halton, E., Bernal, Y., Bouhassira, D.C.G., Arcila, M.E., Gonen, M., Roboz, G.J., Maslak, P., Douer, D., Frattini, M.G., Giralt, S., Sadelain, M., Brentjens, R., 2014. Efficacy and toxicity management of 19- 28z CAR T cell therapy in B cell acute lymphoblastic leukemia. Sci. Transl. Med. 6. https://doi.org/10.1126/scitranslmed.3008226

Eyquem, J., Mansilla-Soto, J., Giavridis, T., van der Stegen, S.J.C., Hamieh, M., Cunanan, K.M., Odak, A., Gönen, M., Sadelain, M., 2017. Targeting a CAR to the TRAC locus with CRISPR/Cas9 enhances tumour rejection. Nature. https://doi.org/10.1038/nature21405

Fesnak, A.D., June, C.H., Levine, B.L., 2016. Engineered T cells: the promise and challenges of cancer immunotherapy. Nat. Rev. Cancer 16, 566–581. https://doi.org/10.1038/nrc.2016.97

Fraietta, J.A., Nobles, C.L., Sammons, M.A., Lundh, S., Carty, S.A., Reich, T.J., Cogdill, A.P., Morrissette, J.J.D., DeNizio, J.E., Reddy, S., Hwang, Y., Gohil, M., Kulikovskaya, I., Nazimuddin, F., Gupta, M., Chen, F., Everett, J.K., Alexander, K.A., Lin-Shiao, E., Gee, M.H., Liu, X., Young, R.M., Ambrose, D., Wang, Y., Xu, J., Jordan, M.S., Marcucci, K.T., Levine, B.L., Garcia, K.C., Zhao, Y., Kalos, M., Porter, D.L., Kohli, R.M., Lacey, S.F., Berger, S.L., Bushman, F.D., June, C.H., Melenhorst, J.J., 2018. Disruption of TET2 promotes the therapeutic efficacy of CD19-targeted T cells. Nature.

17 https://doi.org/10.1038/s41586-018-0178-z

He, R., Hou, S., Liu, C., Zhang, A., Bai, Q., Han, M., Yang, Y., Wei, G., Shen, T., Yang, X., Xu, L., Chen, X., Hao, Y., Wang, P., Zhu, C., Ou, J., Liang, H., Ni, T., Zhang, X., Zhou, X., Deng, K., Chen, Y., Luo, Y., Xu, J., Qi, H., Wu, Y., Ye, L., 2016. Follicular CXCR5- expressing CD8+T cells curtail chronic viral infection. Nature. https://doi.org/10.1038/nature19317

Hodi, F.S., O’Day, S.J., McDermott, D.F., Weber, R.W., Sosman, J.A., Haanen, J.B., Gonzalez, R., Robert, C., Schadendorf, D., Hassel, J.C., Akerley, W., van den Eertwegh, A.J.M., Lutzky, J., Lorigan, P., Vaubel, J.M., Linette, G.P., Hogg, D., Ottensmeier, C.H., Lebbé, C., Peschel, C., Quirt, I., Clark, J.I., Wolchok, J.D., Weber, J.S., Tian, J., Yellin, M.J., Nichol, G.M., Hoos, A., Urba, W.J., 2010. Improved survival with ipilimumab in patients with metastatic melanoma. N. Engl. J. Med. 363, 711–23. https://doi.org/10.1056/NEJMoa1003466

Im, S.J., Hashimoto, M., Gerner, M.Y., Lee, J., Kissick, H.T., Burger, M.C., Shan, Q., Hale, J.S., Lee, J., Nasti, T.H., Sharpe, A.H., Freeman, G.J., Germain, R.N., Nakaya, H.I., Xue, H.H., Ahmed, R., 2016. Defining CD8+T cells that provide the proliferative burst after PD-1 therapy. Nature. https://doi.org/10.1038/nature19330

Jackson, H.J., Rafiq, S., Brentjens, R.J., 2016. Driving CAR T-cells forward. Nat. Rev. Clin. Oncol. 13, 370–383. https://doi.org/10.1038/nrclinonc.2016.36

June, C.H., Sadelain, M., 2018. Chimeric Antigen Receptor Therapy. N. Engl. J. Med. https://doi.org/10.1056/nejmra1706169

Kaech, S.M., Cui, W., 2012. Transcriptional control of effector and memory CD8+ T cell differentiation. Nat. Rev. Immunol. 12, 749–761. https://doi.org/10.1038/nri3307

Kao, C., Oestreich, K.J., Paley, M.A., Crawford, A., Angelosanto, J.M., Ali, M.A.A., Intlekofer, A.M., Boss, J.M., Reiner, S.L., Weinmann, A.S., Wherry, E.J., 2011. Transcription factor T- bet represses expression of the inhibitory receptor PD-1 and sustains virus-specific CD8+ T cell responses during chronic infection. Nat. Immunol. https://doi.org/10.1038/ni.2046

Larkin, J., Chiarion-Sileni, V., Gonzalez, R., Grob, J.J., Cowey, C.L., Lao, C.D., Schadendorf, D., Dummer, R., Smylie, M., Rutkowski, P., Ferrucci, P.F., Hill, A., Wagstaff, J., Carlino, M.S., Haanen, J.B., Maio, M., Marquez-Rodas, I., McArthur, G.A., Ascierto, P.A., Long, G. V., Callahan, M.K., Postow, M.A., Grossmann, K., Sznol, M., Dreno, B., Bastholt, L., Yang, A., Rollin, L.M., Horak, C., Hodi, F.S., Wolchok, J.D., 2015. Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma. N. Engl. J. Med. 373, 23–34. https://doi.org/10.1056/NEJMoa1504030

Leong, Y.A., Chen, Y., Ong, H.S., Wu, D., Man, K., Deleage, C., Minnich, M., Meckiff, B.J., Wei, Y., Hou, Z., Zotos, D., Fenix, K.A., Atnerkar, A., Preston, S., Chipman, J.G., Beilman, G.J., Allison, C.C., Sun, L., Wang, P., Xu, J., Toe, J.G., Lu, H.K., Tao, Y., Palendira, U., Dent, A.L., Landay, A.L., Pellegrini, M., Comerford, I., McColl, S.R., Schacker, T.W., Long, H.M., Estes, J.D., Busslinger, M., Belz, G.T., Lewin, S.R., Kallies, A., Yu, D., 2016. CXCR5+follicular cytotoxic T cells control viral infection in B cell follicles. Nat. Immunol. https://doi.org/10.1038/ni.3543

Long, A.H., Haso, W.M., Shern, J.F., Wanhainen, K.M., Murgai, M., Ingaramo, M., Smith, J.P.,

18 Walker, A.J., Kohler, M.E., Venkateshwara, V.R., Kaplan, R.N., Patterson, G.H., Fry, T.J., Orentas, R.J., Mackall, C.L., 2015. 4-1BB costimulation ameliorates T cell exhaustion induced by tonic signaling of chimeric antigen receptors. Nat. Med. 21, 581–590. https://doi.org/10.1038/nm.3838

Martinez, G.J., Pereira, R.M., Äijö, T., Kim, E.Y., Marangoni, F., Pipkin, M.E., Togher, S., Heissmeyer, V., Zhang, Y.C., Crotty, S., Lamperti, E.D., Ansel, K.M., Mempel, T.R., Lähdesmäki, H., Hogan, P.G., Rao, A., 2015. The Transcription Factor NFAT Promotes Exhaustion of Activated CD8+ T Cells. Immunity 42. https://doi.org/10.1016/j.immuni.2015.01.006

Maude, S.L., Frey, N., Shaw, P.A., Aplenc, R., Barrett, D.M., Bunin, N.J., Chew, A., Gonzalez, V.E., Zheng, Z., Lacey, S.F., Mahnke, Y.D., Melenhorst, J.J., Rheingold, S.R., Shen, A., Teachey, D.T., Levine, B.L., June, C.H., Porter, D.L., Grupp, S.A., 2014. Chimeric Antigen Receptor T Cells for Sustained Remissions in Leukemia. N. Engl. J. Med. 371, 1507–1517. https://doi.org/10.1056/NEJMoa1407222

McLane, L.M., Abdel-Hakeem, M.S., Wherry, E.J., 2019. CD8 T cell Exhaustion During Chronic Viral Infection and Cancer. Annu. Rev. Immunol. https://doi.org/https://doi.org/10.1146/annrev-immunol-041015-055318

Milner, J.J., Toma, C., Yu, B., Zhang, K., Omilusik, K., Phan, A.T., Wang, D., Getzler, A.J., Nguyen, T., Crotty, S., Wang, W., Pipkin, M.E., Goldrath, A.W., 2017. Runx3 programs CD8+T cell residency in non-lymphoid tissues and tumours. Nature. https://doi.org/10.1038/nature24993

Mognol, G.P., Spreafico, R., Wong, V., Scott-Browne, J.P., Togher, S., Hoffmann, A., Hogan, P.G., Rao, A., Trifari, S., 2017. Exhaustion-associated regulatory regions in CD8 + tumor- infiltrating T cells. Proc. Natl. Acad. Sci. 201620498. https://doi.org/10.1073/pnas.1620498114

Moon, E.K., Wang, L.-C.S., Dolfi, D. V, Wilson, C.B., Ranganathan, R., Sun, J., Kapoor, V., Scholler, J., Pure, E., Milone, M.C., June, C.H., Riley, J., Wherry, E.J., Albelda, S.M., 2014. Multifactorial T cell Hypofunction That is Reversible Can Limit the Efficacy of Chimeric Antibody Receptor-transduced Human T cells in Solid Tumors. Clin. Cancer Res. 20, 4262–73. https://doi.org/10.1158/1078-0432.CCR-13-2627

Morsut, L., Roybal, K.T., Xiong, X., Gordley, R.M., Coyle, S.M., Thomson, M., Lim, W.A., 2016. Engineering Customized Cell Sensing and Response Behaviors Using Synthetic Notch Receptors. Cell. https://doi.org/10.1016/j.cell.2016.01.012

Moskophidis, D., Lechner, F., Pircher, H., Zinkernagel, R.M., 1993. Virus persistence in acutely infected immunocompetent mice by exhaustion of antiviral cytotoxic effector T cells. Nature. https://doi.org/10.1038/362758a0

O’Rourke, D.M., Nasrallah, M.P., Desai, A., Melenhorst, J.J., Mansfield, K., Morrissette, J.J.D., Martinez-Lage, M., Brem, S., Maloney, E., Shen, A., Isaacs, R., Mohan, S., Plesa, G., Lacey, S.F., Navenot, J.M., Zheng, Z., Levine, B.L., Okada, H., June, C.H., Brogdon, J.L., Maus, M. V., 2017. A single dose of peripherally infused EGFRvIII-directed CAR T cells mediates antigen loss and induces adaptive resistance in patients with recurrent glioblastoma. Sci. Transl. Med. 9. https://doi.org/10.1126/scitranslmed.aaa0984

19 Pauken, K.E., Sammons, M.A., Odorizzi, P.M., Manne, S., Godec, J., Khan, O., Drake, A.M., Chen, Z., Sen, D.R., Kurachi, M., Barnitz, R.A., Bartman, C., Bengsch, B., Huang, A.C., Schenkel, J.M., Vahedi, G., Haining, W.N., Berger, S.L., Wherry, E.J., 2016. Epigenetic stability of exhausted T cells limits durability of reinvigoration by PD-1 blockade. Science 354. https://doi.org/10.1126/science.aaf2807

Philip, M., Fairchild, L., Sun, L., Horste, E.L., Camara, S., Shakiba, M., Scott, A.C., Viale, A., Lauer, P., Merghoub, T., Hellmann, M.D., Wolchok, J.D., Leslie, C.S., Schietinger, A., 2017. Chromatin states define tumour-specific T cell dysfunction and reprogramming. Nature. https://doi.org/10.1038/nature22367

Philip, M., Fairchild, L., Sun, L., Horste, E.L., Camara, S., Shakiba, M., Scott, A.C., Viale, A., Lauer, P., Merghoub, T., Hellmann, M.D., Wolchok, J.D., Leslie, C.S., Schietinger, A., 2017b. Chromatin states define tumour-specific T cell dysfunction and reprogramming. Nature 545, 452.

Ribas, A., Wolchok, J.D., 2018. Cancer immunotherapy using checkpoint blockade. Science 359. https://doi.org/10.1126/science.aar4060

Roybal, K.T., Rupp, L.J., Morsut, L., Walker, W.J., McNally, K.A., Park, J.S., Lim, W.A., 2016. Precision Tumor Recognition by T Cells with Combinatorial Antigen-Sensing Circuits. Cell 164, 770–779. https://doi.org/10.1016/j.cell.2016.01.011

Sadelain, M., 2017. Chimeric Antigen Receptors: A Paradigm Shift in Immunotherapy. Annu. Rev. Cancer Biol. https://doi.org/10.1146/annurev-cancerbio-050216-034351

Scharer, C.D., Bally, A.P.R., Gandham, B., Boss, J.M., 2017. Cutting Edge: Chromatin Accessibility Programs CD8 T Cell Memory. J. Immunol. https://doi.org/10.4049/jimmunol.1602086

Schietinger, A., Philip, M., Krisnawan, V.E., Chiu, E.Y., Delrow, J.J., Basom, R.S., Lauer, P., Brockstedt, D.G., Knoblaugh, S.E., Hämmerling, G.J., Schell, T.D., Garbi, N., Greenberg, P.D., 2016. Tumor-Specific T Cell Dysfunction Is a Dynamic Antigen-Driven Differentiation Program Initiated Early during Tumorigenesis. Immunity 45. https://doi.org/10.1016/j.immuni.2016.07.011

Scott-Browne, J.P., López-Moyado, I.F., Trifari, S., Wong, V., Chavez, L., Rao, A., Pereira, R.M., 2016. Dynamic Changes in Chromatin Accessibility Occur in CD8+ T Cells Responding to Viral Infection. Immunity 45, 1327–1340. https://doi.org/10.1016/j.immuni.2016.10.028

Sen, D.R., Kaminski, J., Barnitz, R.A., Kurachi, M., Gerdemann, U., Yates, K.B., Tsao, H.-W., Godec, J., LaFleur, M.W., Brown, F.D., Tonnerre, P., Chung, R.T., Tully, D.C., Allen, T.M., Frahm, N., Lauer, G.M., Wherry, E.J., Yosef, N., Haining, W.N., 2016. The epigenetic landscape of T cell exhaustion. Science 354. https://doi.org/10.1126/science.aae0491

Sharma, P., Allison, J.P., 2015. Immune checkpoint targeting in cancer therapy: Toward combination strategies with curative potential. Cell. https://doi.org/10.1016/j.cell.2015.03.030

Siddiqui, I., Schaeuble, K., Chennupati, V., Fuertes Marraco, S.A., Calderon-Copete, S., Pais Ferreira, D., Carmona, S.J., Scarpellino, L., Gfeller, D., Pradervand, S., Luther, S.A.,

20 Speiser, D.E., Held, W., 2019. Intratumoral Tcf1+PD-1+CD8+ T Cells with Stem-like Properties Promote Tumor Control in Response to Vaccination and Checkpoint Blockade Immunotherapy. Immunity. https://doi.org/10.1016/j.immuni.2018.12.021

Suarez, E.R., Chang, D.-K., Sun, J., Sui, J., Freeman, G.J., Signoretti, S., Zhu, Q., Marasco, W.A., 2014. Chimeric antigen receptor T cells secreting anti-PD-L1 antibodies more effectively regress renal cell carcinoma in a humanized mouse model. Oncotarget. https://doi.org/10.18632/oncotarget.9114

Topalian, S.L., Drake, C.G., Pardoll, D.M., 2015. Immune checkpoint blockade: A common denominator approach to cancer therapy. Cancer Cell. https://doi.org/10.1016/j.ccell.2015.03.001

