Lysophosphatidic Acid Signaling Via Lpa5 Inhibits Cd8 T Cell

+ LYSOPHOSPHATIDIC ACID SIGNALING VIA LPA5 INHIBITS CD8 T CELL

ACTIVATION AND CONTROL OF TUMOR PROGRESSION

SHANNON KUNIKO ODA

B.A., Linfield College, McMinnville, 2004

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirements for the degree of

Doctor of Philosophy

Immunology Program

2013

This thesis for the Doctor of Philosophy degree by

Shannon Kuniko Oda

has been approved for the

Immunology Program

David Riches, Chair

John Cambier

Laurent Gapin

John Kappler

Dennis Voelker

Raul Torres, Advisor

Date 1/22/13 .

Oda, Shannon Kuniko (Ph.D., Immunology)

+ Lysophosphatidic Acid Signaling via LPA5 Inhibits CD8 T Cell Activation and

Control of Tumor Progression

Thesis directed by Associate Professor Raul Torres.

CD8+ T lymphocytes are an important component of adaptive immunity, capable of specific identification of a foreign antigen, rapid proliferation, and cytotoxic activity to kill cells expressing the targeted antigen. T cell activation occurs via engagement of the T cell receptor (TCR) by its cognate antigen within the context of the MHC molecule and is modulated by co-receptors that function to either support or inhibit activation. The adaptive immune system, and T cells in particular, are important for their ability to recognize and eliminate nascent tumors in cancer immunosurveillance and tumor rejection. However multiple inhibitory mechanisms within the tumor microenvironment have been described that protect the tumor from T cells, including increased signaling of inhibitory co- receptors. Lysophosphatidic acid (LPA) is a lysophospholipid that is present at low nanomolar concentrations in the plasma of healthy individuals. It has been well documented that many cancers aberrantly produce LPA and the increased level of LPA has been shown to be beneficial to the tumor, promoting tumorigenesis, invasion, metastases, and vascularization. However the effects of elevated levels of LPA on the adaptive immune response have not been addressed. We demonstrate here that LPA inhibits CD8+ T cell activation in vitro and in vivo. Specifically, we show that CD8+ T cells express LPA receptor 5

iii

(LPA5) that, in the presence of LPA, is able to inhibit TCR signaling and subsequent cell activation and proliferation. These data document that LPA is able to negatively regulate T cell activation and that LPA5 functions as an inhibitory T cell co-receptor. Finally, we have shown that transfer of LPA5- deficient T cells is able to abate tumor progression. We propose that lysophospholipid signaling is an additional protective mechanism of the tumor microenvironment to avoid anti-tumor CD8+ T cell immunity.

The form and content of this abstract are approved. I recommend its publication.

Approved: Raul Torres

ACKNOWLEDGEMENTS

I would first like to thank my mentor, Dr. Raul Torres, for his guidance, enthusiasm, and tireless support for my research project and for my progression as a scientist. My committee members have also made major contributions to this project and served as a valuable source of mentors. I would like to thank the members of the R&R lab, especially Pamela Strauch, for technical assistance and helpful discussions. I have also been fortunate to have a variety of science experiences that influenced me to pursue a career in research, including work in the Lubchenco-Menge lab and the Mourich group at Sarepta Therapeutics.

I would especially like to thank my parents, Stephen and Karen Oda, for their continued support throughout my lengthy education and wisdom for life events. I would also like to thank my sister, Adrienne Oda, for her support and motivation.

Finally, I am indebted to my husband, Freddie Elchlepp, for his patience, love, and encouragement and I look forward to conquering new challenges with him.

TABLE OF CONTENTS

CHAPTER

I. INTRODUCTION AND BACKGROUND ...... 1

The adaptive immune response ...... 1

CD8+ T cells and cancer ...... 4

Lysophosphatidic acid ...... 5

Structure ...... 6

Metabolism ...... 8

LPA signaling ...... 10

LPA receptor 5 ...... 12

LPA and cancer ...... 13

LPA and immune cells ...... 15

Summary ...... 18

II. MATERIALS AND METHODS ...... 20

Mice ...... 20

Calcium mobilization ...... 22

qPCR ...... 23

Flow cytometry ...... 24

Lipid preparation ...... 25

Transwell assay ...... 27

In vitro T cell proliferation and activation assays ...... 27

In vitro effector function assays ...... 28

Generation of bone marrow-derived dendritic cells ...... 30

In vivo SIINFEKL-BMDC stimulation ...... 30

IL-2 ELISA ...... 31

B16.cOVA tumor transfer ...... 31

In vivo cytotoxicity assay with tumor burden ...... 32

Statistical analyses ...... 33

III. LPA SIGNALING INHIBITS CD8+ T CELL ACTIVATION AND PROLIFERATION VIA LPA5 ...... 34

Introduction ...... 34

Results ...... 36

LPA signaling inhibits TCR-mediated calcium mobilization + in CD8 T cells via LPA5 ...... 36

LPA does not influence chemotaxis in naïve T cells ...... 43

Increased LPA signaling inhibits proliferation, accumulation, and activation of OT-I T cells in vitro ...... 44

OTP treatment does not affect APC function and inhibits activation and proliferation to a lesser degree to a high affinity antigen ...... 52

LPA signaling modestly inhibits effector function in vitro ...... 54

Increased LPA signaling dampens proliferation and activation of CD8+ T cells in response to SIINFEKL-BMDC stimulation in vivo ...... 58

Conclusion ...... 65

+ IV. LPA5-DEFICIENT CD8 T CELLS EXHIBIT ENHANCED IMMUNE RESPONSES AND ANTI-TUMOR ACTIVITY ...... 69

Introduction ...... 69

Results ...... 70

vii

Characterization of LPA5-deficient mice ...... 70

LPA5-deficient T cells show enhanced proliferation to SIINFEKL-BMDC in vivo ...... 72

+ LPA5-deficient CD8 T cells do not produce nor consume IL-2 differently ...... 76

–/– + Transfer of LPA5 OT-I CD8 T cells controls tumor growth in vivo ...... 78

Single transfer of LPA5-deficient OT-I T cells is not sufficient to ablate B16.cOVA tumor ...... 83

–/– + LPA5 OT-I CD8 T cells accumulate within the tumor ...... 85

–/– LPA5 T cells kill target cells and produce intracellular cytokines similarly to wild-type T cells...... 90

Conclusion ...... 94

V. DISCUSSION AND FUTURE DIRECTIONS ...... 98

A model for LPA inhibition of CD8+ T cells ...... 98

LPA signaling in CD8+ T cells in vivo: implications for enhancing the CD8+ T cell response ...... 103

Implications for tumor studies and drug development ...... 110

LPA5-deficient T cells more effectively abate tumor progression ...... 110

Considerations for enhancing the CD8+ antitumor response ...... 114

REFERENCES ...... 119

viii

LIST OF FIGURES

FIGURE

1.1 Chemical structures of LPA ...... 7

1.2 LPA metabolism ...... 9

1.3 LPA receptor homology ...... 11

1.4 LPA induces chemorepulsion in splenic B cells ...... 17

–/– 2.1 Generation of LPA5 mice ...... 21

2.2 Inhibition of calcium mobilization by ether- versus ester-LPA ...... 26

3.1 LPA dampens calcium mobilization in major splenic lymphocyte populations ...... 37

3.2 LPA inhibition of CD8+ T cell calcium mobilization is dose dependent ...... 38

3.3 LPA treatment dampens calcium release from intracellular calcium stores ...... 39

3.4 CD8+ T cells express Lpar1, 5, and 6 ...... 40

3.5 Inhibition of calcium mobilization by LPA is ablated in –/– + LPA5 CD8 T cells ...... 41

–/– +/– 3.6 LPA5 T and B cells are immune to LPA inhibition whereas LPA5 lymphocytes exhibit an intermediate phenotype ...... 42

3.7 LPA does not influence chemotaxis of naïve splenic T cells ...... 44

3.8 The ability of LPA to inhibit calcium mobilization is ablated by culture in FBS-media ...... 46

3.9 Media supplemented with BSA is insufficient to support T cell proliferation in vitro ...... 47

3.10 LPA treatment inhibits activation with suboptimal stimulation ...... 48

3.11 LPA signaling ablates OT-I proliferation to G4 peptide stimulation ...... 50

3.12 Increased LPA signaling inhibits activation and prevents accumulation of CD8+ T cells in vitro ...... 51

3.13 OTP treatment does not affect peptide presentation ...... 53

3.14 OTP treatment dampens OT-I activation and proliferation to the higher affinity peptide, SIINFEKL ...... 54

3.15 OTP treatment modestly dampens effector cytokine production ...... 56

3.16 LPA signaling induces a minor inhibition of the killing ability of CD8+ T cells ...... 58

3.17 Peptide pulse of 1-4 hours results in equivalent expression of OVA peptide in H2-Kb ...... 60

3.18 Schematic of peptide-BMDC transfer to induce proliferation of transferred OT-I CD8+ T cells ...... 61

3.19 LPA signaling inhibits TCR-mediated CD8+ T cell activation in vivo ...... 63

3.20 OTP treatment inhibits proliferation of LPA5-deficient OT-I; –/– vehicle-treated LPA5 T cells accumulate to a greater extent ...... 65

–/– 4.1 Characterization of lymphocytes in LPA5 mice ...... 71

4.2 Schematic of adoptive transfer of OT-I CD8+ T cells and subsequent activation to SIINFEKL-BMDC ...... 73

+ 4.3 LPA5-deficient CD8 T cells stimulated by peptide in vivo exhibit enhanced proliferation and accumulation ...... 74

+ 4.4 LPA5-deficient CD8 T cells exhibit reduced activation ...... 76

+/+ –/– 4.5 IL-2 production and consumption is similar by LPA5 and LPA5 CD8+ OT-I T cells ...... 78

4.6 Schematic of tumor implantation and OT-I CD8+ T cell adoptive transfer .. 80

+ 4.7 LPA5-deficient OT-I CD8 T cells control B16.cOVA growth ...... 81

+ 4.8 Transfer of LPA5-deficient CD8 T cells results in smaller tumors ...... 82

+ 4.9 A single transfer of LPA5-deficient CD8 T cells is insufficient to protect mice from B16.cOVA tumors ...... 84

4.10 Proliferation of LPA5-deficient OT-I T cells trends to be increased at day 3 ...... 86

4.11 Trend of slightly increased concentration of LPA5-deficient OT-I CD8+ T cells in the tumor at 3 and 5 days post-transfer ...... 89

–/– +/+ 4.12 Killing ability is similar in the dLN and tumor of LPA5 and LPA5 OT-I T cell recipients 5 days after T cell transfer ...... 91

–/– 4.13 LPA5 OT-I T cells may produce less IFNγ but equivalent TNFα +/+ as LPA5 OT-I T cells ...... 93

5.1 A model for LPA regulation of CD8+ T cell activation to different affinity peptides ...... 100

+ 5.2 Does LPA5 signaling influence CD8 T cell differentiation? ...... 107

–/– 5.3 LPA5 T cells may have decreased memory population formation or an earlier contraction ...... 108

–/– 5.4 A higher proportion of LPA5 OT-I T cells produce granzyme B ...... 110

+ 5.5 A model: interference with LPA5 signaling improves CD8 T cell anti-tumor immunity ...... 113

LIST OF ABBREVIATIONS

AGP alkyl glycerol phosphate AgR antigen receptor APC antigen-presenting cell(s) ATX autotaxin BMDC bone marrow-derived dendritic cell(s) Ca2+ calcium CFSE carboxyfluorescein succinimidyl ester dLN draining lymph node Edg endothelial differentiation gene faf fatty acid-free GPR g-protein coupled receptor LPA lysophosphatidic acid LPAR lysophosphatidic acid receptor LPC lysophosphatidylcholine LPP lipid phosphate phosphatase mAb monoclonal antibody MAG monoacylglycerol MFI mean fluorescent intensity MHC major histocompatibility complex ndLN non-draining lymph node OTP octadecenyl thiophosphate PA phosphatidic acid PLA1,2 phospholipase A1 or A2 S1P sphingosine-1-phosphate s.c. sub-cutaneous TCR T cell receptor

xii

CHAPTER I

INTRODUCTION AND BACKGROUND

The adaptive immune response

The adaptive immune system is composed of specialized cells including B and T lymphocytes. In contrast to the innate immune system which targets foreign invaders by recognition of components shared by many pathogens, the adaptive immune response is capable of specifically identifying and responding to a particular epitope of a pathogen, contributing to pathogen clearance, and subsequently generating a population of memory cells that will mount a stronger attack in the event of a future repeat assault. CD8+ T cells are a subset of T cells that are capable of tremendous expansion and differentiation to effector cells that clear infection throughout the body via cytotoxic activity (1).

The importance of CD8+ T cells was demonstrated (though not fully understood) at the beginning of the last century, when James Murphy experimented with rat tumor cell implantation into chick embryos (2). Although the tumor cells “would not take” in older chicks, tumor cells would successfully grow in chick embryos until the chicks aged (3) and as we now know, developed an immune system (2). Resistance to the tumor could be restored by graft of adult lymphoid tissue near the tumor transplant (4). Although B cells had already been identified, CD8+ T cells would be later discovered to be a distinct cell type by their origin (thymus-derived) and cytotoxic abilities (2).

1 The T cell receptor (TCR) was eventually characterized and revealed that

T cells require foreign antigen to be presented by an antigen presenting cell

(APC) in the context of an endogenous protein, the major histocompatibility complex (MHC) (5-9). T cell activation is initiated when the T cell receptor (TCR) is engaged by the MHC-peptide complex (10, 11). This activation can be bolstered by co-stimulation by cell surface molecules on the APC that engage the appropriate receptor on the T cell, such as the B7 molecule of the APC binding to

CD28 on the T cell (12). Although it was once believed that CD28 signaling was required for T cell activation, it is now known that strong TCR engagement can negate the need for co-stimulation (13). ICOS and 4-1BB are other T cell costimulatory receptors that function to enhance the T cell response (14-16).

At the other end of the spectrum, co-receptors can also function to dampen T cell activation, such as the receptors CTLA-4, PD-1, and TIM-3 (13-15,

17-19). CTLA-4 is expressed at low levels on resting T cells and is upregulated upon activation (13). The ligands for both the inhibitory CTLA-4 and stimulatory

CD28 T cell co-receptors are B7 molecules on APC, however CTLA-4 has a higher affinity for B7 and thus outcompetes CD28 to result in preference to inhibition over costimulation (14).

The importance of inhibitory “brakes” for the T cell response is evidenced in the impressive magnitude of response of which T cells are capable. In response to infection, the naïve CD8+ T cells can yield up to 15 rounds of division, generating more than 10,000 daughter cells within a week (20). In addition, it is becoming increasingly appreciated that T cells must exert a degree of

2 promiscuity, as Bob Mason notably calculated that if T cells were monospecific, the volume of the lymphoid system required to include every naïve CD4+ T cell for each possible MHC-associated peptide (1015 CD4+ T cells) would be over 70 times larger than the volume of a mouse (21, 22). Thus he concluded that the

TCR repertoire must be substantially smaller and necessarily cross-reactive to generate effective immunity (22), and recent studies have estimated that fewer than 108 distinct TCRs exist in the human naïve T cell repertoire (21). The activation of cross-reactive T cells is not equivalent for all peptides and activation of T cells by lower affinity antigens typically results in reduced CD8+ T cell activation and expansion (23, 24). This modulation of response is now being more carefully analyzed, with studies interrogating the differences between peptide affinity and co-stimulation, and resultant clearance of infection, differentiation, and memory response (23-28).

It is because of this ability to generate a powerful response that the activation of these cross-reactive CD8+ T cells must be carefully regulated.

Inappropriate activation of CD8+ T cells has been associated with various autoimmune diseases, including type I diabetes and multiple sclerosis (29).

However, the potency of CD8+ T cell immunity is increasingly being explored for therapeutic strategies to combat infections and cancer, and studies are ongoing to enhance the CD8+ T cell response by modulating signaling of inhibitory and co-stimulatory receptors (30-32).

3 CD8+ T cells and cancer

Cancer immunotherapy is an expanding field of research that can potentially offer a targeted, effective approach by harnessing the patient’s immune system to reject the tumor. Attempts to stimulate antitumor immunity were first performed by surgeon William Coley in 1891 by injecting bacteria into the sarcoma tumors of patients. Although he experienced variable success in inducing cancer remission, ‘Coley’s toxins’ may have promoted antitumoral immunity by activating phagocytes to kill bystander tumor cells (32).

Analyses of genetically manipulated mouse strains have demonstrated that the immune system is important to recognize and eradicate tumor cells (33-

35). CD8+ T cells can specifically identify and eliminate tumor cells expressing tumor antigens recognized by their TCRs, in a process termed “cancer immunosurveillance” (33). The presence of tumor-infiltrating T cells is a positive prognostic factor for various cancers including melanoma, breast, and ovarian cancers (36, 37) and a high proportion of intratumoral CD8+ T cells (of total T cells) favors tumor rejection in both human and mouse (37).