Wei, S.C., Duffy, C.R., Allison, J.P., 2018. Fundamental mechanisms of immune checkpoint blockade therapy. Cancer Discov. https://doi.org/10.1158/2159-8290.CD-18-0367

Wei, S.C., Levine, J.H., Cogdill, A.P., Zhao, Y., Anang, N.A.A.S., Andrews, M.C., Sharma, P., Wang, J., Wargo, J.A., Pe’er, D., Allison, J.P., 2017. Distinct Cellular Mechanisms Underlie Anti-CTLA-4 and Anti-PD-1 Checkpoint Blockade. Cell. https://doi.org/10.1016/j.cell.2017.07.024

Wherry, E.J., 2011. T cell exhaustion. Nat. Immunol. 131, 492–499. https://doi.org/10.1038/ni.2035

Wherry, E.J., Blattman, J.N., Murali-Krishna, K., van der Most, R., Ahmed, R., 2003. Viral Persistence Alters CD8 T-Cell Immunodominance and Tissue Distribution and Results in Distinct Stages of Functional Impairment. J. Virol. 77, 4911–4927. https://doi.org/10.1128/JVI.77.8.4911-4927.2003

Wherry, E.J., Ha, S.J., Kaech, S.M., Haining, W.N., Sarkar, S., Kalia, V., Subramaniam, S., Blattman, J.N., Barber, D.L., Ahmed, R., 2007. Molecular Signature of CD8+ T Cell Exhaustion during Chronic Viral Infection. Immunity 27, 670–684. https://doi.org/10.1016/j.immuni.2007.09.006

Wherry, E.J., Kurachi, M., 2015. Molecular and cellular insights into T cell exhaustion. Nat. Rev. Immunol. 15, 486–499. https://doi.org/10.1038/nri3862

Yong, C.S.M., Dardalhon, V., Devaud, C., Taylor, N., Darcy, P.K., Kershaw, M.H., 2017. CAR T- cell therapy of solid tumors. Immunol. Cell Biol. https://doi.org/10.1038/icb.2016.128

Youngblood, B., Oestreich, K.J., Ha, S.J., Duraiswamy, J., Akondy, R.S., West, E.E., Wei, Z., Lu, P., Austin, J.W., Riley, J.L., Boss, J.M., Ahmed, R., 2011. Chronic Virus Infection Enforces Demethylation of the Locus that Encodes PD-1 in Antigen-Specific CD8+ T Cells. Immunity 35, 400–412. https://doi.org/10.1016/j.immuni.2011.06.015

Zajac, A.J., Blattman, J.N., Murali-Krishna, K., Sourdive, D.J., Suresh, M., Altman, J.D., Ahmed, R., 1998. Viral immune evasion due to persistence of activated T cells without effector function. J. Exp. Med. https://doi.org/10.1084/jem.188.12.2205

Zhou, X., Yu, S., Zhao, D.M., Harty, J.T., Badovinac, V.P., Xue, H.H., 2010. Differentiation and Persistence of Memory CD8+ T Cells Depend on T Cell Factor 1. Immunity. https://doi.org/10.1016/j.immuni.2010.08.002

21 Chapter 2:

CD8+ CAR Tumor-Infiltrating T Cells Exhibit Similarity to

CD8+ Endogenous Tumor-Infiltrating T Cells

22 Abstract

CAR T cell therapies have shown great promise in B cell malignancies, but have exhibited less success against solid tumors, in part due to T cell exhaustion. To identify the transcriptional regulators in CAR T cell function, we established a murine CAR T cell tumor model in which mice bearing human CD19+ murine melanomas were adoptively transferred with

CD19-targeted CAR T cells. Using phenotyping by flow cytometry, and genome-wide analyses by RNA-seq and ATAC-seq, we show that PD-1highTIM3high exhausted CAR T cells exhibit similar transcriptional and chromatin accessibility compared to PD-1highTIM3high endogenous

CD8+ T cells. Previous studies have suggested that Nr4a may contribute to the exhaustion phenotype in transgenic TCR mouse models or mouse models of chronic infection, and we validate that this observation in CAR T cells as well, and further confirm the upregulation of

Nr4a at the protein level. Finally, we compared the transcriptional and chromatin accessibility data from the CAR T cells to data from human patients and show similar results. These data suggest that the murine CAR T cell tumor model is a good model system in which to study effectors of T cell exhaustion and that Nr4a does have a prominent role in T cell exhaustion.

23 2.1 Introduction

Clinical advances in cancer immunotherapies have been due largely to mechanistic insights gleaned from the use of mouse models, in particular, insights regarding T cell function in response to a persistent antigen. The initial observations of T cell exhaustion (Gallimore et al., 1998; Moskophidis et al., 1993; Zajac et al., 1998) were in fact made using the chronic

LCMV model (Ahmed et al., 1984), arguably the most well-characterized model for studying responses to persistent viral infection or persistent antigenic stimulation. These studies

(Gallimore et al., 1998; Moskophidis et al., 1993; Zajac et al., 1998) identified the first

“exhausted T cells,” virus-specific cells that showed decreased cytotoxic capabilities (decreased cytolysis, decreased cytokine production) and subsequent studies showed that observations in the LCMV mouse model translated to chronically-infected patients in the clinic – exhausted T cells were reported in human patients chronically-infected with HIV and HCV, amongst other chronic viral infections (Betts et al., 2006; Goepfert et al., 2000; Lee et al., 1999). PD-1 was identified as an important receptor in the regulation of exhaustion using the chronic LCMV mouse model, and abrogation of PD-1 using anti-PDL1 resulted in restoration of function in mice infected with LCMV (Barber et al., 2006). These observations were also found in HIV and SIV infection (Day et al., 2006; Trautmann et al., 2006).

There are a variety of mouse tumor models that are used today for studies involving T cell function and cancer immunotherapy, the three most common being the transplantable tumor model, genetically-engineered tumor models, or humanized mouse models. In the transplantable tumor model, many mouse tumor cell lines can be injected subcutaneously, intravenously, or orthotopically into a syngeneic mouse. The initial discovery detailing the anti- tumor effects of targeting CTLA-4 (Leach et al., 1996) was performed using a subcutaneously- transplanted colon carcinoma, and those results led to the development of anti-CTLA-4 checkpoint blockade therapies in cancer patients (Hodi et al., 2010). Models using orthotopic injection of tumor cells offer the added advantage of modeling tumor growth in a specific tumor

24 environment – medulloblastomas in the brain, for example (Pei et al., 2012). While there are limitations to the modeling of the disease process, it is of note that the transplantable tumor models offer a simple and quick experimental setup for the assessment of therapeutic responses, and that many of the early studies showing the effects of checkpoint blockade were performed in transplantable tumor models prior to their validation in the clinic.

Advances in cancer genetics have allowed for the generation of genetically-engineered tumor models, which can promote tissue-specific tumor growth through deletion/loss of function of tumor suppressor genes or forced expression of tumor oncogenes in a spatiotemporally- controlled fashion. These models have the advantage of modeling more precisely the pathogenesis of tumors in situ, and potentially capturing the interplay between the progression of neoplasm and the immune system that are more translatable to clinical studies. A study by

DuPage and colleagues (DuPage et al., 2011) described an autochthonous lung cancer model in which inhalation of Cre recombinase by genetically-modified mice resulted in expression of oncogenic KRAS and deletion of p53, leading to development of tumors in the lung. The authors reported that over the course of tumor growth, the endogenous T cells lost the ability to produce cytokines and effectively migrate into tumors. Separately, Schietinger and colleagues

(Schietinger et al., 2016) developed a tamoxifen-inducible autochthonous liver cancer model using the SV40 large T antigen as both oncogene and antigen, and this (Schietinger et al.,

2016) and a subsequent study (Philip et al., 2017) showed that tumor-specific cells first enter an early state of function responsive to anti-PD1 blockade, versus a later, more fixed state of dysfunction. The subsequent study (Philip et al., 2017) further showed that the chromatin accessibility profiles of PD-1hghi TILs from human melanoma and small cell lung cancer bore a striking resemblance to the chromatin accessibility profiles of the mouse TILs in the state of fixed dysfunction.

Finally, humanized mouse models of xenografts implanted in immunodeficient mice have generated many observations that lead to clinical trials in the CAR T cell field (Jackson et al.,

25 2016). For example, anti-tumor activity of first-, second-, and third-generation CARs were compared in a xenograft model, showing that first-generation CARs exhibited less effective anti- tumor activity than the second- and third-generation CARs, and this was confirmed in consequent clinical trials (Davila et al., 2014; Maude et al., 2014; Till et al., 2008). Of course, the obvious drawbacks to the humanized mouse model are the cost and the incomplete reconstitution of the human immune response.

As mentioned in Chapter 1, studies from our lab and others (Mognol et al., 2017; Pauken et al., 2016; Scott-Browne et al., 2016; Sen et al., 2016) have hinted that Nr4a transcription factors may have a role in CD8+ T cell exhaustion. In light of the recent advancements in immunotherapy, we wanted to explore the role of Nr4a transcription factors in CD8+ T cell exhaustion in a way that would allow us to also explore its use in cancer immunotherapy.

Towards that end, we present below the establishment and validation of a murine CAR T cell tumor model, in which immunocompetent mice bearing human-CD19+ murine melanomas are adoptively transferred with murine CAR T cells targeting human CD19.

2.2 Results and Discussion

To identify changes in gene expression and associated regulatory elements in CAR T cells infiltrating solid tumors, we compared the properties of three distinct classes of CD8+ tumor-infiltrating lymphocytes (TILs): CAR TILs that recognize human CD19 (hCD19), OT-I T cell receptor (TCR)-transgenic TILs (OT-I TILs) that recognize the SIINFEKL peptide from chicken ovalbumin (OVA) presented by H-2Kb, and endogenous TILs from the recipient congenic mice. As targets for the CAR T cells, we engineered the B16-OVA mouse melanoma cell line, the EL4 mouse thymoma cell line, and the MC38 colon adenocarcinoma cell line to express human CD19 (hCD19) (Figure 2.1a, left panels); the resulting B16-OVA-hCD19 cells are recognized by OVA-reactive OT-I as well as by hCD19-reactive CAR-expressing CD8+ T

26 cells. We confirmed that the B16-OVA-hCD19 cell line stably maintained hCD19 expression after subcutaneous growth in syngeneic C57BL/6J mice for 18 days and subsequent culture for

7 days ex vivo (Figure 2.1a, right panel). We also verified that the B16-OVA and B16-OVA- hCD19 cells grew at the same rate in vivo, indicating that addition of the hCD19 antigen did not cause rejection of the tumor cells (Figure 2.1b, left). Based on the growth rate of the tumors in mice, we chose to inoculate mice with 500,000 B16-OVA-hCD19 tumor cells (Figure 2.1b, right).

The CAR T cells express a second-generation CAR in which a myc epitope-tagged single-chain variable fragment specific for hCD19 (Brogdon et al., 2014; Nicholson et al., 1997) was fused to the transmembrane domain of mouse CD28 and the intracellular signaling portions of mouse CD28 and CD3z (Kochenderfer et al., 2010), and was followed in the retroviral construct by a 2A self-cleaving peptide and a mouse Thy1.1 reporter (Figure 2.1c). We achieved 95.5 + 4.0% transduction efficiency of the CAR retrovirus in mouse CD8+ T cells

(Figure 2.1d). The CAR T cells were functional, as they produced the cytokines TNF and IFNg upon restimulation with EL4-hCD19 target cells in culture (Figure 2.1e, f), and exhibited dose- dependent target cell lysis against the B16-OVA-hCD19 cells in vitro (Figure 2.1g). The CAR T cells resembled mock-transduced cells in their surface expression of PD-1, TIM3, and LAG3 under resting conditions (Figure 2.1h).

To assess CAR T cell function, CD45.1+ CD8+ T cells were transduced with the CAR retrovirus and adoptively transferred into C57BL/6J mice bearing B16-OVA-hCD19 tumors, 13 days after tumor inoculation (Figure 2.2a). Eight days after adoptive transfer, we isolated CD8+

CD45.1+ Thy1.1+ CAR TILs as well as endogenous CD45.2+ host T cells (Figure 2.2b, left). For comparison, we performed a parallel analysis with adoptively transferred CD45.1+ CD8+ OT-I

TCR-transgenic cells (Mognol et al., 2017) infiltrating the same B16-OVA-hCD19 tumor (Figure

2.3a, b; for gating scheme see Figure 2.3e, f). Tumor growth rates were comparable in mice

27 that received CAR or OT-I CD8+ T cells (Figure 2.3c), allowing direct comparison of these two

TIL populations; the number of transferred CAR T cells was kept low to minimize tumor rejection

(Figure 2.3d), allowing the effects of genetic manipulations to be more easily observed. On average, the CAR and OT-I T cells comprised ~18% and ~9% of CD8+ TILs in the tumor

(Figure 2.2c).

At 8 days after transfer, all three TIL populations produced low levels of the cytokines

TNF and IFNg upon restimulation (Figure 2.2d, e), confirming their decreased function. All three

TIL populations also contained PD-1highTIM3high cells that are thought to be highly exhausted

(populations A, C and F in Figure 2.2b, 2.3b, top right), as well as PD-1highTIM3low cells, thought to be antigen-specific memory precursors that proliferate after treatment with anti-PD-1/PD-L1

(populations B and D). The endogenous TILs also contained a population of PD-1lowTIM3low T cells (Figure 2.2b, 2.3b, lower right) that resembled naïve T cells (see ATAC-seq data below).

Thus, all three TIL populations – CAR, OT-I and endogenous – developed similar phenotypes of decreased cytokine production and increased inhibitory receptor expression.

We used RNA-sequencing (RNA-seq) (Picelli et al., 2014) and assay for transposase- accessible chromatin followed by sequencing (ATAC-seq) (Buenrostro et al., 2013) to compare the gene expression and chromatin accessibility profiles (Dobin et al., 2013; Lawrence et al.,

2013; Love et al., 2014) of the different populations (A-F) of CD8+ tumor-infiltrating lymphocytes in the CAR T cell system (Figure 2.2b, 2.3b). Principal component analysis of the RNA-seq data showed that the transcriptional profiles of PD-1highTIM3 high CAR TIL populations (A) were similar to those of endogenous PD-1highTIM3 high TILs (C), but distinct from those of CAR and endogenous PD-1highTIM3 low cells (B, D) and endogenous PD-1lowTIM3low cells (E) (Figure 2.4a,

2.5). Similarly, ATAC-seq analysis showed that the genome-wide chromatin accessibility profiles of the endogenous, OT-I and CAR PD-1highTIM3 high and PD-1highTIM3 low TIL subsets were

28 similar to one another, but distinct from those of PD-1lowTIM3low endogenous TILs (Figure 2.4b,

2.6).

We previously showed in the B16-OVA melanoma model that tumor-reactive OT-I TILs

(PD-1highTIM3 high) showed increased expression of genes encoding the transcription factors

Nr4a2 and Tox, various inhibitory surface receptors, and effector genes including granzymes and cytokines, compared to PD-1lowTIM3low tumor-nonreactive P14 TILs (Mognol et al., 2017).