Although tumor masses can support activation and differentiation of naïve

T cells (38), multiple inhibitory mechanisms within the tumor microenvironment have been described that impede tumor rejection by T cells (39-42). These mechanisms range from induced apoptosis via tumoral FasL expression (42, 43) to increased signaling of T cell inhibitory receptors (19), such as the well characterized CTLA-4 molecule (31). An additional obstacle is that high affinity tumor-specific T cells have been reported to be deleted during thymic education,

4 therefore T cells specific for tumor antigens typically display low affinity for these

‘self tumor antigens’ (44-47), resulting in reduced CD8+ T cell activation and expansion (23, 24).

Various therapies are being explored to enhance the T cell response to tumor. One such therapy is to isolate T cells from the patient, stimulate T cells with tumor antigen outside of the immunosuppressive tumor environment, then transfer T cells back into the patient (48). A similar therapy is to determine which cancer associated antigens will stimulate anti-tumor T cells and then immunize patients with those antigens or modified, more immunogenic antigens to induce a specific T cell response (49-51). Other studies have focused on enhancing the function of the APC, such as enhancing dendritic cell activation and antigen presentation, as an indirect route to enhance the T cell response (52, 53).

Importantly, drugs and combinations of drugs to block inhibitory signaling and enhance costimulation are being tested in labs and the clinic (54-56). Of note, blocking CTLA-4 signaling by monoclonal antibody therapy (Yervoy) was approved by the FDA in 2011 and has made significant progress in treating metastatic melanoma patients when conventional therapies have failed (32, 37).

The success of these therapies illustrates the importance of targeting CD8+ T cells to eradicate cancer cells.

Lysophosphatidic acid

Lysophosphatidic acid (LPA) is a phospholipid that is present in all eukaryotic tissues at low concentrations and at higher concentrations in the blood

5 (57). LPA is related to another phospholipid, sphingosine-1-phosphate (S1P), which is best known in the immunology field for its role in directing lymphocyte migration and retention in lymphoid organs (58). The influence of LPA on cell activities is diverse and research is ongoing to better understand the impact of this lipid.

Structure

LPA is a structurally simple molecule, consisting of a phosphate head group, glycerol backbone, and acyl chain. Different molecular species of LPA exist and each LPA receptor has a distinct ligand preference for LPA species

(59). Sources of variation exist in the position of esterification of the fatty acid to the glycerol backbone (sn-1 or sn-2), the type of chemical linkage to the glycerol backbone (alkyl or alkenyl-ether), and the degree of saturation or unsaturation of the aliphatic chain (Figure 1.1)(60). For our purposes, we will refer to LPA molecular species by the length and degree of unsaturation of the carbon chain

(i.e. 16:0 indicates a 16-carbon chain with no degree of unsaturation) and the type of linkage (ether versus ester). LPA is challenging to work with: it easily adheres to plastic surfaces and is notoriously difficult to extract and quantify from tissue and blood, resulting in variable reported concentrations of LPA within tissues and serum (61, 62). However it appears that a consensus of the most common species in human plasma include 18:2, 16:0, 18:1, and 20:4, though the relative abundance of these 4 species is still debated (57, 63-65). The 18:1 molecular species is most commonly used in studies (57).

6 LPA is readily degraded both in vitro and in vivo with a reported half-life as rapid as 3 minutes in the blood after intravenous injection of radiolabeled LPA

(66). Therefore, various metabolically stable LPA analogs have been synthesized to modulate LPA signaling (63, 67). One of these analogs, octadecenyl thiophosphate (OTP) is an 18:1 LPA analog that has been reported to have a longer half-life than LPA in vivo (Figure 1.1)(68, 69).

16:0 esterLPA

16:0 etherLPA

18:1 esterLPA S

OTP

Figure 1.1 Chemical structures of LPA Moieties of LPA can differ in the length of carbon chain and degree of unsaturation (e.g. 18:1 is an aliphatic substituent with an 18-carbon chain with 1 degree of unsaturation). Species of LPA can also differ in the linkage to the glycerol backbone; by either an ester or ether linkage (ether- LPAs are also known as alkyl glycerophosphate, AGP). OTP is a metabolically stable LPA analog that lacks the glycerol backbone and has a sulfur substitution for an oxygen within the phosphate headgroup. Figure adapted from Yanagida et al. (59).

7 Metabolism

LPA is primarily generated in vivo via an extracellular lysophospholipase D, autotaxin (ATX), which converts lysophosphatidylcholine (LPC) to LPA (Figure

1.2A). ATX was originally isolated from a melanoma cell line and identified as a cancer cell motility stimulating factor (70). ATX-null mice die before birth however

ATX+/– mice feature half of the plasma concentration of LPA of normal mice (71,

72), indicating that production by ATX is likely the major source of plasma LPA.

This discovery has led researchers to develop ATX inhibitors as a method of reducing LPA concentrations in vivo (73-75). Another method of LPA generation is the conversion of intracellular phosphatidic acid (PA) to LPA by the enzymes, phospholipase A1 and phospholipase A2 (PLA1, PLA2) (Figure 1.2A) (57, 76).

This mechanism has been shown to be important for LPA production by activated platelets during blood coagulation (77, 78).

Extracellular LPA levels in healthy individuals is a balance of synthesis by

ATX from the abundantly available precursor molecule, LPC, which is secreted by the liver (>200 µM in plasma)(79, 80) and subsequent degradation of LPA

(Figure 1.2A)(81-83). Physiological levels of LPA are tightly regulated with concentrations in plasma reported in the hundred nanomolar range (63). LPA is degraded in vivo by lipid phosphate phosphatases (LPP1-3), which inactivate

LPA by conversion to monoacylglycerol (MAG, Figure 1.2A). In particular, LPP1 has been shown to be important for LPA degradation as LPP1–/– mice metabolize injected LPA 4 times slower than wild-type (66). Significantly elevated

8 LPA levels, up to 10 µM, are observed in individuals with certain tumors and other pathophysiological conditions (84-90).

The structure of autotaxin was solved in 2011 by two independent groups with complementary results (91, 92). Mutational analysis has shown that ATX binds to cells via the SMB2 domain to β3 integrins on the surface of mammalian cells, thus directly releasing LPA to surface LPA receptors and efficiently initiating LPA signaling to that cell (Figure 1.2B) (91-94). The strength of LPA signaling may thus depend on the proximity of ATX and LPP to the LPA receptor

(61).

A B

ATX SMB2 β3

PLA1 or 2 ATX

LPP

MAG

Figure 1.2 LPA metabolism (A) LPA is generated primarily in vivo by the actions of ATX from the precursor molecule LPC. LPA can also be generated from PA by PLA1 and PLA2. LPA is degraded by LPP to MAG. (B) ATX binds to cells via integrins and releases LPA to surface receptors. Figure 1.2A adapted from Moolenaar et al, 2004 (95). Figure 1.2B adapted from Moolenaar and Perrakis, 2011 (94).

9 LPA signaling

Although LPA was previously better known as a lipid metabolism intermediate (63), it is now appreciated for a variety of effects on cell function by binding to extracellular receptors and induction of intracellular signaling pathways.

There are currently 6 confirmed LPA receptors that signal by coupling to and activating G proteins (GPCR)(57) and these receptors are annotated LPA1-6 (96).

Although the first three receptors were discovered in quick succession (LPA1,

1996 (97); LPA2, 1998 (97, 98); LPA3, 1999 (99)), LPA4-6 were discovered up to

10 years later and other proposed LPA receptors are still awaiting validation (57,

96). LPA4-6 were de-orphaned later than LPA1-3 as they were discovered to belong to a different branch of the GPCR family (76), revealing convergent evolution of these two families of receptors. Phylogenetic analysis reveals that while LPA1-3 share 50-57% amino acid sequence and LPA4-6 share 35-55% homology, the LPA homology between phylogenetic groups is low (Figure

1.3)(76). LPA1-3 share homology with classical S1P receptors (S1P1-5) whereas

LPA4-6 share similarities with purinergic receptors (57, 60), allowing diverse functions. In addition, different LPA receptors preferentially couple to specific G proteins and the combination of LPA receptor with a given G protein increases the variety of signaling pathways (61).

10 Purinergic receptors

Figure 1.3 LPA receptor homology A phylogenetic tree for selected human GPCR. LPA receptors are circled in red and S1P receptors are circled in blue. LPA1-3 share homology with S1P receptors and other Edg family receptors while LPA4-6 are more closely aligned with the purinergic receptors. Figure adapted from Ishii et al, 2009 (76).

In addition to these extracellular receptors, an intracellular receptor,

PPARγ, has been described that is stimulated by unsaturated LPA molecular species (100-102). Extracellular LPA had no effect on PPARγ signaling (101) and

PPARγ–/– mice have been shown to have defective arterial wall remodeling, though the physiological role(s) of LPA-PPARγ interactions remain to be fully elucidated (103).