Consistent with these findings, genes encoding transcription factors (Nr4a2, Tox, Tbx21), inhibitory receptors (Pdcd1, Ctla4, Havrc2, Tigit), several granzymes, cytokines (Il21, Ifng, Tnf), and cytokine receptors (Il2ra, Il10ra) were also upregulated in PD-1high CAR TILs and endogenous TILs (populations A-D), compared to endogenous PD-1lowTIM3 low TILs (population

E) (Figure 2.5, panels comparing populations A vs E, B vs E, C vs E and D vs E, Suppl. Table

1). Exhausted PD-1highTIM3high CAR and endogenous TILs (populations A, C) did not show significant differences in Nr4a mRNA expression compared to PD-1highTIM3low (populations B, D)

(Figure 2.5, panels comparing populations A vs B, C vs D, Suppl. Table 1), but Nr4a protein expression was clearly increased in the former TIL populations (A, C) compared to the latter (B,

D) (see below). Finally, consistent with the published literature (Mognol et al., 2017; Scott-

Browne et al., 2016; Wherry et al., 2007), the PD-1highTIM3low antigen-specific memory precursors resembled naïve CD8+ T cells in expressing higher levels of Tcf7, Il7r, Ccr7 and Sell

(encoding CD62L (L-selectin)) mRNAs compared to exhausted PD-1highTIM3high TILs (Figure

2.5, panels comparing populations A vs E, B vs E, C vs E and D vs E, Suppl. Table 1).

Analysis of the ATAC-seq data showed that endogenous PD-1lowTIM3low TILs (population

E) resembled naïve CD8+ T cells in their pattern of chromatin accessibility, and were distinct from PD-1highTIM3high and PD-1highTIM3low TILs (populations A-D) (Figure 2.4b, top panel;

Figure 2.6). Moreover, regions selectively accessible in populations A-D (PD-1high CAR and endogenous TILs) were enriched for consensus Nr4a binding motifs, as well as consensus

29 binding motifs for NFAT, NFkB, bZIP and IRF-bZIP motifs (Figure 2.4b, clusters 8 and 9; discussed below). Again consistent with the published literature (Mognol et al., 2017; Scott-

Browne et al., 2016; Wherry et al., 2007), the chromatin accessibility profile of memory- precursor PD-1highTIM3low TILs (populations B, D) resembled that of naïve CD8+ T cells in showing substantial enrichment for Tcf7 binding motifs (Figure 2.4b, cluster 6).

As mentioned above, PD-1highTIM3high and PD-1highTIM3low TIL populations displayed no significant difference in expression of Nr4a family members at the mRNA level (Figure 2.5, panels comparing populations A vs B, C vs D). However, flow cytometric analysis showed clearly that at the protein level, expression of all three Nr4a transcription factors was higher in

PD-1highTIM3high compared to PD-1highTIM3low TIL populations (Figure 2.7, 2.8). Together, the increased expression of Nr4a family members in exhausted PD-1highTIM3high compared to PD-

1highTIM3low TIL populations, as well as the enrichment for Nr4a binding motifs in the differentially accessible regions of these cells, pointed strongly to Nr4a family members as potential transcriptional effectors of the CD8+ T cell response to chronic antigen stimulation. We therefore focused in this study on the Nr4a family.

Data from human TILs and T cells from chronically infected HIV patients provided further justification for our focus on the Nr4a family. Single-cell RNA-seq data derived from CD8+ T cells infiltrating a human melanoma (Tirosh et al., 2016) revealed that NR4A1 and NR4A2 expression showed a strong positive correlation with PDCD1 (encoding PD-1) and HAVCR2

(encoding TIM3) expression, whereas NR4A3 showed a moderate positive correlation (Figure

2.9a). Consistent with our mouse RNA-seq data, we observed a positive correlation of PDCD1 and HAVCR2 mRNA expression with expression of mRNAs encoding the surface receptors

CD38, TIGIT, and CTLA4 (Figure 2.9b, Suppl. Table 2); and a negative correlation with expression of mRNAs encoding the transcription factor TCF1, the cell surface receptor SELL (L- selectin, or CD62L), and the chemokine receptor CCR7 (Figure 2.9c, Suppl. Table 2). The

30 expression of mRNAs encoding other transcription factors of interest – TOX, TOX2 and IRF4 – also correlated positively with the expression of PDCD1 and HAVCR2 mRNAs (Figure 2.9d,

Suppl. Table 2). Analysis of human ATAC-seq data (Philip et al., 2017; Sen et al., 2016) showed enrichment for Nr4a nuclear receptors, NFAT, bZIP and IRF:bZIP motifs in regions uniquely accessible in CD8+ PD-1high TILs from melanoma and non-small cell lung cancer, and HIV- antigen specific CD8+ T cells from infected humans (Figure 2.10). Taken together, our data show that Nr4a family members are upregulated, and Nr4a nuclear receptor binding motifs are enriched in regions accessible to PD-1high T cells (Mognol et al., 2017; Pauken et al., 2016;

Scott-Browne et al., 2016; Sen et al., 2016) in both human and mouse CD8+ T cells exposed to chronic antigen stimulation.

2.3 Conclusion

As in any scientific endeavor, it is necessary to validate the animal model and show that it accurately models the system to be studied. While others have shown that CAR T cells also exhibit T cell exhaustion by way of decreased cytokine production or upregulation of inhibitory surface receptors, our study is the first to show comprehensive comparisons of transcriptional and chromatin accessibility of CAR T cells to that of transgenic and endogenous tumor- infiltrating T cells. We have also shown comparisons of our genomic data to tumor-infiltrating T cells from human patients, establishing that there is a similar phenotype of exhaustion. These comparisons suggest that the murine CAR T cell tumor model presented here may provide clinically translatable observations of responses to immunotherapeutic interventions or effects of transcriptional regulators in T cell exhaustion. Given the difficulties involved in obtaining CAR T cells from patients for sequencing, difficulties of comparison encountered in the heterogeneity of human samples, and the cost of human xenograft models, we propose this murine CAR T cell model as an attractive alternative for studies in T cell exhaustion and cancer immunotherapy.

31 2.4 Figures

Figure 2.1. Functional assessment of a human CD19 (hCD19)-reactive chimeric antigen receptor (CAR). (a) Left three panels, EL4, MC38 and B16-OVA cell lines expressing hCD19. Gray, parental; black, hCD19-expressing cells. Right; B16-OVA-hCD19 cells recovered after growth in a C57BL/6J mouse followed by culture for 7 days. Gray, isotype control; black, anti- hCD19. Data from one biological replicate in each case. (b) Left, growth curves (mean ± s.e.m., 15 mice per group) of 250,000 B16-OVA parental or B16-OVA-hCD19 tumor cells in vivo after inoculation into C57BL/6J mice. There is no significant difference at any timepoint (ordinary two- way ANOVA, p >0.9999). Right, growth curves (mean ± s.e.m.) of 250,000 (n=5 mice) or 500,000 (n=6 mice) B16-OVA-hCD19 tumor cells in vivo after inoculation (significant difference between the two groups at day 21; ordinary two-way ANOVA; *p=0.0146). (c) Diagram of the CAR construct. LS, leader sequence; SS, signal sequence; myc, myc epitope-tag; scFv, single chain variable fragment against human CD19; followed by the mouse (m) CD28 and CD3ζ signaling domains, the 2A self-cleaving peptide and the mouse Thy1.1 reporter. (d) CAR surface expression monitored by myc epitope-tag and Thy1.1 expression. Mock-transduced CD8+ T cells were used as controls. (e) Cytokine (TNF, IFNg) production by CAR CD8+ T cells after restimulation with EL4-hCD19 cells or with PMA/ionomycin. (f) Quantification of the data shown in (e); p-values (TNF: ****p<0.0001, IFNg: ***p=0.0009) were calculated using a two- tailed unpaired t-test. (g) In vitro killing assay (mean ± s.e.m.) of CD8+ CAR and mock- transduced T cells; data from two biologically independent experiments, each with three technical replicates. (h) Inhibitory surface receptor expression on CAR- and mock-transduced CD8+ T cells cultured in vitro for 5 days; data representative of three biological replicates. Gray shading, isotype control, black line, mock or CAR. Data in (d, e, h) are representative of 3 independent experiments. For all p-value calculations, *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.

Figure 2.2. CAR, OT-I, and endogenous CD8+ TILs isolated from B16-OVA-hCD19 tumors exhibit similar phenotypes. (a) Experimental design to assess CAR and endogenous TILs; 1.5x106 CAR-T cells were adoptively transferred into C57BL/6J mice 13 days after tumor inoculation. (b) Left, representative flow cytometry plot identifying CD45.2+ endogenous TILs and CD45.1+Thy1.1+ CAR-TILs (Thy1.1 encoded in the CAR retroviral vector). Right, flow cytometry plots showing PD-1 and TIM3 surface expression on CD8+ CAR and endogenous TILs. (c) Bar graph showing the percentage of CAR and OT-I TILs in total CD8+ TILs. Bars show mean values with data points for 6, 5 and 11 independent experiments for CAR, OT-I and endogenous TILs respectively. (d) Quantification of cytokine production after restimulation of CAR, OT-I and endogenous CD8+ TILs, compared to cultured CD8+ CAR-T cells stimulated with PMA/ionomycin or left unstimulated. Bars show mean values with data points for 3 independent experiments. All p-values were calculated using two-tailed unpaired t-tests with Welch’s correction, *p≤0.05, **p≤0.01, ***p≤0.001, ****p≤0.0001. (e) Representative flow cytometry plots of cytokine production after restimulation.

Figure 2.3. Tumor-bearing mice adoptively transferred with CAR and OT-I cells exhibit similar tumor growth rates. (a, b) Experimental design to assess CD45.1+ OT-I and CD45.2+ endogenous TILs; 1.5x106 OT-I T cells were adoptively transferred into C57BL/6J mice 13 days after tumor inoculation. (c) Tumor growth curves (mean ± s.e.m.) of mice adoptively transferred with CAR or OT-I CD8+ T cells; graph is a compilation of 3 independent experiments. At days 7 and 21, mouse numbers were: CAR, n=24, 17; for OT-I, n=21, 20. (d) Tumor growth curves (mean ± s.e.m.) of mice adoptively transferred with CAR or PBS; graph is a compilation of 3 independent experiments. At days 7 and 21, mouse numbers were: CAR n=35, 35; PBS n=8, 6. (c, d) For tumor sizes on day 21 after tumor inoculation, p=0.3527 for CAR compared to OT-I (c) and p=0.6240 for PBS compared to CAR (d); p-values were calculated using a two-tailed unpaired t-test with Welch’s correction, *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001. (e, f) Flow cytometry gating scheme for CAR (e) and OT-I (f) CD8+ TILs.

Figure 2.4. CAR and endogenous CD8+ TILs exhibit similar gene expression and chromatin accessibility profiles. (a) Principal component analysis of RNA-seq data from PD- 1hiTIM3hi (A) and PD-1hiTIM3lo (B) CAR TILs, and endogenous PD-1hiTIM3hi (C), PD- 1hiTIM3lo (D), PD-1loTIM3lo (E) TILs. Data represent 3 independent experiments, each using TILs pooled from 9-14 mice. (b) Top, heatmap of mouse CD8+ T cell ATAC-seq data showing log2 fold change from row mean for 9 k-means clusters. Bottom, heatmap of motif enrichment analysis. Data shown for one representative member of transcription factor families enriched in at least one cluster compared to all accessible regions.

Figure 2.5. Adoptively transferred CD8+ CAR T cells infiltrating B16-OVA-hCD19 tumors exhibit phenotypes similar to those of OT-I and endogenous CD8+ TILs. Mean average (MA) plots of genes differentially expressed in the indicated comparisons. Wald test was performed to calculate p-values, as implemented in DESeq2; p-values were adjusted using the Benjamini-Hochberg method. Genes differentially expressed (adjusted p-value <0.1 and log2FoldChange ≥1 or ≤-1) are highlighted. Selected genes are labeled. Top row, comparisons of the CAR TIL populations amongst themselves and to endogenous PD-1loTIM3lo TILs; middle row, comparisons within the endogenous TIL populations; bottom row, comparisons of CAR and endogenous PD-1hiTIM3hi TILs (left), and CAR and endogenous PD-1hiTIM3lo TILs (right).

37 Figure 2.6. Adoptively transferred CAR-expressing mouse CD8+ T cells infiltrating B16- OVA-hCD19 tumors exhibit chromatin accessibility profiles similar to those of + endogenous CD8 TILs. (a) Pairwise euclidean distance comparisons of log2 transformed ATAC-seq density (Tn5 insertions per kilobase) between all replicates at all peaks accessible in at least one replicate. (b) Scatterplot of pairwise comparison of ATAC-seq density (Tn5 insertions per kb) between samples indicated. (c) Genome browser views of sample loci, Pdcd1 (left), Itgav (right); scale range is from 0-600 for all tracks and data are the mean of all replicates. CD8+ TIL populations are as indicated and defined in Fig. 2.2b, 2.3b: (A) PD- 1hiTIM3hi CAR, (B) PD-1hiTIM3lo CAR, (C) PD-1hiTIM3hi endogenous, (D) PD-1hiTIM3lo endogenous, (E) PD-1loTIM3lo endogenous.

Figure 2.7. CAR and endogenous CD8+ PD-1hiTIM3hi TILs exhibit upregulation of Nr4a family members. (a) Quantification of Nr4a expression (MFI); p-values for CAR comparisons (top) were calculated using two-tailed paired t-tests; p-values for endogenous comparisons (bottom) were calculated using row-matching one-way ANOVA with Greenhouse-Geisser correction and Tukey’s multiple comparisons tests; for both calculations, *p≤0.05, **p≤0.01, ***p≤0.001, ****p≤0.0001. Data show mean and individual values from three independent experiments, each using TILs pooled from 9-14 mice. (d) Representative flow cytometry histograms for Nr4a protein expression, comparing CAR and endogenous TIL populations (A-E) defined in Fig. 2.2b.

40 Figure 2.8. Mouse CD8+ TILs exhibit increased expression of Nr4a1, Nr4a2, Nr4a3. (a, b) Flow cytometry gating scheme for CAR (a) and endogenous (b) CD8+ TILs. (c) Representative flow cytometry histograms of Nr4a proteins in PD-1hiTIM3hi TILs, PD-1hiTIM3lo TILs, and PD- 1loTIM3lo TILs and their corresponding fluorescence minus one controls (in off-white). Data are representative of 3 independent experiments in which the sample from each independent experiment is comprised of TILs pooled together from 9-14 mice.

Figure 2.9. Human CD8+ TILs exhibit increased NR4A family members in PD- 1hiTIM3hi TILs. (a) Scatterplots of RNA-seq data showing expression of PDCD1 (x-axis) and HAVCR2 (y-axis) in single cells of human CD8+ TILs (Tirosh et al., 2016), with expression of the indicated NR4A genes shown in the color scale. Each dot represents a single cell. (b, c, d) Plotting in single cells the expression of PDCD1 and HAVCR2 (x- and y-axis respectively), and (displayed by the color scale) the expression of the following: (b) Genes differentially upregulated in PD-1hiTIM3hi TILs relative to PD-1loTIM3lo TILs. (c) Genes differentially downregulated in PD-1hiTIM3hi TILs relative to PD-1loTIM3lo TILs. (d) Genes coding for selected transcription factors showing differential expression in the comparison of PD-1hiTIM3hi TILs relative to PD-1loTIM3lo TILs. Each dot represents a single cell. Human CD8+ TILs data are from [ref. (Tirosh et al., 2016)].

Figure 2.10. Human CD8+ PD-1hi TILs and antigen-specific T cells from HIV patients show similar chromatin accessibility to PD-1hi mouse TILs. Left, mouse CD8+ T cell ATAC-seq data (reproduced from Fig. 2.4b for side-by-side comparison to human ATAC-seq): Top, + heatmap of mouse CD8 T cell ATAC-seq data showing log2 fold change from row mean for 9 k- means clusters. Bottom, heatmap of motif enrichment analysis. Data shown for one representative member of TF families enriched in at least one cluster compared to all accessible regions. Right, human CD8+ T cell ATAC-seq data: Top, human CD8+ T cell ATAC-seq data from PD-1hi TILs (two samples from melanoma, one sample from non-small cell lung tumor [ref. (Philip et al., 2017)]) and antigen-specific CD8+ T cells from HIV-infected individuals [ref. (Sen et al., 2016)] showing log2 fold change from row mean for 9 k-means clusters. Bottom, heatmap of motif enrichment analysis.