11 LPA has been reported to induce a variety of sometimes opposing activities in different cell types, including proliferation, protection from apoptosis, cell death, migration, adhesion, and activation (63, 79, 104, 105). Although it is perplexing that a structurally simple molecule can have such a varied impact, this heterogeneity is mediated by the diversity of LPA receptors coupled with the different G protein signaling pathways (106). Edg family receptors (LPA1-3) have been reported to induce Ca2+ mobilization, mediate protein-kinase activation, inhibit adenylyl cyclase, and prefer acyl-LPA to alkyl-LPA (60). As the non-Edg family LPA receptors (LPA4-6) were discovered more recently, these signaling pathways are not yet fully characterized (59). There have been reports of associations of LPA4 with vascular development and bone formation, whereas

LPA6 has been found to have an effect on hair growth (59).

LPA receptor 5

GPR92 was deorphaned and identified as LPA5 in 2006 (107, 108), with acceptance by the International Union of Basic and Clinical Pharmacology

2+ (IUPHAR) in 2010 (96). LPA5 has been shown to induce Ca mobilization when coupled to Gαq and induce neurite retraction when coupled to Gα12/13 (108).

Reports have also identified non-LPA ligands for LPA5, including farnesyl pyrophosphate (FPP) and N-arachidonoylglycine (NAG) (109). However, others have confirmed the identity of GPR92 as an LPA receptor by determination that the concentration required to activate LPA5 for non-LPA ligands is much greater than LPA (110, 111). Thus, the physiological relevance of non-LPA ligands as

12 activators of LPA5 remains to be shown (57). The LPA ligand preference for LPA5 is ether-linked (also known as alkyl glycerol phosphate; AGP) over ester-linked molecular species (110).

The Lpar5 gene spans 2 exons; the first exon is non-coding, thus the coding sequence is contained within exon 2 (108). In mice, Lpar5 is located on chromosome 6 and in humans, on chromosome 12 (LPAR5) (57). Lpar5 is expressed in various tissues, including the heart, small intestine, liver, and has moderate to high expression in the spleen of both mice and humans (57).

As LPA5 was discovered recently in 2006 (107, 108), the research regarding the effects of LPA5 signaling is limited. LPA5 has received considerable attention for its role in platelet activation and it has been suggested that inhibition of LPA5 signaling may be a promising anti-thrombotic therapy (110, 112, 113).

Signaling via LPA5 was shown to cause release of the inflammatory protein, MIP-

1β, by mast cells (114). Intestinal CD8+ T cells were found to have high expression of LPA5, though the role of LPA5 signaling in these cells was not elucidated (107). Recently, a group generated an LPA5-deficient mouse and discovered that these mice were protected from neuropathic pain (115).

LPA and cancer

LPA has a strong association with cancer where it is not only aberrantly expressed by a number of different malignant cell types (84-87), but has also has been shown to enhance tumor progression (116-118). Various tumors have been shown to produce aberrant levels of LPA and ascites concentrations up to 10 µM

13 have been reported in patients with certain tumors, including ovarian, cervical, and myeloma cancers (84-87, 90). High expression of ATX has also been associated with a variety of tumor cell types, including thyroid carcinoma, breast cancer, neuroblastoma, renal cell carcinoma, hepatocellular carcinoma, glioblastoma, and non-small cell lung cancer (NSCLC) (63, 116).

The elevated level of LPA has been shown to stimulate migration, survival, metastases, angiogenesis, drug resistance, and proliferation of cancer cells (119-

125). Increased LPA signaling can directly promote tumorigenesis, as it has been demonstrated that overexpression of ATX or LPA1-3 under an MMTV-LTR promoter was sufficient to induce mammary carcinomas (126). Ovarian cancer cells have been shown to upregulate LPA2, thus upregulating LPA signaling resulting in cell survival, proliferation, angiogenesis, increased migration, and tissue invasion (57). The dramatic increase in LPA2 also resulted in a lowered threshold of signaling, to low nanomolar concentrations of LPA. In a separate study, enforced expression of LPA3 increased chemotherapy drug resistance

(cisplatin and doxorubicin) in a mouse mammary tumor (127).

LPA signaling has also been implicated in migration and metastasis of cancer cells. LPA has been shown to stimulate migration in ovarian, prostate, pancreatic, and breast cancer cells (121, 123, 128, 129). Interestingly, one group found that B16 melanoma metastases were reduced in LPA5-deficient hosts, indicating that LPA5 on endogenous tissues facilitates migration of cancer cells

(Gabor Tigyi, unpublished data). Another group recently reported that LPA acts as a chemorepellent for B16 melanoma cells and that silencing of LPA5 on the

14 cancer cells ablated this effect (130). This could suggest that LPA signaling via

LPA5 causes migration from the original tumor site, where LPA levels are the highest, thus promoting metastasis.

LPA and immune cells

The reported effects of LPA signaling on different immune cell types are diverse. As before mentioned, LPA signaling via LPA5 has been reported to be responsible for mast cell release of MIP-1β (CCL4), an inflammatory protein that is a chemoattractant for monocytes, lymphocytes, and other immune cells (114).

In contrast, LPA has been reported to inhibit the cytotoxic activity of NK cells via

LPA2 (131). In the human monocytic line, THP-1, LPA induced migration in a pertussis toxin-sensitive manner (132). Intriguingly, the effect of LPA on dendritic cells (DC) has been found to be dependent on maturation status. In immature DC,

LPA induced chemotaxis however this effect was not observed in mature DC

(133, 134). LPA has also been reported to impact DC by regulating expression and function of CD1 antigen-presenting molecules (135) and inhibiting activation

(136). As a major antigen presenting cell (APC), modulation of DC function would likely also affect the T cell response.

It has been previously demonstrated that a lysophospholipid can influence adaptive immunity through studies of S1P, which is structurally similar to LPA and plays an important role in directing lymphocyte trafficking, localization and development (137-142). Studies from the past 20 years documenting the effects of LPA signaling on T cells are limited, sometimes contradictory, and have

15 focused on human primary CD4+ cells and T cell lines (57). LPA has been reported to induce calcium mobilization in cytokine-induced human Th1 and Th2 cells (143, 144) whereas in contrast, a separate study reported that LPA does not regulate calcium entry into human Jurkat T cells (145). LPA has been reported to have an inhibitory effect on IL-2 production of CD4+ cells (146) but promotes IL-

13 production (a cytokine important for asthma pathogenesis) (147). Interestingly, isolated CD4+ T cells from two human donors produced increased amounts of IL-

13 with LPA treatment with submaximal anti-CD3 stimulation (0.5 µg/mL) but not higher concentrations (1-10 µg/mL). Notably, LPA treatment did not affect IL-4 nor IFNγ production in this study.

The variability in reports of the effect of LPA signaling on T cell activation reveal that more work needs to be done as the LPA receptor(s) involved, precise mechanism and effects are not congruently elucidated (148). As LPA is readily degraded in vivo and in culture, some of the variability in reports could be attributed to this property. Sevastou et al. recently published a review that meticulously summarizes previous work and notes LPA concentration and culture conditions, further emphasizing the variability and difficulty of in vitro LPA experiments (2012).

Recently, others have investigated the effect of LPA signaling on T cell migration, given the finding that S1P has been shown to be important for lymphocyte trafficking (139, 141). As ATX is tethered to cells by integrin receptors (91, 93, 149, 150), the resultant increased LPA concentration has been shown to induce chemokinesis or chemorepulsion in T cells (132, 149, 151). This

16 coincides with data from our lab that LPA causes a small but significant degree of chemorepulsion but not chemotaxis in B cells (Figure 1.4). Injection of enzymatically inactive ATX slightly decreased homing of T cells to lymph nodes, which the authors attributed to displacement of functional endogenous ATX resulting in attenuated chemokinesis and subsequent diapedesis (149). In this study, expression of chemokine receptors CXCR4 and CCR7 were not affected by LPA treatment in vitro and the LPA receptor(s) responsible for this phenotype was not identified.

20 15 10 5

0.8

(% of B cells) * Migrated cells 0.6 0.4 0.2 0.0 Upper LPA LPA CXCL12 CXCL12

Lower LPA LPA CXCL12 CXCL12

Figure 1.4 LPA induces chemorepulsion in splenic B cells Transwell migration assay (5 µm pore size) of erythrocyte-lysed C57BL/6 splenocytes added to the upper chamber with 1 µM LPA added either to the same chamber (to induce chemokinesis/chemorepulsion), to the lower chamber (to induce chemotaxis) or to both chambers (blue bars, n=3, *p<0.05). After a 3h incubation at 37°C, B cells that had migrated to the bottom chamber were quantified by flow cytometry (B220+). Right, CXCL12 (50 ng/ml) serves as a lymphocyte chemoattractant positive control (green bars). 17 In comparison with T cells, far fewer studies have been done to examine how LPA affects the B cell response. In our lab, we have found that LPA functions to dampen antigen receptor signaling in another adaptive immune cell,

B cells. We have found that LPA inhibits B cell antigen receptor signaling via

LPA5 and that this inhibition leads to decreased activation and antibody production (152).