44 2.5 Experimental Section

Construction of retroviral vector (MSCV-myc-CAR-2A-Thy1.1) containing chimeric antigen receptor (CAR). The chimeric antigen receptor was pieced together using published portions of the clone FMC63 human CD19 single chain variable fragment (Brogdon et al., 2014;

Nicholson et al., 1997), and the published portions of the murine CD28 and CD3z sequences

(Kochenderfer et al., 2010). The sequence for the myc tag on the N-terminus was obtained from published work (Roybal et al., 2016). This chimeric antigen construct was then cloned into an

MSCV-puro (Clontech) murine retroviral vector in place of the PGK-puro.

Sequence of CAR construct (nt): (1470 bp)

ATGGCTTTGCCAGTGACAGCTCTTCTCCTTCCACTGGCCCTCCTCCTTCACGCCGCTAGGC

CAGAGCAGAAACTTATTTCAGAGGAAGACCTGGACATTCAAATGACACAAACTACTTCTTCT

CTCTCCGCCTCACTTGGTGACCGCGTCACTATTAGTTGCCGCGCTAGTCAAGATATTAGTA

AGTACCTGAATTGGTATCAACAAAAACCTGACGGGACTGTAAAGCTGCTTATATATCATACT

TCTAGGCTGCATTCTGGAGTACCTTCACGATTTAGCGGTAGCGGATCCGGCACCGACTACT

CCCTCACAATTAGCAATCTGGAGCAAGAGGACATAGCCACCTACTTCTGCCAGCAAGGGA

ATACCTTGCCATACACTTTCGGTGGTGGAACTAAGCTCGAAATTACTGGGGGTGGAGGCA

GTGGCGGAGGGGGGTCAGGTGGGGGAGGTTCAGAAGTCAAACTCCAGGAATCTGGACCT

GGACTCGTTGCCCCTTCCCAATCCCTTAGTGTTACATGCACTGTATCAGGTGTATCCCTCC

CTGATTACGGTGTCTCCTGGATTCGGCAGCCTCCTCGGAAGGGTCTCGAGTGGTTGGGAG

TGATTTGGGGGTCTGAAACTACTTATTATAACAGTGCCCTTAAGAGTAGATTGACTATAATT

AAGGATAACAGTAAGTCACAAGTATTCCTCAAAATGAATTCCTTGCAAACAGACGATACAGC

AATATATTACTGCGCAAAACACTACTACTATGGCGGTAGTTACGCTATGGACTATTGGGGT

CAAGGAACCTCTGTCACAGTTTCTAGCATTGAGTTCATGTATCCCCCACCTTACTTGGACAA

TGAAAGGTCTAATGGGACCATCATACACATTAAAGAGAAACACCTGTGTCATACTCAGAGTT

CTCCAAAATTGTTCTGGGCCTTGGTTGTCGTTGCCGGCGTACTGTTCTGTTACGGTCTCTT

45 GGTTACCGTGGCACTTTGTGTTATCTGGACTAATTCCCGGCGGAATCGGGGTGGACAGAG

CGATTACATGAATATGACCCCAAGAAGACCTGGACTGACCAGGAAACCATATCAACCCTAT

GCTCCTGCTCGGGACTTTGCTGCTTACCGCCCACGCGCAAAGTTTTCTAGGAGCGCTGAA

ACCGCTGCCAACCTCCAAGACCCTAATCAGCTTTACAATGAATTGAACTTGGGACGCCGGG

AGGAGTATGACGTCCTTGAGAAAAAGCGGGCTCGGGATCCAGAAATGGGCGGAAAGCAAC

AGAGGCGAAGAAATCCACAAGAGGGGGTCTATAACGCTCTTCAGAAAGATAAAATGGCTG

AGGCATATAGCGAAATTGGGACCAAGGGGGAGAGAAGAAGAGGCAAGGGACATGACGGG

CTTTACCAGGGTTTGTCTACCGCAACAAAAGACACCTATGATGCTTTGCACATGCAAACACT

GGCTCCTAGA

Sequence of CAR construct (aa): (490 aa)

MALPVTALLLPLALLLHAARPEQKLISEEDLDIQMTQTTSSLSASLGDRVTISCRASQ

DISKYLNWYQQKPDGTVKLLIYHTSRLHSGVPSRFSGSGSGTDYSLTISNLEQEDIATYF

CQQGNTLPYTFGGGTKLEITGGGGSGGGGSGGGGSEVKLQESGPGLVAPSQSLSVTCTVSG

VSLPDYGVSWIRQPPRKGLEWLGVIWGSETTYYNSALKSRLTIIKDNSKSQVFLKMNSLQTDDT

AIYYCAKHYYYGGSYAMDYWGQGTSVTVSSIEFMYPPPYLDNERSNGTIIHIKEKHLCHTQSSP

KLFWALVVVAGVLFCYGLLVTVALCVIWTNSRRNRGGQSDYMNMTPRRPGLTRKPYQPYAPA

RDFAAYRPRAKFSRSAETAANLQDPNQLYNELNLGRREEYDVLEKKRARDPEMGGKQQRRR

NPQEGVYNALQKDKMAEAYSEIGTKGERRRGKGHDGLYQGLSTATKDTYDALHMQTLAPR

Construction of retroviral vector containing hCD19. DNA fragment encoding hCD19 was PCR-amplified and cloned into an MSCV-puro (Clontech) murine retroviral vector.

Construction of retroviral vectors containing Cre (MSCV-Cre-IRES-NGFR), and

Nr4a1, Nr4a2, Nr4a3 (MCSV-HA-Nr4a1-IRES-NGFR, MCSV-HA-Nr4a2-IRES-NGFR, MCSV-

HA-Nr4a3-IRES-NGFR). DNA fragment encoding Cre was PCR-amplified and cloned into

MSCV-IRES-NGFR (Addgene Plasmid #27489). DNA fragment encoding Nr4a1 (a kind gift of

46 C.-W. J. Lio, La Jolla Institute for Immunology, La Jolla, CA) was PCR-amplified with 5’ HA-tag and cloned into MSCV-IRES-NGFR. DNA fragment encoding Nr4a2 (Addgene Plasmid #3500) was PCR-amplified with 5’ HA-tag and cloned into MSCV-IRES-NGFR. DNA fragment encoding

Nr4a3 (DNASU Plasmid # MmCD00080978) was PCR-amplified with 5’ HA-tag and cloned into

MSCV-IRES-NGFR.

Eukaryotic cell lines. The EL4 mouse thymoma cell line was purchased from the

American Type Culture Collection (ATCC): EL4 (ATCC® TIB-39™, Mus musculus T cell lymphoma). The B16-OVA mouse melanoma cell line expressing the ovalbumin protein (a kind gift of S. Schoenberger, La Jolla Institute for Immunology, La Jolla, CA) was previously described(Mognol et al., 2017). The 293T cell line was purchased from ATCC: 293T (ATCC®

CRL-3216™). The Platinum-E Retroviral Packaging Cell Line, Ecotropic (PlatE) cell line was purchased from Cell BioLabs, Inc: RV-101. The MC38 mouse colon adenocarcinoma cell line (a kind gift of A.W. Goldrath, UCSD, La Jolla, CA) was originally purchased from Kerafast, Inc

(ENH204).

Construction of mouse tumor cell lines expressing hCD19. B16-OVA, EL4, and MC-

38 cells were transduced with an amphotropic virus containing the human CD19 (hCD19) and then sorted for cells expressing high levels of hCD19.

Preparation of B16-OVA-hCD19 melanoma cells for tumor inoculation. B16-OVA- hCD19 cells were cultured in Dulbecco’s medium (DMEM) with 10% (vol/vol) FBS, 1% L- glutamine, 1% penicillin/ streptomycin and passaged three times prior to inoculation. At the time of injection, cells were trypsinized and resuspended in Hanks balanced salt solution without phenol red at 10 million cells per milliliter. C57BL/6J male mice (8–12 wk old) were injected intradermally with 500,000 B16-OVA-hCD19 cells (50 μL per injection).

47 Mice. C57BL/6J, B6.SJL-PtprcaPepcb/BoyJ, Rag 1-/- mice were obtained from Jackson

Laboratories. Nr4a gene-disrupted strains were obtained from Takashi Sekiya and Akihiko

Yoshimura, with permission from Pierre Chambon. Both male and female mice were used for studies. Mice were age-matched and between 8-12 weeks old when used for experiments, and tumor-bearing mice were first tumor size-matched and then randomly assigned to experimental groups. All mice were bred and/or maintained in the animal facility at the La Jolla Institute for

Immunology. All experiments were performed in compliance with the LJI Institutional Animal

Care and Use Committee (IACUC) regulations.

B16-OVA-hCD19 tumor model. For analysis of CAR CD8+ TILs and endogenous CD8+

TILs: On Day 0, 8-12 week old C57BL/6J mice were injected intradermally with 5 x 105 B16-

OVA-hCD19 cells. After tumors became palpable, tumor measurements were recorded with a manual caliper every other day and tumor area was calculated in centimeters squared (length x width). On Day 13, 1.5 million CAR transduced CD45.1+ CD8+ T cells were adoptively transferred into tumor size-matched tumor-bearing mice. On Day 21, mice were harvested for tumors and spleens.

Preparation of cells for adoptive transfer. CD8+ T cells were isolated and activated with 1ug/mL anti-CD3 and 1 ug/mL anti-CD28 for 1d, then removed from activation and transduced with retrovirus expressing CAR, Cre or corresponding empty vector, or a combination of the above for 1h at 37°C and 2000g. Immediately after the transduction, cells were replaced with media containing 100U of IL-2/mL. 1d following the first transduction, a second transduction was performed and immediately after the transduction, cells were replaced with media containing 100U of IL-2/mL. On the day of adoptive transfer (either day 3 or day 5 post activation), cells were analyzed by flow cytometry and cell counts were obtained using a hemocytometer. The number of CAR-transduced cells was obtained using the cell counts from the hemocytometer and the population percentages obtained from flow cytometry. Cells were

48 then collected, washed with PBS and resuspended at a concentration equivalent of 1.5 million or 3 million CAR-transduced cells per 200uL of PBS. Mice were then adoptively transferred with

200uL of retro-orbital i.v. injections each.

Isolation of tumor infiltrating lymphocytes (TILs) for subsequent analyses. Sample preparation for flow cytometry of TILs from CAR and OT-I experiments: On Day 21, mice were euthanized and perfused with PBS prior to removal of tumor. Tumors were collected, pooled together by group, homogenized, and then dissociated using the MACS Miltenyi Mouse Tumor

Dissociation kit (Miltenyi Biotec) and the gentleMACs dissociator with Octo Heaters (Miltenyi

Biotec) according to manufacturer’s instructions. Tumors were then filtered through a 70uM filter and spun down. Supernatant was aspirated and the tumors were resuspended in the equivalent of 4-5 grams of tumor per 5mL of 1%FBS/PBS for CD8 positive isolation using the Dynabeads

FlowComp Mouse CD8 isolation kit (Invitrogen). After positive isolation, cells were either divided into equal amount for staining and phenotyping with flow cytometry, or stained for cell sorting.

Transfections. Transfections were performed in 10cm dish format, following manufacturer’s instructions for the TransIT®-LT1 Transfection Reagent (Mirus Bio LLC) and using the pCL10A1 and pCL-Eco packaging vectors (the former for the hCD19 virus, and the latter for all other viruses produced).

Retroviral transduction. Retroviral transductions were performed in 6-well plate format, using 3mL of 0.45uM filtered virus and 15 ug/mL of polybrene per well. Double transductions were performed using 1.5mL of each virus for a total of 3mL. Cells were spun at 2000g for 1 h at

37°C in a pre-warmed centrifuge. Immediately after the transduction, cells were replaced with media containing 100U of IL-2/mL. A second transduction is performed the following day.

49 Antibodies. Fluorochrome-conjugated antibodies were purchased from Biolegend, BD

Sciences, eBioscience, and Cell Signaling Technologies. Primary antibody used for chromatin- immunoprecipitation was purchased from Cell Signaling Technologies.

Surface marker staining. Cells were spun down and stained with 1:200 final concentration of antibodies in 50% of 2.4G2 (Fc block) and 50% of FACS Buffer (PBS + 1%

FBS, 2mM EDTA) for 15 min.

Cytokine restimulation and staining. Prior to staining, cells were incubated in media containing 10nM of PMA and 500nM of ionomycin, and 1ug/mL of Brefeldin A at 37ºC for 4 hours. After restimulation, cells were then stained for surface markers and with live/dead dye as described in the surface marker staining protocol above. Cells were then fixed with 4% paraformaldehyde for 30 min, permeabilized with 1X BD Perm/Wash (BD Biosciences) for

30min, and then stained for cytokines at a final concentration of 1:200 in 1X BD Perm/Wash buffer. 1X BD Perm/Wash buffer was prepared according to manufacturer’s instructions. All wash steps were performed with FACS Buffer (PBS + 1% FBS, 2mM EDTA).

Transcription factor staining. Cells were stained for surface markers and with live/dead dye as described in the surface marker staining protocol above. Cells were then fixed, permeabilized, and stained using the Foxp3 / Transcription Factor Staining Kit (eBioscience) according to manufacturer’s instructions. All transcription factor antibodies were used at 1:200 final concentration. The antibody for Ki67 was used at 1:100 final concentration.

Flow Cytometry analysis. All flow cytometry analysis was performed using the

LSRFortessa (BD Biosciences) or the LSR-II (BD Biosciences). Flow data was analyzed using

FlowJo v.10 (Tree Star, Inc). Relevant sample gating has been provided in extended data figures.

50 Statistical analyses. Statistical analyses on flow cytometric data and tumor growth data for experiments involving were performed using the appropriate statistical comparison, including paired or unpaired two-tailed t-tests with Welch’s correction as needed, one-way ANOVA with multiple comparisons test (Tukey’s or Dunnett’s), or ordinary two-way ANOVA (Prism 7,

GraphPad Software). Statistical analyses for survival curves were performed using the Log-rank

(Mantel-Cox) test (Prism 7, GraphPad Software). A p value of ≤ 0.05 was considered statistically significant.

In vitro killing assay. 10,000 B16-OVA-hCD19 cells (target cells) were plated in 100uL of T cell media (or media only for background) in each well in E-plate 96 (ACEA Biosciences

Inc, San Diego, CA). Plate was placed in xCELLigence Real-Time Cell Analysis (RTCA) instrument (ACEA Biosciences Inc, San Diego, CA) after 30 minutes and incubated overnight.

The following day, the plate was removed from the xCELLigence RTCA machine and CD8+

CAR T cells (effector cells) were added in an additional 100uL of T cell media for 30 minutes

(for lysis positive control, 0.2% TritonX was used, for lysis negative control, only media was used). The plate was then placed back into the incubator, and data acquisition began. 5 hours after, the Cell Index (CI) was obtained from each well. Percentage of specific lysis was calculated for each well as follows: % specific lysis = 100 - (CIeach well/(CIpos-CIneg))*100. B16-

OVA-hCD19 cells were thawed out 3 days prior to plating on day 4 (when inoculation would usually occur); mouse CD8+ CAR T cells were prepared prior to the experiment to be added to target cells on day 5 post activation of CD8+ T cells.

Cell sorting. Cell sorting was performed by the LJI Flow Cytometry Core, using the

FACSAria-I, FACSAria-II, or FACSAria-Fusion (BD Biosciences). For ATAC-seq, 50,000 cells were sorted from the isolated CD8+ TILs, with the exception of the OT-I samples, for which

15,000 – 30,000 cells were sorted. In some cases, a second ATAC-seq technical replicate using

50,000 additional cells was prepared in parallel. For RNA-seq, 10,000 cells were sorted from the

51 isolated CD8+ TILs. For the CAR and OT-I experiments, the populations sorted were as follows:

CD8+ CD45.1+ Thy1.1+ PD-1high TIM3high CAR (population A), CD8+ CD45.1+ Thy1.1+ PD-1high

TIM3low CAR (population B), CD8+ CD45.1- Thy1.1- PD-1high TIM3high endogenous cells

(population C), CD8+ CD45.1- Thy1.1- PD-1high TIM3low endogenous cells (population D) and

CD8+ CD45.1- Thy1.1- PD-1low TIM3low endogenous cells (population E), and CD8+ CD45.1+ PD-

1high TIM3high OT-I (population F).

ATAC-seq sample and library preparation. ATAC-seq samples were prepared as in

[ref. (Buenrostro et al., 2013)] with minor modifications. Briefly, cells were sorted into

50%FBS/PBS, spun down, washed once with PBS, and then lysed. Transposition reaction was performed using Nextera enzyme (Illumina) and purified using the MinElute kit (Qiagen) prior to

PCR amplification (KAPA Biosystems) with 10-12 cycles using barcoded primers and 2 x 50 cycle paired-end sequencing (Illumina).

ATAC-seq analysis. Sequencing reads in FASTQ format were generated from Illumina

Basespace (for mouse datasets) or were from published data (Philip et al., 2017; Sen et al.,

2016). Reads were mapped to mouse (mm10) or human (hg19) genomes using bowtie (version

1.0.0, [ref. (Langmead et al., 2009)] with parameters "-p 8 -m 1 --best --strata -X 2000 -S --fr -- chunkmbs 1024." Unmapped reads were processed with trim_galore using parameters “--paired

--nextera --length 37 --stringency 3 -- three_prime_clip_R1 1 --three_prime_clip_R2 1" before attempting to map again using the above parameters. These two bam files were merged and processed to remove reads mapping to the mitochondrial genome and duplicate reads (with picard MarkDuplicates). For mouse datasets, technical replicates were merged together into one single biological replicate at this point. For human datasets, samples with low coverage or which did not meet quality control metrics were excluded. For the one human sample with two technical replicates, these matched closely and number 1 was chosen for the analysis. Genomic coverage for individual replicates were computed on 10 bp windows with MEDIPS [ref. (Chavez

52 et al., 2010)] using full fragments captured by ATAC-seq and used to generate average coverage with the Java Genomics Toolkit [ref. (Palpant, 2016)] for each group.

To identify peaks, the bam files containing unique, non-chrM reads were processed with samtools and awk using "'{if(sqrt($9*$9)<100){print $0}}’" to identify nucleosome free DNA fragments less than 100 nt in length. These subnucleosomal fragments were used to call peak summits for each replicate with MACS2 using parameters "--nomodel –nolambda --keep-dup all

--call-summits." For peak calling, we used a q value cutoff of 0.0001 for mouse datasets and

0.01 for human datasets. The summits for each peak from all replicates were expanded to regions with a uniform size of 200 bp for mouse datasets and 300 bp for human datasets. These regions from all replicates were merged into one global set of peaks and were filtered to remove peaks on the Y chromosome or those that overlapped ENCODE blacklisted regions

(Consortium et al., 2012; Kundaje, 2014).

We used summarizeOverlaps to compute the number of transposase insertions overlapping each peak from all replicates (Lawrence et al., 2013). For differential coverage, raw

ATAC-seq counts in each peak for all replicates of all samples were normalized between replicates using voom [ref. (Ritchie et al., 2015)]. Pairwise contrasts were performed with limma and differentially accessible regions were filtered based on an fdr adjusted p-value of less than

0.01 and an estimated fold-change of at least 4. ATAC-seq density (number of transposase insertion sites per kilobase per million mapped reads) per peak and accessible regions were defined as those with a mean of 5 normalized insertions per kilobase. We used HOMER [ref.

(Heinz et al., 2010)] to identify motifs for transcription factor binding sites enriched in different groups of peaks.

RNA-seq sample and library preparation. Total RNA was extracted using the RNeasy

Micro Kit (Qiagen). SMARTseq2 libraries were prepared as described (Picelli et al., 2014).

Briefly, purified RNA was hybridized to polyA to enrich for mRNA, and then mRNA underwent

53 reverse transcription and template switching prior to an 18-cycle PCR preamplification step.

PCR cleanup was then performed using AMPure XP beads (Beckman Coulter). Quality check of the cDNA library was performed using an Agilent high-sensitivity DNA chip, and 1ng of input cDNA was further used for library preparation using the Nextera XT LibraryPrep kit (Illumina).

Tagmented DNA was amplified with a 12-cycle PCR and again purified with AMPureXP beads.

Library size distribution and yield were evaluated using the Agilent high-sensitivity DNA chip.

Libraries were pooled at equimolar ratios and sequenced with the rapid run protocol on a

HiSEq2500 (Illumina) with 50-nt single-end cycling.

RNA-seq analysis. Quality and adapter trimming was performed on raw RNA-seq reads using TrimGalore! v0.4.5 (http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) with default parameters, retaining reads with minimal length of 36 bp. Resulting single-end reads were aligned to mouse genome mm10 using STAR v2.5.3a [ref. (Dobin et al., 2013)] with alignment parameter outFilterMismatchNmax 4. Technical replicates were merged. RNA-seq analysis was performed at the gene level, employing the transcript annotations of the mouse genome mm10. Reads aligning to annotated features were counted using the summarizeOverlaps function (mode="Union") of the Bioconductor package GenomicAlignments v1.10.1 [ref. (Lawrence et al., 2013)]. The DESeq2 package v1.14.1 [ref. (Love et al., 2014)] was used to normalize the raw counts and identify differentially expressed genes (FDR cutoff of p<0.1, unless otherwise specified). Genes with less than 10 reads total were pre-filtered in all comparisons as an initial step. Transformed values (rlog) were calculated within DESeq2 for data visualization.

Single cell RNA-seq analysis. Data was obtained from a previously published study on the cellular ecosystem of human melanoma tumors (Tirosh et al., 2016). Briefly, [ref. (Tirosh et al., 2016)] profiled by single cell RNA-seq (scRNA-seq) malignant and non-malignant cells

(including immune, stromal, and endothelial cells). Normalized expression values

54 (Ei,j=log2(TPMi,j/10+1), where TPMi,j refers to transcripts per million (TPM) for gene i in cell j) were obtained from Gene Expression Omnibus (GSE72056). For the analysis, we only kept genes with non-zero expression values in at least 10 cells. Given the technical noisiness and gene dropout associated with scRNA-seq data, we used the MAGIC algorithm (van Dijk et al., 2018) for imputation in the matrix of normalized expression values, with diffusion parameter t=2. An R implementation of the MAGIC method was downloaded from

(https://www.krishnaswamylab.org/magic-project). Tumor infiltrating T cells were selected based on the inferred cell type annotation described in [ref. (Tirosh et al., 2016)]. CD8+ T cells were selected based on the expression of CD8A (cells with imputed values ≥ 4) and CD4 (cells with imputed values ≤ 1.5). Imputed values were used for gene expression visualizations.

Data Reporting. No statistical methods were used to predetermine sample size. The investigators were not blinded to group allocation during experiments and outcome assessment.

55 2.6 Acknowledgements

Chapter 2, in full, is a reprint with modifications as it appears in NR4A transcription factors limit CAR T cell function in solid tumours, Nature (2019), published online Feb 27, 2019. doi: 10.1038/s41586-019-0985-x. The dissertation author was the primary investigator and author of this paper. Other authors include Isaac F. López-Moyado, Hyungseok Seo, Chan-

Wang J. Lio, Laura J. Hempleman, Takashi Sekiya, Akihiko Yoshimura, James P. Scott-Browne

& Anjana Rao.

56 2.7 References

Ahmed, R., Salmi, A., Butler, L.D., Chiller, J.M., Oldstone, M.B.A., 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. 60, 521–540. https://doi.org/https://doi.org/10.1084/jem.160.2.521

Barber, D.L., Wherry, E.J., Masopust, D., Zhu, B., Allison, J.P., Sharpe, A.H., Freeman, G.J., Ahmed, R., 2006. Restoring function in exhausted CD8 T cells during chronic viral infection. Nature 439, 682–687. https://doi.org/10.1038/nature04444

Betts, M.R., Nason, M.C., West, S.M., De Rosa, S.C., Migueles, S.A., Abraham, J., Lederman, M.M., Benito, J.M., Goepfert, P.A., Connors, M., Roederer, M., Koup, R.A., 2006. HIV nonprogressors preferentially maintain highly functional HIV-specific CD8+ T cells. Blood. https://doi.org/10.1182/blood-2005-12-4818

Brogdon, J., June, C.H., Loew, A., Maus, M., Scholler, J., 2014. Treatment of cancer using humanized anti-cd19 chimeric antigen receptor. US20140271635 A1.

Chavez, L., Jozefczuk, J., Grimm, C., Dietrich, J., Timmermann, B., Lehrach, H., Herwig, R., Adjaye, J., 2010. Computational analysis of genome-wide DNA methylation during the differentiation of human embryonic stem cells along the endodermal lineage. Genome Res. 20, 1441–1450. https://doi.org/10.1101/gr.110114.110

Consortium, E.P., Dunham, I., Kundaje, A., Aldred, S.F., Collins, P.J., Davis, C. a, Doyle, F., Epstein, C.B., Frietze, S., Harrow, J., Kaul, R., Khatun, J., Lajoie, B.R., Landt, S.G., Lee, B.-K.B.-K., Pauli, F., Rosenbloom, K.R., Sabo, P., Safi, A., Sanyal, A., Shoresh, N., Simon, J.M., Song, L., Trinklein, N.D., Altshuler, R.C., Birney, E., Brown, J.B., Cheng, C., Djebali, S., Dong, X., Ernst, J., Furey, T.S., Gerstein, M., Giardine, B., Greven, M., Hardison, R.C., Harris, R.S., Herrero, J., Hoffman, M.M., Iyer, S., Kelllis, M., Kheradpour, P., Lassman, T., Li, Q., Lin, X., Marinov, G.K., Merkel, A., Mortazavi, A., Parker, S.C.J.S.L., Reddy, T.E., Rozowsky, J., Schlesinger, F., Thurman, R.E., Wang, J., Ward, L.D., Whitfield, T.W., Wilder, S.P., Wu, W., Xi, H.S., Yip, K.Y., Zhuang, J., Bernstein, B.E., Green, E.D., Gunter, C., Snyder, M., Pazin, M.J., Lowdon, R.F., Dillon, L. a L., Adams, L.B., Kelly, C.J., Zhang, J., Wexler, J.R., Good, P.J., Feingold, E. a, Crawford, G.E., Dekker, J., Elinitski, L., Farnham, P.J., Giddings, M.C., Gingeras, T.R., Guigó, R., Hubbard, T.J.T.J., Kellis, M., Kent, W.J., Lieb, J.D., Margulies, E.H., Myers, R.M., Starnatoyannopoulos, J. a, Tennebaum, S. a, Weng, Z., White, K.P., Wold, B., Yu, Y., Wrobel, J., Risk, B. a, Gunawardena, H.P., Kuiper, H.C., Maier, C.W., Xie, L., Chen, X., Mikkelsen, T.S., Gillespie, S., Goren, A., Ram, O., Zhang, X., Wang, L., Issner, R., Coyne, M.J., Durham, T., Ku, M., Truong, T., Eaton, M.L., Dobin, A., Lassmann, T., Tanzer, A., Lagarde, J., Lin, W., Xue, C., Williams, B. a, Zaleski, C., Röder, M., Kokocinski, F., Abdelhamid, R.F., Alioto, T., Antoshechkin, I., Baer, M.T., Batut, P., Bell, I., Bell, K., Chakrabortty, S., Chrast, J., Curado, J., Derrien, T., Drenkow, J., Dumais, E., Dumais, J., Duttagupta, R., Fastuca, M., Fejes-Toth, K., Ferreira, P., Foissac, S., Fullwood, M.J., Gao, H., Gonzalez, D., Gordon, A., Howald, C., Jha, S., Johnson, R., Kapranov, P., King, B., Kingswood, C., Li, G., Luo, O.J., Park, E., Preall, J.B., Presaud, K., Ribeca, P., Robyr, D., Ruan, X., Sammeth, M., Sandu,

57 K.S., Schaeffer, L., See, L.-H., Shahab, A., Skancke, J., Suzuki, A.M., Takahashi, H., Tilgner, H., Trout, D., Walters, N., Wang, H.H., Hayashizaki, Y., Reymond, A., Antonarakis, S.E., Hannon, G.J., Ruan, Y., Carninci, P., Sloan, C. a, Learned, K., Malladi, V.S., Wong, M.C., Barber, G.P., Cline, M.S., Dreszer, T.R., Heitner, S.G., Karolchik, D., Kirkup, V.M., Meyer, L.R., Long, J.C., Maddren, M., Raney, B.J., Grasfeder, L.L., Giresi, P.G., Battenhouse, A., Sheffield, N.C., Showers, K. a, London, D., Bhinge, A. a, Shestak, C., Schaner, M.R., Kim, S.K., Zhang, Z.Z.Z., Mieczkowski, P. a, Mieczkowska, J.O., Liu, Z., McDaniell, R.M., Ni, Y., Rashid, N.U., Kim, M.J., Adar, S., Wang, T., Winter, D., Keefe, D., Iyer, V.R., Sandhu, K.S., Zheng, M., Wang, P., Gertz, J., Vielmetter, J., Partridge, E.C., Varley, K.E., Gasper, C., Bansal, A., Pepke, S., Jain, P., Amrhein, H., Bowling, K.M., Anaya, M., Cross, M.K., Muratet, M. a, Newberry, K.M., McCue, K., Nesmith, A.S., Fisher- Aylor, K.I., Pusey, B., DeSalvo, G., Balasubramanian, S.S., Davis, N.S., Meadows, S.K., Eggleston, T., Newberry, J.S., Levy, S.E., Absher, D.M., Wong, W.H., Blow, M.J., Visel, A., Pennachio, L. a, Elnitski, L., Petrykowska, H.M., Abyzov, A., Aken, B., Barrell, D., Barson, G., Berry, A., Bignell, A., Boychenko, V., Bussotti, G., Davidson, C., Despacio-Reyes, G., Diekhans, M., Ezkurdia, I., Frankish, A., Gilbert, J., Gonzalez, J.M., Griffiths, E., Harte, R., Hendrix, D. a, Hunt, T., Jungreis, I., Kay, M., Khurana, E., Leng, J., Lin, M.F., Loveland, J., Lu, Z., Manthravadi, D., Mariotti, M., Mudge, J., Mukherjee, G., Notredame, C., Pei, B., Rodriguez, J.M., Saunders, G., Sboner, A., Searle, S., Sisu, C., Snow, C., Steward, C., Tapanari, E., Tress, M.L., van Baren, M.J., Washieti, S., Wilming, L., Zadissa, A., Zhengdong, Z., Brent, M., Haussler, D., Valencia, A., Raymond, A., Addleman, N., Alexander, R.P., Auerbach, R.K., Bettinger, K., Bhardwaj, N., Boyle, A.P., Cao, A.R., Cayting, P., Charos, A., Cheng, Y., Eastman, C., Euskirchen, G., Fleming, J.D., Grubert, F., Habegger, L., Hariharan, M., Harmanci, A., Iyenger, S., Jin, V.X., Karczewski, K.J., Kasowski, M., Lacroute, P., Lam, H., Larnarre-Vincent, N., Lian, J., Lindahl-Allen, M., Min, R., Miotto, B., Monahan, H., Moqtaderi, Z., Mu, X.J., O’Geen, H., Ouyang, Z., Patacsil, D., Raha, D., Ramirez, L., Reed, B., Shi, M., Slifer, T., Witt, H., Wu, L., Xu, X., Yan, K.-K., Yang, X., Struhl, K., Weissman, S.M., Tenebaum, S. a, Penalva, L.O., Karmakar, S., Bhanvadia, R.R., Choudhury, A., Domanus, M., Ma, L., Moran, J., Victorsen, A., Auer, T., Centarin, L., Eichenlaub, M., Gruhl, F., Heerman, S., Hoeckendorf, B., Inoue, D., Kellner, T., Kirchmaier, S., Mueller, C., Reinhardt, R., Schertel, L., Schneider, S., Sinn, R., Wittbrodt, B., Wittbrodt, J., Jain, G., Balasundaram, G., Bates, D.L., Byron, R., Canfield, T.K., Diegel, M.J., Dunn, D., Ebersol, A.K., Frum, T., Garg, K., Gist, E., Hansen, R.S., Boatman, L., Haugen, E., Humbert, R., Johnson, A.K., Johnson, E.M., Kutyavin, T.M., Lee, K., Lotakis, D., Maurano, M.T., Neph, S.J., Neri, F. V, Nguyen, E.D., Qu, H., Reynolds, A.P., Roach, V., Rynes, E., Sanchez, M.E., Sandstrom, R.S., Shafer, A.O., Stergachis, A.B., Thomas, S., Vernot, B., Vierstra, J., Vong, S., Weaver, M. a, Yan, Y., Zhang, M., Akey, J. a, Bender, M., Dorschner, M.O., Groudine, M., MacCoss, M.J., Navas, P., Stamatoyannopoulos, G., Stamatoyannopoulos, J. a, Beal, K., Brazma, A., Flicek, P., Johnson, N., Lukk, M., Luscombe, N.M., Sobral, D., Vaquerizas, J.M., Batzoglou, S., Sidow, A., Hussami, N., Kyriazopoulou-Panagiotopoulou, S., Libbrecht, M.W., Schaub, M. a, Miller, W., Bickel, P.J., Banfai, B., Boley, N.P., Huang, H., Li, J.J., Noble, W.S., Bilmes, J. a, Buske, O.J., Sahu, A.O., Kharchenko, P. V, Park, P.J., Baker, D., Taylor, J., Lochovsky, L., 2012. An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74. https://doi.org/10.1038/nature11247