Summary

The adaptive immune response and CD8+ T cells in particular are tremendously important not only to protect the host from pathogens but also to combat endogenous threats, such as cancer. Lysophosphatidic acid is a pleiotropic molecule that elicits a variety of responses in different cell types and has been shown to be released at increased concentrations as a result of various tumors, however the effect of LPA signaling on antigen receptor signaling of

CD8+ T cells remains to be addressed. We have found previously in B cells that

LPA functions to inhibit antigen receptor signaling, activation, effector function, and induces chemorepulsion (unpublished observations). Given the signaling similarities between B and T cells, we questioned if this inhibition of antigen receptor signaling also occurred with T cells. We hypothesized that increased

LPA signaling will similarly dampen the CD8+ T cell response and questioned if this inhibition could be an additional inhibitory mechanism featured in the tumor microenvironment. The work presented here shows that LPA signaling inhibits

+ CD8 T cell activation and proliferation via LPA5 and that LPA more potently inhibits activation to a lower affinity peptide. In addition, we show that ablation of

18 + LPA5 signaling on tumor-specific CD8 T cells allows better control of tumor progression and suggest that this may represent a new therapeutic target to enhance the adaptive immune response to tumor.

CHAPTER II

MATERIALS AND METHODS

Mice

Specific pathogen-free, C57BL/6 (CD45.2) and CD45.1 (B6.SJL-Ptprca

Pepcb/BoyJ) were obtained from Jackson Laboratories or bred in the Biological

Resource Center (BRC) at National Jewish Health (NJH). CD45.1 OT-I mice

–/– (153)(gift of Dr. Ross Kedl, University of Colorado), LPA2 mice (154)(gift of Dr.

–/– Jerold Chun, Scripps Research Institute), LPA5 mice (gift of Lexicon

Pharmaceuticals), and TCRα–/– mice (gift of Dr. Philippa Marrack, National

Jewish Health) were bred in the BRC at NJH. All mice used were 8-12 weeks of age and were maintained and treated in accordance with the regulations of the

Institutional Animal Care and Use Committee.

OT-I and CD45.1 mice were typed by flow cytometry. Briefly, peripheral blood was collected and lysed by water, then neutralized with 10X HBSS. The remaining cells were pelleted and stained as described below. OT-I mice were considered >90% Vα2+Vβ5+ and >90% CD8+ of Thy1.2+.

–/– LPA5 mice were generated by targeted replacement of exon 1 with a

LacZ/neo cassette (Figure 2.1A). Deletion was confirmed by PCR and southern blots at Lexicon Pharmaceuticals (Figure 2.B,C). Heterozygous crosses resulted in Mendelian ratios of pups with no obvious phenotypic differences (Figure 2.1D

–/– and data not shown). LPA5 mice were generated on a mixed 129/SvEvBrd and

20 C57BL/6J genetic background and were continually backcrossed with C57BL/6

–/– mice. LPA5 mice were backcrossed 2-6 generations before analysis together

+/+ with gender-matched LPA5 littermates as controls. LPA5 mice were genotyped by PCR with the following primers (Fig 2.1):

Neo-sense GCAGCGCATCGCCTTCTATC,

9-sense CCAACTGCGTGCTGGATC,

22-antisense CATACTGGAGCAGGGTTTC.

5’ internal probe 3’ external probe A B PCR XmaI EcoRI ATG Stop XmaI EcoRI bp 1 kb 600 400 9 22 200 pKOS58 (target vector) neo +/+ +/ +/+ / PCR primers LacZ/neo Wild type 454 bp XmaI EcoRI Mutant 297 bp

C Southern blots D kb 5’ internal probe kb 3’ external probe (20111020 updated) LPA5 typing 12 50 12 Wildtype 10 44% 10 Mutant 8 40 8 30 6 31% 20 25%

6 % of total 10 XmaI digests 4 EcoRI digests Wild type 13.2 kb 0 Wild type 8.5 kb t Mutant 10.5 kb O T Mutant 4.4 kb KO K Het He WW T

–/– Figure 2.1 Generation of LPA5 mice (A) Schematic diagram of Lpar5 targeted deletion. A portion of the Lpar5 gene was replaced with a LacZ/neo selection cassette. (B) PCR genotyping displaying wild-type and mutant Lpar5. Primers are indicated in (A) as “9”, “22”, and “neo”. (C) Southern hybridization indicating the proper targeting event in ESC clones. Restriction sites and probe locations are indicated in (A). –/– +/– (D) The proportion of pups that were LPA5 (KO), LPA5 (Het), or +/+ LPA5 (WT) resultant from heterozygous crosses (n=106).

21 Calcium mobilization

Mouse cells were erythrocyte-lysed with ACK lysis buffer and splenocytes, pooled lymph node cells, or purified CD8+ T cells were suspended at 20 x 106 cells/mL in RPMI 1640 medium (Fisher Scientific) supplemented with 2.5% fatty- acid free (faf) BSA (Calbiochem), to minimize background levels of extracellular

LPA. Indo1-AM was added to a final concentration of 5 µM and incubated for 30 minutes at 37°C. Antibodies for extracellular staining were added for the final 20 minutes. Cells were washed and resuspended at 4-6 x 106 cells/mL before analysis on the LSRII. For data acquisition, a baseline reading was measured for

30 sec, then the sample was removed from the instrument and LPA or vehicle control was added immediately and AgR stimulation (for B cells, 20 µg/mL anti-

IgM F(ab’)2 [goat anti-mouse, Jackson ImmunoResearch Laboratories]; for T cells, 10 µg/mL anti-CD3e biotin pretreatment [145-2c11, BD Biosciences], ligated by 20 µg/mL avidin [Invitrogen]) was added at 39 sec (155). The sample was returned to the machine at 45 sec and the calcium mobilization was recorded for an additional 4-9 minutes, depending on the cell type and experiment. For some experiments, calcium mobilization from intracellular stores was analyzed by addition of 4 mM EGTA (152).

To determine how LPA is affected by culture with various media additives,

16:0 ester- or ether-LPA was used cultured at a high concentration (420 µM) in culture media (IMDM) so that it could be subsequently used at 20 µM in a calcium mobilization assay. LPA was cultured in media alone or media supplemented with 5% fatty acid-free BSA, 5% FBS, or 50 µM β-

22 mercaptoethanol for 3 or 6 hours. The supernatant was then collected and aliquots of the supernatant were used as a source of LPA at a final concentration of 20 µM LPA (based on original LPA concentration) for calcium mobilization assays with C57BL/6 splenic B cells stimulated with anti-IgM. As a control, cells were treated with media cultured in the absence of LPA in place of cultured LPA.

qPCR

Erythrocyte-lysed splenocytes and pooled LN cells were sorted for

Thy1+CD8+CD4- cells using the MoFlo XDP sorter (Beckman Coulter) to >85% purity. RNA was isolated using TRIzol (InvitrogenLife Technologies) and further purified of DNA with a DNA-free kit (Ambion). cDNA was prepared from equivalent amounts of RNA using a SuperScript III First-Strand Synthesis

System for RT-PCR (Invitrogen Life Technologies). Quantitative PCR amplification was performed using Platinum SYBR Green qPCR SuperMix-UDG

(Invitrogen Life Technologies) and detected on an MJ Research DNA Engine

Opticon 2 real-time PCR machine.

Primers for the Lpar1-3,6 receptors and Hprt were designed and then purchased (Integrated DNA Technologies) with the following sequences:

Lpar1–sense CTGTGGTCATTGTGCTTGGTG,

Lpar1–antisense CATTAGGGTTCTCGTTGCGC,

Lpar2–forward GGCTGCACTGGGTCTGGG,

Lpar2–sense GCTGACGTGCTCCGCCAT,

23 Lpar3–sense GCGCACAGGAATGGGAGAG,

Lpar3–antisense GAGCTGGAGGATGTTGGGAG,

Lpar6-sense TCTGGCAATTGTCTACCCATT,

Lpar6-antisense TCAAAGCAGGCTTCTGAGG,

Hprt-sense CTGGTGAAAAGGACCTCTCG,

Hprt-antisense TGAAGTACTCATTATAGTCAAGGGCA.

Primers from Lpar4 and Lpar5 were purchased (Super Array Biosciences).

The control for Lpar1-6 expression was thymus tissue. qPCR for Lpar6 was run separately, therefore expression was normalized to Lpar5 expression. The cycle threshold (Ct) was determined for each experiment based on background from control samples (no cDNA) and expression of Lpar mRNA of interest was normalized to Hprt mRNA.