58 28z CAR T cell therapy in B cell acute lymphoblastic leukemia. Sci. Transl. Med. 6. https://doi.org/10.1126/scitranslmed.3008226

Day, C.L., Kaufmann, D.E., Kiepiela, P., Brown, J.A., Moodley, E.S., Reddy, S., Mackey, E.W., Miller, J.D., Leslie, A.J., DePierres, C., Mncube, Z., Duraiswamy, J., Zhu, B., Eichbaum, Q., Altfeld, M., Wherry, E.J., Coovadia, H.M., Goulder, P.J.R., Klenerman, P., Ahmed, R., Freeman, G.J., Walker, B.D., 2006. PD-1 expression on HIV-specific T cells is associated with T-cell exhaustion and disease progression. Nature. https://doi.org/10.1038/nature05115

Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M., Gingeras, T.R., 2013. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21. https://doi.org/10.1093/bioinformatics/bts635

DuPage, M., Cheung, A.F., Mazumdar, C., Winslow, M.M., Bronson, R., Schmidt, L.M., Crowley, D., Chen, J., Jacks, T., 2011. Endogenous T cell responses to antigens expressed in lung adenocarcinomas delay malignant tumor progression. Cancer Cell. https://doi.org/10.1016/j.ccr.2010.11.011

Gallimore, A., Glithero, A., Godkin, A., Tissot, A.C., Plückthun, A., Elliott, T., Hengartner, H., Zinkernagel, R., 1998. Induction and exhaustion of lymphocytic choriomeningitis virus- specific cytotoxic T lymphocytes visualized using soluble tetrameric major histocompatibility complex class I-peptide complexes. J. Exp. Med. https://doi.org/10.1084/jem.187.9.1383

Goepfert, P.A., Bansal, A., Edwards, B.H., Ritter, G.D., Tellez, I., McPherson, S.A., Sabbaj, S., Mulligan, M.J., 2000. A significant number of human immunodeficiency virus epitope- specific cytotoxic T lymphocytes detected by tetramer binding do not produce gamma interferon. J. Virol. https://doi.org/10.1128/jvi.74.21.10249-10255.2000

Heinz, S., Benner, C., Spann, N., Bertolino, E., Lin, Y.C., Laslo, P., Cheng, J.X., Murre, C., Singh, H., Glass, C.K., 2010. Simple Combinations of Lineage-Determining Transcription Factors Prime cis-Regulatory Elements Required for Macrophage and B Cell Identities. Mol. Cell 38, 576–589. https://doi.org/10.1016/j.molcel.2010.05.004

Jackson, H.J., Rafiq, S., Brentjens, R.J., 2016. Driving CAR T-cells forward. Nat. Rev. Clin. Oncol. 13, 370–383. https://doi.org/10.1038/nrclinonc.2016.36

Kochenderfer, J.N., Yu, Z., Frasheri, D., Restifo, N.P., Rosenberg, S.A., 2010. Adoptive transfer of syngeneic T cells transduced with a chimeric antigen receptor that recognizes murine CD19 can eradicate lymphoma and normal B cells. Blood 116, 3875–3886. https://doi.org/10.1182/blood-2010-01-265041

Kundaje, A., 2014. Blacklisted genomic regions for functional genomics analysis [WWW

59 Document]. URL https://sites.google.com/site/anshulkundaje/projects/blacklists

Langmead, B., Trapnell, C., Pop, M., Salzberg, S., 2009. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25. https://doi.org/10.1186/gb-2009-10-3-r25

Lawrence, M., Huber, W., Pagès, H., Aboyoun, P., Carlson, M., Gentleman, R., Morgan, M.T., Carey, V.J., 2013. Software for Computing and Annotating Genomic Ranges. PLoS Comput. Biol. 9. https://doi.org/10.1371/journal.pcbi.1003118

Leach, D.R., Krummel, M.F., Allison, J.P., 1996. Enhancement of antitumor immunity by CTLA- 4 blockade. Science 271. https://doi.org/10.1126/science.271.5256.1734

Lee, P.P., Yee, C., Savage, P.A., Fong, L., Brockstedt, D., Weber, J.S., Johnson, D., Swetter, S., Thompson, J., Greenberg, P.D., Roederer, M., Davis, M.M., 1999. Characterization of circulating T cells specific for tumor-associated antigens in melanoma patients. Nat. Med. https://doi.org/10.1038/9525

Love, M.I., Huber, W., Anders, S., 2014. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15. https://doi.org/10.1186/s13059-014-0550-8

Nicholson, I.C., Lenton, K.A., Little, D.J., Decorso, T., Lee, F.T., Scott, A.M., Zola, H., Hohmann, A.W., 1997. Construction and characterisation of a functional CD19 specific single chain Fv fragment for immunotherapy of B lineage leukaemia and lymphoma. Mol. Immunol. 34, 1157–1165. https://doi.org/10.1016/S0161-5890(97)00144-2

Palpant, T., 2016. Java Genomics Toolkit [WWW Document]. URL https://github.com/timpalpant/java-genomics-toolkit

Pauken, K.E., Sammons, M.A., Odorizzi, P.M., Manne, S., Godec, J., Khan, O., Drake, A.M., Chen, Z., Sen, D.R., Kurachi, M., Barnitz, R.A., Bartman, C., Bengsch, B., Huang, A.C., Schenkel, J.M., Vahedi, G., Haining, W.N., Berger, S.L., Wherry, E.J., 2016. Epigenetic stability of exhausted T cells limits durability of reinvigoration by PD-1 blockade. Science 354. https://doi.org/10.1126/science.aaf2807

Pei, Y., Moore, C.E., Wang, J., Tewari, A.K., Eroshkin, A., Cho, Y.J., Witt, H., Korshunov, A., Read, T.A., Sun, J.L., Schmitt, E.M., Miller, C.R., Buckley, A.F., McLendon, R.E.,

60 Westbrook, T.F., Northcott, P.A., Taylor, M.D., Pfister, S.M., Febbo, P.G., Wechsler-Reya, R.J., 2012. An Animal Model of MYC-Driven Medulloblastoma. Cancer Cell. https://doi.org/10.1016/j.ccr.2011.12.021

Picelli, S., Faridani, O.R., Björklund, Å.K., Winberg, G., Sagasser, S., Sandberg, R., 2014. Full- length RNA-seq from single cells using Smart-seq2. Nat. Protoc. 9, 171–181. https://doi.org/10.1038/nprot.2014.006

Ritchie, M.E., Phipson, B., Wu, D., Hu, Y., Law, C.W., Shi, W., Smyth, G.K., 2015. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47. https://doi.org/10.1093/nar/gkv007

Till, B.G., Jensen, M.C., Wang, J., Chen, E.Y., Wood, B.L., Greisman, H.A., Qian, X., James, S.E., Raubitschek, A., Forman, S.J., Gopal, A.K., Pagel, J.M., Lindgren, C.G., Greenberg, P.D., Riddell, S.R., Press, O.W., 2008. Adoptive immunotherapy for indolent non-hodgkin lymphoma and mantle cell lymphoma using genetically modified autologous CD20-specific T cells. Blood. https://doi.org/10.1182/blood-2007-12-128843

Tirosh, I., Izar, B., Prakadan, S.M., Wadsworth, M.H., Treacy, D., Trombetta, J.J., Rotem, A., Rodman, C., Lian, C., Murphy, G., Fallahi-Sichani, M., Dutton-Regester, K., Lin, J.R., Cohen, O., Shah, P., Lu, D., Genshaft, A.S., Hughes, T.K., Ziegler, C.G.K., Kazer, S.W., Gaillard, A., Kolb, K.E., Villani, A.C., Johannessen, C.M., Andreev, A.Y., Van Allen, E.M., Bertagnolli, M., Sorger, P.K., Sullivan, R.J., Flaherty, K.T., Frederick, D.T., Jané-Valbuena, J., Yoon, C.H., Rozenblatt-Rosen, O., Shalek, A.K., Regev, A., Garraway, L.A., 2016. Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science 352. https://doi.org/10.1126/science.aad0501

61 Trautmann, L., Janbazian, L., Chomont, N., Said, E.A., Gimmig, S., Bessette, B., Boulassel, M.R., Delwart, E., Sepulveda, H., Balderas, R.S., Routy, J.P., Haddad, E.K., Sekaly, R.P., 2006. Upregulation of PD-1 expression on HIV-specific CD8+ T cells leads to reversible immune dysfunction. Nat. Med. https://doi.org/10.1038/nm1482 van Dijk, D., Sharma, R., Nainys, J., Yim, K., Kathail, P., Carr, A.J., Burdziak, C., Moon, K.R., Chaffer, C.L., Pattabiraman, D., Bierie, B., Mazutis, L., Wolf, G., Krishnaswamy, S., Pe’er, D., 2018. Recovering Gene Interactions from Single-Cell Data Using Data Diffusion. Cell. https://doi.org/10.1016/j.cell.2018.05.061

62 Chapter 3:

Nr4a Transcription Factors Limit CAR T Cell Function

in Solid Tumors

63 Abstract

As previously noted, CAR T cell therapies have shown great promise in B cell malignancies, but have exhibited less success against solid tumors, in part due to T cell exhaustion. In our study from the previous chapter as well as from data of other groups, it seems that Nr4a family members play a role in T cell exhaustion. There are three Nr4a family members and they have been shown to have redundant roles in regulatory T cell (Treg cell) development. Considering this redundancy, and using the murine CAR T cell model validated in the previous chapter along with Nr4a-genetically modified mice as donors for CAR T cell production, we show that mice bearing hCD19+ tumors and adoptively transferred with Nr4a

TKO CAR T cells exhibit tumor regression and prolonged survival in comparison to tumor- bearing mice adoptively transferred with wild-type CAR T cells. In additional functional experiments, we show that Nr4a family members do exhibit partially redundant functions in

CD8+ T cell responses. Finally, in our examination of WT and Nr4a TKO tumor-infiltrating lymphocytes, we observed that Nr4a TKO CAR TILs displayed phenotypes and gene expression profiles characteristic of CD8+ effector T cells, and chromatin regions uniquely accessible in Nr4a TKO CAR TILs compared to wild type were enriched for binding motifs for

NFkB and AP-1, transcription factors involved in T cell activation. Our data identify Nr4a transcription factors as major players in the cell-intrinsic program of T cell hyporesponsiveness and point to Nr4a inhibition as a promising strategy for cancer immunotherapy.

64 3.1 Introduction

While the role of Nr4a transcription factors in CD8+ T cell exhaustion was not known at the onset of this thesis work, Nr4a transcription factors have been studied extensively in various contexts. The Nr4a transcription factors, Nr4a1 (Nur77), Nr4a2 (Nurr1), and Nr4a3 (Nor1) are orphan nuclear receptors that contain an N-terminal transactivation domain, followed by zinc finger DNA-binding domains, and the C-terminal ligand domain which has been reported to be enshrouded by hydrophobic residues (Cainan et al., 1995). Nr4a transcription factors have been shown to have roles in inflammation across a number of cell types, and in the development and homeostasis of immune cells.

Early studies showed that Nur77 (Nr4a1) was required for TCR-mediated apoptosis;

Nur77 expression was increased in apoptotic T cell hybridomas and double-positive thymocytes but not in growing cells (Liu et al., 1994; Woronicz et al., 1994), and it was hypothesized and subsequently shown that Nur77 plays a role in apoptosis in the negative selection stage of T cell development – a dominant-negative form of Nr4a lacking the transactivation domain resulted in inhibition of antigen-induced negative selection (Cainan et al., 1995; Zhou et al., 1996). In CD8+

T cell development, Nr4a1 was shown to repress Runx3, regulating the development and frequency of CD8+ T cells; without Nr4a1, Runx3 expression is increased and results in a 2-fold increase in the number and frequency of CD8+ T cells (Nowyhed et al., 2015).

In addition to roles in T cell development and homeostasis, Nr4a also has anti- inflammatory roles in T cells. Sekiya and colleagues (Sekiya et al., 2013) showed that mice in which three Nr4a receptors were deleted in T cells (mice in which floxed Nr4a1 and Nr4a2 genes were conditionally disrupted only in T cells by introducing CD4Cre, and Nr4a3 was globally knocked out) developed severe multi-organ autoimmunity that resulted in death within

21 days. The phenotype of fatal multiorgan autoimmunity was also observed in mice lacking all three Nr4a receptors only in the Treg compartment (mice in which Nr4a1, Nr4a2 genes were

65 conditionally disrupted in Treg cells by introducing Foxp3Cre, and Nr4a3 was globally knocked out), suggesting that Nr4a has a role in mature Treg cells as well (Sekiya et al., 2015). mRNA expression analysis showed that genes exclusively expressed in Treg cells were downregulated in Nr4a TKO Treg cells compared to wildtype, showing that in addition to repressing the autoimmune response, Nr4a also maintains the Treg compartment (Sekiya et al., 2015). Targets of Nr4a are context-dependent – Nr4a1 was able to bind to the Foxp3 promoter in Treg cells but not in Th2 or Tfh cells (Sekiya et al., 2015). In a separate paper, Sekiya and colleagues (Sekiya

+ et al., 2011) showed that in CD4 T cells, Nr4a2 binds directly to the Foxp3 promoter in Treg cells and represses inflammatory cytokines (Sekiya et al., 2011); the suppression of IFNg production results in decreased differentiation to Th1 cells.

Finally, Nr4a also exhibits homeostatic and anti-inflammatory responses outside of T cells. In patrolling monocytes, which resolve inflammation by scavenging debris from the vascular endothelium, Nr4a deficiency results in increased lung metastasis in vivo after tumor inoculation of murine melanoma (Hanna et al., 2015). Transfer of wild-type patrolling monocytes into Nr4a1 KO mice prevented lung tumor metastasis. In a separate study, Nr4a1 deficiency in mice prone to cardiovascular disease resulted in accelerated development of atherosclerosis

(Hanna et al., 2012); Nr4a1 deficiency in macrophages resulted in abnormal toll-like receptor signaling and polarization of the macrophages towards an inflammatory phenotype. Nr4a1 and

Nr4a2 have also been shown to limit neuroinflammatory responses (Saijo et al., 2009; Shaked et al., 2015).

As apparent from the above short survey of Nr4a, targets and functions of Nr4a are cell- type dependent and context-dependent. Based on previous studies both from our lab and in the literature (Mognol et al., 2017; Pauken et al., 2016; Scott-Browne et al., 2016; Sen et al., 2016), we hypothesized that Nr4a transcription factors were performing redundant roles in CD8+ T cell

66 exhaustion, and we present below our study of the role Nr4a transcription factors in exhausted tumor-infiltrating CD8+ CAR T cells.

3.2 Results and Discussion

Based on the results presented in Chapter 1, we focused on the Nr4a family for this initial study. All three members of the Nr4a family are essential for the development of regulatory T cells (Sekiya et al., 2013), suggesting that they may function redundantly in other biological contexts. Considering this expected redundancy, we compared Nr4a-sufficient (WT)

CAR TILs with CAR TILs triply deficient in all three Nr4a transcription factors (Nr4a TKO)

(Figure 3.1a). Because the Nr4a gene-disrupted mice were originally derived from 129/SvJ ES cells (Kadkhodaei et al., 2009), and their genetic background might not have been fully compatible with that of inbred C57BL/6J mice despite stringent backcrossing, we used Rag1- deficient mice as recipients for tumor inoculation to avoid variable rejection of the CAR T cells.

Naïve CD8+ T cells from Nr4a1 fl/fl Nr4a2 fl/fl Nr4a3-/- mice were simultaneously transduced with two retroviruses, the first encoding the CAR-2A-Thy1.1 and the second encoding Cre followed by an IRES-NGFR cassette, to yield Nr4a triple knockout (Nr4a TKO) CAR T cells

(Figure 3.2). As controls, naïve CD8+ T cells from Nr4a1 fl/fl Nr4a2 fl/fl Nr4a3+/+ mice were retrovirally transduced with the CAR-2A-Thy1.1 retrovirus and the empty retrovirus with IRES-

NGFR alone, to yield Nr4a WT CAR T cells (WT) (Figure 3.2). The CAR T cells were adoptively transferred into mice that had been injected with B16-OVA-hCD19 melanoma cells 7 days previously, and tumor growth was monitored for an additional 83 days. Compared to control mice adoptively transferred with WT CD8+ CAR T cells, mice adoptively transferred with Nr4a

TKO CD8+ CAR T cells lacking all three Nr4a transcription factors showed pronounced tumor regression and enhanced survival (Figure 3.1b, d), with the tumor size difference between the three groups (PBS, WT and Nr4a TKO) apparent as early as day 21 after tumor inoculation (i.e.

67 14 days after adoptive transfer) (Figure 3.1c). Nr4a TKO CAR T cells promoted tumor rejection and prolonged survival even in immunocompetent recipient mice (Figure 3.3). Thus, Nr4a transcription factors suppress tumor rejection in the CAR T cell model.

To explore the redundancy of the Nr4a family members in CD8+ T cell function in vivo, we compared the anti-tumor effects of CD8+ CAR T cells lacking individual Nr4a proteins to those of WT and Nr4a TKO CAR T cells (Figure 3.4). The CAR T cells were adoptively transferred into Rag1-deficient mice that had been injected with B16-OVA-hCD19 melanoma cells 7 days previously, and tumor growth was monitored for an additional 83 days (Figure

3.4a). Tumor growth curves and survival curves showed that Nr4a TKO CAR T cells exhibited superior anti-tumor activity compared to CAR T cells from mice lacking any single Nr4a protein

(Figure 3.4b, d). Again, tumor size differences between the WT and the various Nr4a knockouts were observed as early as day 21 after tumor inoculation (Figure 3.4c).

To further confirm the redundant functions of the three Nr4a family members in CD8+ T cell function, we used retroviruses to express Nr4a1, Nr4a2, and Nr4a3 in mouse CD8+ T cells in vitro. Ectopic expression of any of the three Nr4a transcription factors resulted in increased expression of inhibitory surface receptors PD-1, TIM3, 2B4 and GITR and decreased production of the cytokines TNF and IFNg upon restimulation (Figure 3.5). Principal component analysis of the RNA-seq data of cells expressing any given Nr4a transcription factor or the empty vector control indicated that the majority of the variance between these groups was at genes with a similar profile in cells expressing Nr4a family members compared to empty vector (Figure 3.6a,

Suppl. Table 3). In both RNA-seq and ATAC-seq, pairwise comparisons showed very few if any differences between Nr4a family members (Figure 3.6b, c). Thus all three Nr4a proteins induce overlapping changes in gene expression and regulatory element accessibility profiles of CD8+ T cells in vitro as well as in vivo.

68 To assess the phenotypic and genome-wide changes associated with anti-tumor function in Nr4a TKO and WT, we modified the experimental conditions to delay tumor regression (Figure 3.7) and ensure similar tumor sizes between the two groups (Figure 3.8a), for gating scheme of TILs see Figure 3.8b). The number of Nr4a TKO TILs recovered per gram of tumor was not significantly different from the number of WT TILs recovered (Figure 3.8c). At day 21 following tumor inoculation (8 days after adoptive transfer), there was a mild but statistically significant decrease in PD-1 expression when comparing WT and Nr4a TKO TILs, as well as a striking skewing of the total PD-1high population towards low TIM3 expression

(Figure 3.7b, top panels); in the Nr4a TKO TILs, there is a skewing of the TIM3low populations towards low TCF1 expression (Figure 3.8d, top). In tests of effector function, we found that the percentage of cells expressing TNF and both IFNg and TNF was significantly higher in Nr4a

TKO compared to WT TILs (Figure 3.7c). On a bulk level, there was no significant difference in

MFI of TCF1, Tbet, or Eomes (Figure 3.8d, bottom).

We assessed genome-wide changes associated with effector function in the Nr4a TKO and WT TILs by RNA-seq and ATAC-seq. RNA-seq identified 1,076 differentially expressed genes, of which 536 genes were more highly expressed in the Nr4a TKO TILs and 540 were more highly expressed in the WT TILs (Figure 3.9b, Figure 3.10, Suppl. Table 4). Gene set enrichment analysis (Subramanian et al., 2005) (GSEA) using gene sets from effector, memory and exhausted (PD-1highTIM3high) populations isolated from LCMV-infected mice (Scott-Browne et al., 2016) showed that broadly, Nr4a TKO TILs were enriched for genes related to effector function (Figure 3.9a; Figure 3.10; Suppl. Table 4). Specifically, mRNAs encoding IL-2Ra,

TNF, and granzymes were upregulated in Nr4a TKO TILs, consistent with the increased production of TNF observed upon restimulation (Figure 3.7c). In contrast, genes that are typically upregulated in naïve/memory T cells compared to effector populations, such as Sell

(encoding L-selectin/CD62L) and Ccr7 were downregulated in Nr4a TKO compared to WT TILs

69 (Figure 3.9b). Inhibitory surface receptors usually upregulated in hyporesponsive T cells, including Pdcd1, Havcr2, Cd244, Tigit, and Cd38, were also downregulated in Nr4a TKO compared to WT TILs (Figure 3.9b).

To identify the transcriptional targets of individual Nr4a proteins, we considered the genes differentially expressed in Nr4a TKO TILs compared to WT TILs (Figure 3.9b, Suppl.

Table 4), and clustered them by whether their expression changed when Nr4a was ectopically expressed (Figure 3.9c, Suppl. Table 5). Clusters 1 and 2 contain genes that are downregulated in the absence of Nr4a, and upregulated in cells ectopically expressing Nr4a – these include Pdcd1, Havcr2, Cd244 in cluster 1, and Tox, Tigit and Cd38 in cluster 2. Cluster 4 contains genes, notably Tnf and Il21, that were upregulated in the absence of Nr4a, and downregulated in cells ectopically expressing Nr4a. Notably, Runx3 was not among the genes differentially expressed in Nr4a TKO compared to WT TILs, even though previous publications have identified it as a downstream target of Nr4a in the context of CD8+ T cell development

(Nowyhed et al., 2015), and as a gene whose overexpression contributes to tumor regression

(Milner et al., 2017).

ATAC-seq revealed ~2500 differentially accessible regions between WT and Nr4a TKO

TILs (Fig. 4c). Among regions lost in Nr4a TKO TILs, a substantial fraction (~36%) contained

Nr4a binding motifs and a smaller subset contained NFAT binding sites (Figure 3.11a). Of regions more accessible in Nr4a TKO compared to WT TILs, ~71% were enriched for consensus bZIP family motifs and 25% for consensus Rel/ NFkB binding motifs, confirming the established role of bZIP (Fos, Jun, ATF, CREB, etc) and Rel/ NFkB family members in T cell activation and effector function. These data are also consistent with a previous publication suggesting a negative crosstalk between Nr4a and NFkB (Harant and Lindley, 2004; Saijo et al.,

2009). Thus, Nr4a TKO TILs display potent effector function by several independent measures: altered profiles of gene expression that confer decreased expression of inhibitory receptors and

70 increased cytokine production; and strong enrichment of binding motifs for transcription factors involved in effector function in regions of accessible chromatin.

To determine whether differentially accessible regions enriched for Nr4a motifs in fact bound Nr4a, we ectopically expressed individual HA-tagged Nr4a proteins in CD8+ T cells, and confirmed Nr4a binding by chromatin immunoprecipitation using the anti-HA antibody followed by qPCR for selected differentially accessible regions in gene loci that were differentially expressed. For instance, Ccr7, a gene whose expression is high in naïve and memory T cells and decreased in effector cells (Best et al., 2013), is less expressed in effector-like Nr4a TKO compared to WT (more exhausted) TILs (Figure 3.12a); consistent with lower expression, the distal 5’ region of Ccr7 has at least two ATAC-seq peaks that are less prominent in Nr4a TKO than in WT TILs and contain adjacent NFAT and Nr4a binding motifs (Figure 3.12a, left panel, peach lines). We confirmed by ChIP-qPCR that ectopically expressed Nr4a bound these two

Ccr7 enhancer regions (Figure 3.12a, left panel, black lines, and right panels, bar plots). In contrast, the proximal promoter region and first intron of Ccr7 have ATAC-seq peaks that are more prominent in Nr4a TKO compared to WT TILs, and contain bZIP and NFkB motifs (Figure

3.12a, left panel, blue lines). Additional examples (Ifng, Ccr6) are shown in Figure 3.12a (Ifng,

Ccr6: left panels, genome browser views with NFAT and Nr4a motifs marked in peach and bZIP and NFkB motifs in blue; right panels, ChIP-qPCR for Nr4a binding to the indicated accessible regions). Examples of genes with enrichment of bZIP motifs in differentially accessible regions include Il21, which encodes a cytokine involved in effector function (Spolski and Leonard, 2008) and is more highly expressed in Nr4a TKO compared to WT TILs (Figure 3.9b); two regions of the Il21 promoter gain accessibility in Nr4a TKO compared to WT TILs, one of which contains a bZIP motif (Figure 3.12b). Similarly, the cytokine TNF is more highly expressed at both mRNA

(Figure 3.9b) and protein (Figure 3.7c, bottom) levels in Nr4a TKO TILs, and the Tnf locus

71 shows broadly increased accessibility across the promoter and the entire gene in Nr4a TKO

TILs (Figure 3.12b).

Inspection of chromatin accessibility at the Pdcd1 locus (encoding PD-1) shows that each of the Nr4a family members is at least partly responsible for increased accessibility of an enhancer located at ~23 kb 5’ of the Pdcd1 transcription start site (Figure 3.11b, gray box), which has been noted in all mouse models of exhaustion or dysfunction investigated so far

(Mognol et al., 2017; Pauken et al., 2016; Philip et al., 2017; Scott-Browne et al., 2016; Sen et al., 2016). The ATAC-seq peak marking this enhancer is diminished in Nr4a TKO compared to

WT CAR T cells, and increased in T cells ectopically expressing Nr4a1, Nr4a2 or Nr4a3 compared to cells transduced with empty vector alone. ChIP-qPCR shows that all three Nr4a family members bind at this enhancer region of the PD-1 locus (Figure 3.11b, right). Notably, the small decrease in PD-1 MFI that we observe in the Nr4a TKO TILs is supported by a previous publication (Sen et al., 2016) showing that deletion of this ~23 kb enhancer region results in a small decrease in the mean fluorescence intensity of PD-1 staining in the EL-4 thymoma cell line. PD-1 blockade (Pauken et al., 2016) caused a significant (2-fold) decrease in levels of Nr4a2 mRNA, with a smaller decrease in Nr4a1 and Nr4a3 mRNAs (Figure 3.13).

We previously used an engineered NFAT protein, CA-RIT-NFAT1, to mimic a dephosphorylated nuclear NFAT that cannot form cooperative transcriptional complexes with

AP-1 (Fos-Jun) (Martinez et al., 2015). Compared to mock-treated cells, CA-RIT-NFAT1- transduced cells express higher levels of Nr4a transcription factors as well as inhibitory receptors, and display a transcriptional program that mimics in vivo exhaustion, particularly the early stages of “dysfunction” (Martinez et al., 2015; Mognol et al., 2017; Scott-Browne et al.,

2016). Genome-wide analysis of ATAC-seq data showed that regions that were more accessible in WT compared to Nr4a TKO TILs were also more accessible in cultured cells expressing CA-RIT-NFAT1 compared to mock-transduced cells, as well as in Nr4a-expressing

72 compared to empty vector transduced cells (Figure 3.12c). Thus, the regulatory elements sensitive to in vivo reduction of NR4A activity are sensitive to in vitro induction of constitutive

NFAT and NR4A activity. Conversely, regions that were more accessible in Nr4a TKO TILs compared to WT TILs were more accessible in PMA/ionomycin stimulated cells compared to that of resting cells (Figure 3.12c). As PMA/ionomycin stimulation engages bZIP and NFkB family member activity, these data indicate that the increased in vivo effector function and gene expression of Nr4a TKO TILs compared to WT TILs is associated with increased activity of these transcription factors.

3.3 Conclusion

In this study, we demonstrate that Nr4a transcription factors are prominent, redundant effectors of the CD8+ T cell hyporesponsive program downstream of NFAT (Figure 3.14).

Ectopic expression of each individual Nr4a protein represses cytokine function; conversely, TILs lacking all three Nr4a proteins display a gene expression profile characteristic of effector function, including increased expression of granzymes and cytokines. Nr4a TKO TILs also show increased chromatin accessibility at regions containing binding sites for transcription factors of the bZIP and Rel/ NFkB families, which are involved in the classical program of T cell activation.

Thus lack of Nr4a results in a permissive genomic landscape for T cell activation to occur. In addition, Nr4a TKO CAR TILs exhibit an increased frequency of a TIM3- TCF1- population

(Figure 3.8d, top) that may exhibit increased effector function, and is different from the TIM3-

TCF1+ memory/precursor population (He et al., 2016; Im et al., 2016; Leong et al., 2016;

Utzschneider et al., 2016) that expands after PD-1 blockade (Im et al., 2016) but is less represented in Nr4a TKO than in WT CAR TILs.