Flow cytometry

Cells were stained in 2% BSA-PBS+0.1% sodium azide with blocking Fc receptor antibody (2.4G2) on ice for 20-30 minutes. For live/dead assessment, the viability dye, 7-AAD, was added 10 minutes prior to data acquisition. All flow cytometric analysis was performed on the LSRII flow cytometer (BD) and analyzed with FlowJo v8 (Tree Star) and GraphPad Prism software (v 5.0).

The staining panel for analysis of proliferation by CFSE dilution was as follows: FcR block (2.4G2, homemade), CD8-eFluor450 (53-6.7, eBioscience),

CFSE (CFDA, SE, Invitrogen), 7-AAD (eBioscience), CD25-APC (PC61.5, eBioscience), CD45.1 PE-Cy7 (A20, eBioscience), CD44 APC-eFluor780 (IM7,

24 eBioscience), and CD45.2-Alexa Fluor 700 (104, eBioscience) if needed to distinguish CD45.1+/+ from CD45.1+/-, as for in vivo cytotoxicity assays.

The staining panel for analysis of intracellular cytokine production was as follows: FcR block, CD8-eFluor450, TNFα-FITC (MP6-XT22, eBioscience), IL-2-

PE (JES6-5H4, eBioscience), and IFNγ-APC (XMG1.2, eBioscience) or Lamp-1

(eBio1D4B, eBioscience) and Granzyme B-Alexa Fluor 647 (16G6, eBioscience).

The staining panel for OT-I typing was as follows: FcR block, CD8- eFluor450, CD45.2-AF488 (104, Biolegend), Vβ5-PE (MR9-4, BD), Thy1.2-

PerCP (53-2.1, BioLegend), Vα2-APC (B20.1, eBioscience), and CD45.1 PE-Cy7

(A20, eBioscience).

Additional flow cytometry antibodies used include the following: CD8-APC- eFluor780 (53-6.7, eBioscience), CD8-APC (53.6.7, BD), Thy1.2-APC (53-2.1, eBioscience), CD4-FITC (RM4-4, BD), CD62L-FITC (MEL-14, eBioscience),

CD127-PE (A7R34, eBioscience), CCR7-PE (4B12, eBioscience), KLRG1-FITC

(2F1, eBioscience), PD-1-FITC (29F.1A12, BioLegend), MHCII-Fluos (M5-114, homemade), CD86-APC (GL1, eBioscience), anti-SIINFEKL in H-2Kb (25-D1.16, gift of John Kappler), and CD3-FITC (145-2c11, eBioscience).

Lipid preparation

Ether-LPA (16:0 AGP, Avanti Polar Lipids) and ester-LPA (16:0, Avanti

Polar Lipids) were solubilized to a 5 mM concentration in 0.1% BSA-PBS, aliquoted, and stored at -20°C. Aliquots were diluted to 1 mM in RPMI 1640 medium supplemented with 2.5% charcoal-stripped fatty acid-free BSA (faf-BSA

25 [Calbiochem]) prior to use in calcium mobilization assays. Faf-BSA still contains low levels of lipids that were stripped with charcoal. Charcoal-stripped faf-BSA was prepared by incubation of 10 mL of 1 mM faf-BSA in PBS with 1 mg of activated charcoal (Sigma) overnight at 4°C. Charcoal was removed by centrifugation and sterile filtration.

Ether- and ester-LPA were compared in a preliminary experiment in B cells to determine efficiency of inhibition of calcium mobilization, given that B cells express LPA5, which is preferentially stimulated by ether-LPA over ester-

LPA (110, 156). Ether-LPA was found to have a greater inhibition of calcium mobilization in splenic B cells (Figure 2.2), consistent with the ability of LPA5 to be more potently stimulated by ether-LPA over ester-LPA.

3.5 3 2.5 2

Indo ratio 1.5 1

0 50 100 150 200 Time (s)

Figure 2.2 Inhibition of calcium mobilization by ether- versus ester-LPA Intracellular calcium mobilization was assessed in splenic B cells stimulated with 10 µg/mL anti-IgM in the absence (black line) or presence of 20 µM 16:0 ester-LPA (red line) or ether-LPA (blue line).

OTP was generated as previously described (157), stored as a powder, and solubilized in 95% methanol to create aliquots. Virgin glass tubes and caps

26 were sterilized by autoclave and aliquots of 0.25 to 1µmol were prepared with glass pipettes (lipids readily adhere to plastic). Aliquots were allowed to desiccate sterilely to a film in a tissue culture hood. The concentration was confirmed by phosphorus assay (158).

OTP was solubilized for experimental use by sonication with FBS- or charcoal-stripped faf-BSA-containing culture media or vehicle (2% propanediol,

1% ethanol in PBS) for in vitro or in vivo usage, respectively. For in vitro experiments, OTP was prepared at 50 µM and passed through a 0.2 µm filter for further sterilization. For in vivo experimentation, solubilized OTP was transferred to siliconized eppendorf tubes and animals were dosed at 5 mg/kg every 8 hours.

Transwell assay

C57BL/6 splenocytes (106 cells) were added to the upper chambers of

Transwells (pore size, 5 µm; Costar) and were allowed to migrate for 3 h at

37°C. For analysis of the chemokinetic response of T cells, cells were incubated with 1 µM LPA in the upper, lower, or both chambers. CXCL12 (50 ng/ml) was used as a lymphocyte chemoattractant positive control. Migrated cells were quantified by flow cytometry for B (B220+) and T cells (CD3+).

In vitro T cell proliferation and activation assays

To determine if LPA affected activation of wild-type CD8+ T cells, C57BL/6 splenocytes were erythrocyte-lysed and stimulated with 0, 1, or 10 µg/mL plate- bound anti-CD3 and 0 or 1 µg/mL soluble anti-CD28 in 5% BSA RPMI in the

27 presence of 0, 1, or 10 µM LPA for 24 hours. Cells were harvested and stained for flow cytometry.

OT-I splenocytes were isolated, erythrocyte-lysed, and labeled with CFSE

(Invitrogen). For all CFSE labeling, cells were suspended at 15 x 106 / mL in PBS and CFSE was added to a final concentration of 2.5 µM for 10 minutes and then excess CFSE was removed by media washes. Splenocytes were pulsed with 1

µM of the SIIGFEKL (G4, Anaspec, Inc.) or SIINFEKL (gift of Pippa Marrack) peptide for 4 hours or 90 minutes, respectively, in 5% faf-BSA RPMI with or without 50 µM OTP, then washed. Cells were plated at 2.5 x 106 cells/mL with or without 50 µM OTP in 96 well plates. Cells were enumerated by flow cytometry by counting gated events for the first minute of acquisition after event rate had stabilized. The proportion of cells that had proliferated was calculated by Flowjo analysis. The mean fluorescent intensity (MFI) values of activation marker expression were normalized to account for variations between experiments.

To determine if antigen presentation was affected by LPA signaling, splenocytes from a T cell-deficient mouse (TCRα–/–) were pulsed with 1 µM

SIINFEKL peptide in the presence or absence of OTP, washed, and used to stimulate purified CD8+ OT-I T cells (2:1, splenocytes:T cell) in culture in the presence or absence of OTP.

In vitro effector function assays

To assess in vitro cytokine production, OT-I effector T cells were generated in vitro by pulsing erythrocyte-lysed OT-I splenocytes with 1 µM

28 SIINFEKL and culture with IL-2 for 5 days. On day 5 of culture, target cells (EL4 cells) were pulsed with 1 µM SIINFEKL and cultured at an effector to target ratio of 0.625:1 with OT-I effector T cells for 4 hours in the presence of Brefeldin A, in the presence or absence of 50 µM OTP.

To compare in vitro cytokine production of LPA5-sufficient and LPA5-

+ +/+ –/– deficient T cells, CD8 T cells were purified from LPA5 or LPA5 OT-I splenocytes and lymph nodes, and combined at a 1:2 ratio with C57BL/6 splenocytes before a 4 hour pulse with SIIGFEKL (G4) peptide. Cells were washed and plated at 2.5 x 106 cells/mL. On day 3, cells were transferred to a round-bottom plate and cultured for 5 hours in the presence of brefeldin A. Cells were fixed, permeablized, and stained for flow cytometry.

To assess in vitro cytotoxic ability, OT-I effector T cells were generated as described before. Target EL4 cells were fluorescently labeled CFSEhigh with 2.5

µM CFSE and pulsed with 10-100 nM SIINFEKL. Control EL4 cells were not pulsed with peptide and fluorescently labeled CFSElow by diluting the CFSE stain

9-fold. CFSEhigh and CFSElow EL4 cells were mixed at a 1:1 ratio and combined with OT-I effector T cells at various effector to target (E:T) ratios. As an alternative method to assess cytotoxicity, target EL4 cells were not stained with

CFSE and were pulsed with 10 nM SIINFEKL and combined with OT-I effector T cells at various effector to target (E:T) ratios. Cell death was assessed by Non-

Radioactive Cytotoxicity Assay (LDH ELISA, CytoTox 96, Promega) and adjusted for background lysis of OT-I T cells co-cultured with EL4 cells without SIINFEKL peptide (159).

29 Generation of bone marrow-derived dendritic cells

Gender-matched bone marrow-derived dendritic cells (BMDC) were generated by flushing of femur and tibia and culture at 106 cells/mL in RPMI 1640 supplemented with 20 ng/mL GM-CSF (gift from Dr. Ross Kedl), 10% FBS

(Omega Scientific), Penicillin-Streptomycin (Invitrogen), and GlutaMAX

(Invitrogen). Media was refreshed on days 2 and 4 of culture. On day 7, BMDC were harvested from culture and stimulated with 1 ng/mL LPS for 90 minutes and pulsed with peptide for the last hour of LPS treatment. BMDC were washed 5 times with media to remove LPS and peptide before use.

In vivo SIINFEKL-BMDC stimulation

Bone marrow derived dendritic cells were generated as described in section 2.8. One day prior to immunization, CD8+ T cells were purified from OT-I spleen and LN cells with a CD8+ magnetic bead enrichment kit (Isolation Kit II,

Miltenyi) to a purity of ≥95%, and 106 CFSE-labeled CD8+ T cells were transferred to CD45 allotype mismatched recipient C57BL/6 mice. BMDC (106) were suspended in PBS and transferred by sub-cutaneous (s.c.) injection in the scruff to individual recipients. On d3 post-immunization, animals were sacrificed and dLN (axilary, brachial, cervical), ndLN (inguinal, mesenteric), and spleen were harvested. After erythrocyte lysis, cells were counted by Z2 Coulter Particle

Count and Size Analyzer (Beckman-Coulter) and 10 x 106 cells were stained for flow cytometry. Cells were suspended in FACS buffer and stained with 7-AAD for viability before analysis on the BD LSRII.

30 IL-2 ELISA

To determine IL-2 production of LPA5-sufficient versus LPA5-deficient OT-I,

CD8+ T cells were purified from pooled splenocytes and lymph nodes from

+/+ –/– 5 + LPA5 or LPA5 OT-I mice. 10 CD8 T cells were cultured with SIINFEKL- pulsed BMDC at a 2:1 or 10:1 ratio. Supernatants were collected at the indicated timepoints and assayed for IL-2 by ELISA. Briefly, wells were coated with anti-IL-

2 (JES6-1A12, eBioscience) and samples were detected with biotinylated anti-IL-

2 (JES6-5H4, BD) and AP-avidin (Southern Biotech).

To determine IL-2 consumption of LPA5-sufficient versus LPA5-deficient

OT-I, an in vitro experiment was conducted similarly, except that 15 ng/mL rmIL-

2 was added to culture.

B16.cOVA tumor transfer

The OVA-transfected B16 tumor cell line (B16-OVA) was kindly provided by Dr. Ross Kedl. Cells were maintained in RPMI 1640 (Cellgro), supplemented with 10% heat-inactivated FBS (Omega Scientific), GlutaMAX, Penicillin-

Streptomycin, MEM NEAA, sodium pyruvate (Invitrogen), and 0.75 mg/mL G418 sulfate selection (Cellgro). Cells were passaged after trypsin digestion every 2 days.

–/– 5 To determine how LPA5 OT-I T cells responded to tumor, 5 x 10

B16.cOVA cells were transferred s.c. into the hind leg of recipients and 5 days later 106 CFSE-labeled OT-I T cells were transferred s.c. into the tail vein.

Recipients were sacrificed 5 days later and tumor and dLN were harvested.

31 Lymphoid organs were mashed through a cell strainer (100 µm, BD), erythrocyte- lysed, and counted. Tumors were digested in 0.5 mg/mL Collagenase D (Fisher

Scientific) and 60 U/mL DNase (Sigma-Aldrich) for 30 minutes with perturbation every 10 minutes, then neutralized with 5 mM EDTA for 5 minutes. Tumor cells were homogenized by pipette and passed through a 100 µm filter, erythrocyte lysed, and counted.

To measure OT-I CD8+ T cell control of tumor progression, 105 B16.cOVA cells were delivered s.c. into the leg of a C57BL/6 recipient mouse. On d5, 0.5 x

106 OT-I T cells were transferred i.v. via tail vein injection. Tumor diameter was measured with calipers and body weight and overall appearance were assessed every 2 days. Mice were sacrificed when tumor diameter exceeded 10 mm.

In vivo cytotoxicity assay with tumor burden

C57BL/6 recipients were implanted s.c. in the leg with 0.5 x 105 B16.cOVA

6 +/+ –/– cells. Five days post-implantation, 10 CFSE-labeled LPA5 or LPA5 OT-I

CD8+ T cells were transferred i.v. C57BL/6 splenocytes were divided to two populations and one population was labeled CFSEhigh and pulsed with SIINFEKL and the second population was labeled CFSElow and not pulsed with peptide. The

CFSEhigh and CFSElow cells were combined at a 1:1 ratio and injected i.v. 1 day before harvest. Tumors and dLNs were harvested as before described and the ratio of CFSEhigh and CFSElow cells was assessed (160).

32 Statistical analyses

All statistical analyses were performed with FlowJo (v8, Tree Star) or

GraphPad Prism software (v 5.0) using a two-tailed unpaired Student’s t test or

Mantel-Cox log rank test.

CHAPTER III

LPA SIGNALING INHIBITS CD8+ T CELL ACTIVATION AND

PROLIFERATION VIA LPA5

Introduction

Lysophosphatidic acid (LPA) is a bioactive lysophospholipid found in all eukaryotic tissues and in blood (57). Lysophospholipids, including LPA and sphingosine 1-phosphate (S1P), induce a diversity of biological and pathophysiological effects by binding and activating specific G-protein coupled receptors (GPCR) (63, 87, 103). Similar to S1P, the LPA lysophospholipid is recognized by a set of GPCR that are differentially expressed by different cell types and that are divided into two groups based on similarity: LPA1,2,3 belong to the Edg family of receptors whereas LPA4,5,6 belong to a purinergic-like family of receptors (60, 63). While S1P is now recognized to play an important role in orchestrating lymphocyte development, trafficking and function (137-142), the immune regulatory activities of LPA have been less studied despite the finding that many immune cells express LPA receptors (57, 87, 103).

LPA is primarily generated in vivo via an extracellular lysophospholipase D, autotaxin (ATX), which has been proposed to associate with the cell surface and directly release LPA to receptors, thus efficiently initiating LPA signaling to that cell (91-94). LPA levels in healthy individuals is the result of synthesis by ATX and degradation of LPA by lipid phosphate phosphatases (81-83) and

34 physiological levels of LPA are maintained in the hundred nanomolar range in plasma (62, 63). However, elevated LPA levels in the micromolar range have been reported in certain pathological conditions, such as inflammatory disease and in individuals with certain tumors (161) and given the expression of LPA receptors on immune cells, we sought to determine how the increase in LPA may affect the adaptive immune response.

T cell activation is induced by TCR signaling initiated by the appropriate

MHC-peptide complex on an APC (10, 11). This activation can be modulated by

TCR co-receptors in a positive (e.g. CD28, ICOS) or negative manner (e.g.

CTLA-4, PD-1) (13-15, 17). In previous studies of B cells in our lab, we found that

LPA signaling dampened calcium mobilization in primary B cells and B cell lines

(152, 162). Thus, we questioned if LPA signaling could be an additional mechanism that modulates TCR signaling and activation.

We have further found that retroviral shRNA knock-down of LPA5 dampened inhibition of LPA (152, 162). We substantiate this finding by showing that LPA5-deficient B and T cells are insensitive to LPA inhibition. Additionally, our lab found that LPA inhibited B cell activation and dampened antibody production in vivo (152). Given that LPA more potently inhibited calcium mobilization in CD8+ T cells than in B cells, we hypothesized that this inhibition of activation would be more pronounced in CD8+ T cells. Our data demonstrate that LPA signaling can profoundly inhibit T cell activation and proliferation in response to peptide stimulation and that this inhibition is stronger with a weaker affinity peptide. Finally, we show that these findings translate to in vivo relevance,

35 as treatment with an LPA analog results in inhibition of CD8+ T cell activation in response to stimulation with peptide-loaded dendritic cells.