73 There is a clear relationship between PD-1 and Nr4a, both in our own studies and in previously published data from other labs (Mognol et al., 2017; Pauken et al., 2016; Philip et al.,

2017; Scott-Browne et al., 2016; Sen et al., 2016). Ectopic expression of any individual Nr4a family member in vitro results in upregulation of PD-1 at both the protein and mRNA levels, as well as in increased accessibility at the -23kb upstream enhancer region of the Pdcd1 locus.

Nr4a TKO CAR TILs exhibit decreased accessibility at this enhancer compared to WT CAR

TILs, and all three Nr4a family members bound this enhancer in cells. Taken together, these data indicate that the effect of triple Nr4a deficiency is functionally somewhat similar to PD-1 blockade (Figure 3.14), but Nr4a deletion has a broader effect than PD-1 blockade alone, by affecting a wide range of regulatory elements. In other words, PD-1 is likely to be only one of many genes regulated by the NFAT/ Nr4a axis in both mouse and human T cells – the others encode other inhibitory receptors including CTLA4, TIM3, LAG3 and TIGIT.

Immune cell therapies offer considerable promise for the treatment of cancer (Fesnak et al., 2016; Jackson et al., 2016). In some cases, endogenous CD8+ T cells that do not reject tumors efficiently can be rendered functional by antibodies that block inhibitory receptors such as PD-1 and CTLA-4 (Hodi et al., 2010; Larkin et al., 2015; Ribas et al., 2016). However, treatment with individual blocking antibodies rarely achieves complete cures, and hence the field of cancer immunotherapy is moving towards treatment with blocking antibodies to multiple inhibitory receptors. The NFAT/ Nr4a axis controls the expression of multiple inhibitory receptors, and functionally, treatment of tumor-bearing mice with CAR T cells lacking all three

Nr4a transcription factors resulted in tumor regression and prolonged survival. Thus inhibiting the function of Nr4a family members in tumor-infiltrating T cells could be a promising strategy for cancer immunotherapy.

74 3.4 Figures

Figure 3.1. Nr4a-deficient CAR TILs promote tumor regression and prolong survival. (a) Experimental design; 3x106 WT or Nr4a TKO CAR T cells were adoptively transferred into Rag1-/- mice 7 days after tumor inoculation. PBS was injected as a control. (b) Tumor growth in individual mice. (c) Tumor sizes of individual mice at day 21 (mean ± s.d.); p-values calculated using an ordinary one-way ANOVA with Tukey’s multiple comparisons test. (d) Survival curves; ****p<0.0001 calculated using log-rank (Mantel-Cox) test. Surviving mouse numbers at d7, d21 and d90 were n=21, 14, 0 (PBS); n=35, 25, 1 (WT); n=39, 36, 27 (Nr4a TKO).

75 Figure 3.2. Preparation of wild-type and Nr4a TKO CAR T cells for adoptive transfer. (a) CD8a only staining control (previously tested to be the same as fluorescence minus one controls for CAR+ expression and NGFR+ expression) of CAR T cells prior to adoptive transfer. (b) CAR and NGFR expression of CD8+ WT CAR T cells prior to adoptive transfer. (c) CAR and NGFR expression of CD8+ Nr4a TKO CAR T cells prior to adoptive transfer. Data in (a-c) are representative of 4 independent experiments.

77 Figure 3.3. Prolonged survival of immunocompetent tumor-bearing mice adoptively transferred with CD8+ Nr4a TKO compared to WT CAR T cells. (a) 6x106 CAR T cells were adoptively transferred into C57BL/6J mice 7 days after tumor inoculation. (b) Growth of B16- OVA-hCD19 (left; 13-15 mice per condition) and MC38-hCD19 (right; 10 mice per condition) tumors in individual mice. (c) B16-OVA-hCD19 (left) and MC38-hCD19 (right) tumor sizes (mean ± s.d.) at day 21 and 19 post inoculation respectively. p-values were calculated using an ordinary one-way ANOVA with Tukey’s multiple comparisons test: B16-OVA-hCD19, no significant difference; MC38-hCD19, PBS vs Nr4a TKO ***p=0.0001; PBS vs WT, p=0.3252; WT vs Nr4a TKO, *p=0.0120. (d) Survival curves for mice bearing B16-OVA-hCD19 tumors (left) and MC38-hCD19 tumors (right). p-values calculated using log-rank (Mantel-Cox) test. For B16-OVA-hCD19, surviving mouse numbers at d7, d21, d90 were: PBS, n=13, 11, 0; WT, n=15, 11, 0; Nr4a TKO, n=14, 13, 2; * p=0.0026. For MC38-hCD19, surviving mouse numbers at d7 and d19 were: PBS, n=10, 9; WT, n =10, 7; Nr4a TKO, n=10, 10; all mice died by d23; *p=0.0138. For all p-value calculations in (f, g), *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.

Figure 3.4. Tumor-bearing mice adoptively transferred with CAR CD8+ T cells lacking all three Nr4a family members exhibit prolonged survival compared to mice transferred with wild-type CAR CD8+ T cells or CAR CD8+ T cells lacking only one of the three Nr4a family members. (a) Experimental design; 3x106 WT, Nr4a TKO, Nr4a1 KO, Nr4a2 KO, or Nr4a3 KO CAR T cells were adoptively transferred into Rag1-/- mice 7 days after tumor inoculation. (b) Growth of B16-OVA-hCD19 tumors in individual mice, comprised of 17 or more mice per condition (these data include the WT and Nr4a TKO data from Figure 3.1). (c) Graph shows mean ± s.d. and the individual values of B16-OVA-hCD19 tumor sizes at day 21 after inoculation. p-values were calculated using an ordinary one-way ANOVA with Tukey’s multiple comparisons test; PBS vs WT, *p=0.0395; WT vs Nr4a1 KO, p=n.s.=0.0511; WT vs Nr4a2 KO, **p=0.002, WT vs Nr4a3 KO, *p=0.0161; and WT vs Nr4a TKO, ****p<0.0001. (d) Survival curves. **** p< 0.0001, calculated using log-rank (Mantel-Cox) test. Surviving mouse numbers at d7, d21, and d90 were n=31, 14, 0 for PBS; n=35, 25, 1 for WT; n=17, 12, 0 for Nr4a1 KO; n=17, 15, 1 for Nr4a2 KO; n=32, 22, 11 for Nr4a3 KO; and n=39, 36, 27 for Nr4a TKO. For all p- value calculations, *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.

80 Figure 3.5. Phenotypic features of mouse CD8+ T cells expressing Nr4a1, Nr4a2 or Nr4a3. Mouse CD8+ T cells were isolated, activated, transduced with empty retrovirus or retroviruses encoding HA-tagged Nr4a1, Nr4a2, or Nr4a3 with human NGFR reporter, and assayed on day 5 post activation. (a) Flow cytometry gating of CD8+ NGFR+ empty vector control, Nr4a1, Nr4a2, and Nr4a3-expressing cells at a constant expression level of NGFR reporter. (b) Quantification of surface receptor expression (data from 3 independent replicates). (c) Representative flow cytometry plots of cytokine production upon restimulation with PMA/ionomycin. (d) Quantification of the data in (c). All p-values were calculated using an ordinary one-way ANOVA with Dunnett’s multiple comparisons test; *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.

Figure 3.6. Genomic features of mouse CD8+ T cells expressing Nr4a1, Nr4a2 or Nr4a3. (a) PCA plot of RNA-seq data from in vitro resting mouse CD8+ T cells ectopically expressing Nr4a1, Nr4a2, Nr4a3, or empty vector control. (b) MA plots of genes differentially expressed in the comparisons of ectopic expression of Nr4a1, Nr4a2, or Nr4a3 against empty vector (top row), and pairwise comparisons between the ectopic expression of various Nr4a family members (bottom row). Wald test was performed to calculate p-values, as implemented in DESeq2. p-values were adjusted using the Benjamini-Hochberg method. Genes differentially expressed (adjusted p-value <0.1 and log2FoldChange ≥1 or ≤-1) are highlighted using different colors as indicated in the PCA plot as in (a). Selected genes are labeled. (c) Scatterplot of pairwise comparison of ATAC-seq density (Tn5 insertions per kb) between the indicated samples. Data in (a-c) from two independent experiments, each with two technical replicates.

Figure 3.7. Nr4a-deficient CAR TILs show rescued effector function. (a) Experimental design; 1.5x106 WT or Nr4a TKO CAR T cells were adoptively transferred into Rag1-/- mice 13 days after tumor inoculation, and analyzed 8 days later. (b) Surface PD-1 and TIM3 expression on CAR+ NGFR+ cells with a set level of CAR expression (103 – 104). Representative flow cytometry plots (top), histograms (middle and bottom, left) and means and individual values (right) of 6 independent experiments, each using TILs pooled from 3-8 mice. p-values were calculated using two-tailed paired t-tests with Welch’s correction. (c) Top, representative flow cytometry plots for TNF and IFNg production. Bottom, quantification of 5 independent experiments, each using TILs pooled from 3-8 mice. IL-2 was not detectable above background (not shown). p-values were calculated using two-tailed paired t-tests between unstimulated and stimulated WT and Nr4a TKO CAR TILs. For all p-value calculations, *p≤0.05, **p≤0.01, ***p≤0.001, ****p≤0.0001.

Figure 3.8. CD8+ Nr4a TKO CAR TILs show increased effector function compared to WT CAR TILs. (a) Tumor growth curves (mean ± s.e.m.) after adoptive transfer of 1.5x106 CAR T cells into Rag1-/- mice on day 13 after tumor inoculation. Mouse numbers at d7 and d21 were: WT, n=47, 35; Nr4a TKO, n=41, 32. p-values were calculated using an ordinary 2-way ANOVA with Tukey’s multiple comparisons test; for WT vs Nr4a TKO, p=0.5463. (b) Flow cytometry gating scheme for surface markers, cytokines, and TFs expressed by WT (top) and Nr4a TKO (bottom) TILs. All samples are gated on cells with a set level of CAR expression (103 – 104) within the CAR+ NGFR+ population. (c) Bar plots (mean ± s.d.) showing (left) number of WT and Nr4a TKO CAR TILs per gram of tumor (5 independent experiments; p-value was calculated using a two-tailed ratio paired t-test) and (right) MFI of Ki67 of WT and Nr4a TKO CAR TILs (2 independent experiments). (d) Top, representative flow cytometry for TIM3 and TCF1 expression in WT and Nr4a TKO CAR TILs (2 independent experiments). Bottom, bar plots (mean ± s.d.) of transcription factor expression by WT and Nr4a TKO CAR TILs (6 independent experiments). p-values were calculated using two-tailed paired t-tests. For all p-value calculations, *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.

Figure 3.9. Gene expression profiles indicate increased effector function of Nr4a TKO compared to WT CAR TILs. (a) PCA plot of RNA-seq data from Nr4a TKO or WT CAR TILs. (b) Mean average plots of genes differentially expressed in Nr4a TKO versus WT CAR TILs; p- values calculated using Wald test (as implemented in DESeq2), and adjusted using the Benjamini-Hochberg method. Differentially expressed genes (adjusted p-value <0.1, log2FoldChange ≥1 or ≤-1) are highlighted; selected genes are labeled. (c) Heatmap of genes with opposing expression changes between Nr4a deletion and Nr4a overexpression. Fold change values (log2scale) of genes differentially expressed in Nr4a TKO relative to WT CAR TILs were compared to corresponding values in cells ectopically expressing Nr4a1, Nr4a2, or Nr4a3, and 7 k-means clusters were identified. Genes downregulated after Nr4a deletion/ upregulated after Nr4a overexpression (e.g. Pdcd1, Havcr2, Tox), or upregulated after Nr4a deletion/ downregulated after Nr4a overexpression (e.g. Tnf, Il21) are indicated.

Figure 3.10. Gene set enrichment analyses. (a) Normalized enrichment scores (NES) of gene sets defined from pairwise comparisons of effector, memory and exhausted CD8+ T cells from LCMV-infected mice (Scott-Browne et al., 2016). Enrichment score calculated using a Kolmogorov-Smirnov test, as implemented in gene set enrichment analysis (GSEA). (b) GSEA of RNA-seq data from WT and Nr4a TKO CAR TILs displayed as enrichment plots, ranking genes by log2 fold change in expression between those conditions. The false discovery rate (FDR) for both (a, b) is controlled at a level of 5% by the Benjamini–Hochberg correction. For (e- g), data are from two independent experiments each consisting of 1-2 technical replicates.

Figure 3.11. Chromatin accessibility profiles indicate increased effector function of Nr4a TKO compared to WT CAR TILs. (a) Scatterplot of pairwise comparison of ATAC-seq density (Tn5 insertions per kb) between WT and Nr4a TKO CAR TILs, showing differentially accessible regions and associated de novo identified motifs. (b) Left, genome browser view of the Pdcd1 locus in all previously-mentioned ATAC-seq samples and CA-RIT-NFAT1-transduced cells. Gray bar, exhaustion-specific enhancer ~23 kb 5’ of the Pdcd1 TSS. Top right, histogram of Nr4a expression in cells expressing HA-tagged Nr4a1, Nr4a2, or Nr4a3 (representative of two biological replicates). Bottom right, ChIP-qPCR showing enrichment of HA-tagged Nr4a over background at the Pdcd1 enhancer (technical replicates from one of two independent experiments).

88 Figure 3.12. Nr4a family members bind to predicted Nr4a binding motifs that are more accessible in the WT CAR TILs compared to the Nr4a TKO CAR TILs, and regions more accessible in WT compared to Nr4a TKO CAR TILs are more accessible in CA-RIT- NFAT1- and Nr4a1/2/3-transduced cells. (a) Right top, histogram view showing expression of Nr4a in cells ectopically expressing HA-tagged versions of Nr4a1, Nr4a2, Nr4a3; data are representative of 2 independent experiments. Middle, genome browser views of the Ccr7, Ccr6, Ifng loci for WT CAR TILs compared to Nr4a TKO CAR TILs, including binding motifs for NFAT, Nr4a (Nur77), bZIP, NFkB, and the location of the qPCR primers used. Scale range is 0-1000 for all tracks and data are mean of two independent experiments. Right, bar plots showing enrichment of Nr4a at regions probed; data representative of 2 independent experiments consisting of three technical replicates each. (b) Genome browser views of the Il21 (top), Tnf (bottom) loci incorporating WT CAR TILs compared to Nr4a TKO CAR TILs, including binding motifs for NFAT, Nr4a (Nur77), bZIP, NFkB. Scale range is 0-600 for all tracks except Tnf for which the scale is 0-1000; data are mean of two independent experiments. (c) Top four panels, ATAC-seq data from Nr4a TKO and WT CAR TILs compared with data from cells ectopically expressing CA-RIT-NFAT1, Nr4a1, Nr4a2, or Nr4a3. Bottom panel, ATAC-seq data from Nr4a TKO and WT CAR TILs compared with data from cultured cells stimulated with PMA/ionomycin.

Figure 3.13. Nr4a family members show a moderate decrease in mRNA expression in antigen-specific cells from LCMV-infected mice treated with anti-PDL1 or IgG control. (a) MA plots of genes differentially expressed in cells treated with anti-PDL1 compared to cells treated with IgG control, highlighting two different categories of differentially expressed genes: those with adjusted p-value <0.1 and log2FoldChange ≥0.5 or ≤-0.5 (lighter colors); and those with adjusted p-value < 0.1 and log2FoldChange ≥1 or ≤-1 (darker colors). Selected genes are labeled. Displayed are the number of genes in each category. The sequencing data in this analysis was obtained from [ref. (Pauken et al., 2016)]. Wald test was performed to calculate p- values, as implemented in DESeq2; p-values were adjusted using the Benjamini-Hochberg method.

Figure 3.14. Proposed role of Nr4a in T cells chronically stimulated with antigen.

92 3.5 Experimental Section