Results

LPA signaling inhibits TCR-mediated calcium mobilization in CD8+ T cells via LPA5

By measuring intracellular calcium mobilization by flow cytometry we can determine the immediate effects of LPA on activation simultaneously in different cell types. We previously found that LPA signaling dampens B cell calcium mobilization in response to BCR signaling (152) and that ether-LPA was a more potent inhibitor than ester-LPA (Figure 2.2). Given the many similarities in antigen receptor signaling by B and T lymphocytes, we questioned if LPA similarly inhibited TCR signaling. LPA is present in plasma in the nanomolar range however is found at micromolar concentrations in serum and in the ascites fluid of patients with certain cancers (63, 117, 163). Therefore, we tested the influence of 1-20 µM concentrations of 16:0 ether-LPA on TCR-mediated intracellular calcium mobilization. The CD3 molecule is a component of the T cell receptor complex and T cells can be activated by engaging CD3 molecules with anti-CD3 antibody and subsequent crosslinking with an anti-IgG antibody or streptavidin (the latter with biotinylated anti-CD3) (155). The results from these experiments revealed that LPA inhibited intracellular calcium mobilization not only in marginal zone and follicular B cells, as we had previously shown, but also in both CD4+ and CD8+ T cells in response to TCR stimulation achieved by anti-

CD3 monoclonal antibody (mAb) crosslinking (Figure 3.1). LPA treatment in the

36 absence of antigen receptor stimulation did not affect calcium mobilization.

Interestingly, when inhibition of calcium mobilization of similar magnitude was

compared across lymphocyte subsets, we found that LPA inhibition appeared

most pronounced in CD8+ T cells (Figure 3.1). We therefore decided to elucidate

the mechanism of LPA-mediated dampening of calcium mobilization in CD8+ T

cells and used 16:0 ether-LPA for experimentation.

3 3 2.5 FO MZ 2 2 B cells 1.5 1 1 Indo ratio 0.5 AgR stim AgR stim + 20 M LPA 0 50 100 150 200 250 0 50 100 150 200 250 µ 3 2.5 20µM LPA 2.5 CD4 CD8 2 T cells 2 1.5 1.5Indo2/Indo1 Indo2/Indo1

3 3 Indo ratio 1 1 2.5 FO MZ 0 100 200 300 400 0 100 200 300 400 2 2 B cells 1.5 Time (s) Time (s) 1 1 Indo ratio 0.5 AgR stim AgR stim + 20 M LPA 0 50 100 150 200 250 0 50 100 150 200 250 µ 3 2.5 20µM LPA 2.5 CD4 CD8 2 T cells 2 1.5 1.5Indo2/Indo1 Indo2/Indo1 Figure 3.1 LPA dampens calcium mobilization in major

Indo ratio 1 1 splenic lymphocyte populations 0 100 200 300 400 0 100Intracellular200 300 calcium400 mobilization was assessed in splenic B (marginal Time (s) zone,Time (s) MZ; follicular, FO) and T cells stimulated with 10 µg/mL anti-IgM or 10 µg/mL anti-CD3, respectively, with or without 20 µM 16:0 ether-LPA (thin and bold lines, respectively). 20 µM LPA treatment in the absence of antigen receptor (AgR) stimulation is shown in gray lines. Representative of at least 5 independent experiments.

To determine if LPA inhibition of calcium mobilization occurred directly on

CD8+ T cells, we repeated the calcium mobilization assay with purified CD8+ T

cells and found that LPA inhibition was cell-intrinsic (Figure 3.2). We titrated the

37 amount of LPA used in the calcium mobilization assay to determine the sensitivity of inhibition and found that inhibition was achieved with all concentrations tested, was dose dependent, and inhibition was sensitive to LPA concentrations as low as 1µM (Figure 3.2).

3 αCD3 αCD3 + 1µM LPA 2 Indo ratioIndo ratio αCD3 + 5µM LPA

Indo ratio 1 αCD3 + 20µM LPA 0 100 200 300 400 500 Time (s)

Figure 3.2 LPA inhibition of CD8+ T cell calcium mobilization is dose dependent Intracellular calcium mobilization was assessed in purified splenic and pooled lymph node CD8+ T cells stimulated with 10 µg/mL anti-CD3 in the absence (black line) or presence of the indicated concentrations of LPA. Representative of 4 independent experiments.

In B cells, we have shown that LPA inhibition of BCR signaling manifests by reducing intracellular calcium stores release (152). To determine if LPA signaling similarly affected intracellular calcium release after TCR stimulation, purified CD8+ T cells were stimulated in the presence of EGTA to selectively monitor release of calcium from intracellular stores. These results revealed that in the absence of extracellular calcium, LPA maintained ability to inhibit an increase in intracellular calcium levels, even at the lowest LPA concentration tested (1µM, Figure 3.3). These results suggest that LPA inhibits TCR-induced intracellular calcium stores release, similar to that observed with B lymphocytes

(152).

EGTA 2 αCD3 1.5 αCD3 + 1µM LPA 1 αCD3 + 5µM LPA

Indo ratio αCD3 + 20µM LPA 0.5 0 200 400 Time (s)

Figure 3.3 LPA treatment dampens calcium release from intracellular calcium stores Intracellular calcium mobilization was assessed in the presence of 4 mM EGTA to prevent extracellular influx. Purified splenic and pooled lymph node CD8+ T cells were stimulated with 10 µg/mL anti-CD3 in the absence (black line) or presence of the indicated concentrations of LPA. Representative of 2 independent experiments.

We next wanted to determine the LPA receptor(s) responsible for this inhibition by first measuring LPA receptor expression on CD8+ T cells. There are currently six described LPA receptors: LPA1-6 (96). Flow cytometric compatible antibodies to the extracellular region of murine LPA receptors are not commercially available therefore we evaluated LPA receptor expression by quantitative RT-PCR (qRT-PCR). These results revealed that purified CD8+ T cells express LPA receptors 2, 5, and 6 (Figure 3.4), similar to B cells (152) and previous findings (107, 164).

39 1,000 CD8+ PCR control 100

10 Relative expression 1 Lpar : 1 2 3 4 5 6

Figure 3.4 CD8+ T cells express Lpar1, 5, and 6 Expression of Lpar1-6 in purified splenic and lymph node CD8+ T cells as measured by qRT-PCR. PCR control is thymus tissue. Representative of 1-3 independent experiments.

It had previously been shown in our lab that LPA5 knockdown by retroviral shRNA resulted in the A20 B cell line reduced sensitivity to LPA inhibition (152).

To determine which of these LPA receptor(s) was responsible for LPA inhibition

–/– in T cells, we compared calcium mobilization in splenocytes from LPA2 and

–/– –/– LPA5 mice (LPA6 mice were not commercially available). We found that

+ although LPA inhibition was intact in the LPA2-deficient CD8 T cells, this

+ inhibition was ablated in the LPA5-deficient CD8 T cells, demonstrating that

LPA5 is required for LPA inhibition of TCR-mediated calcium mobilization (Figure

3.5).

/ 5 LPA 4 / 2 LPA 4 3 5 3 αCD3 Indo2/1 2Indo ratio 2 αCD3 + 20µM LPA Indo ratio 1 1 0 100 200 300 0 100 200 300 400

/ Time (s) 5 LPA 4 / 2 LPA 4 3 5 3 αCD3 Indo2/1 2Indo ratio 2 αCD3 + 20µM LPA

Indo ratio 1 1 0 100 200 300 0 Figure100 3.5200 300Inhibition400 of calcium mobilization by LPA is –/– + Time (s) ablated in LPA5 CD8 T cells –/– –/– Intracellular calcium mobilization was assessed in LPA2 or LPA5 splenic CD8+ T cells in response to 10 µg/mL anti-CD3 stimulation in the presence (bold line) or absence (thin line) of 20 µM LPA. Representative of 2 independent experiments

+ + LPA5-deficient B cells (B220 ) and CD4 T cells were also resistant to LPA

inhibition of calcium mobilization (Figure 3.6A, right panels). Interestingly, LPA5

heterozygous mice had an intermediate phenotype of calcium inhibition by LPA

for all three cell types tested (Figure 3.6A, center panels). This intermediate

phenotype is emphasized when the 20 µM treated calcium traces (bold lines in

Figure 3.6A) were overlaid (Figure 3.6B). This indicates that a single copy of the

Lpar5 gene is sufficient to transmit inhibitory signaling, though not to the same

+/+ extent as the LPA5 homozygote. Further this suggests that a single copy of

Lpar5 results in reduced surface expression of the LPA5 protein. Since we

observed an intermediate phenotype for the heterozygous animals, we compared

–/– LPA5 mice with gender-matched littermate wild-type mice in all future

–/– experiments using LPA5 mice.

A LPA5+/+ LPA5+/ LPA5/ 4 5 4 CD8+ 4 3 3 3 2 Indo ratio 2 2 1 1 1 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400

4 4 CD4+ 3 3 3 2 Indo ratio2 Indo ratio2 1 1 1 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400

2.2 B220+ 2.5 2 1.9 2 1.5 1.6 1.3Indo ratio Indo ratio1.5 1 1 1 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400

Indo ratio Time (s) αAgR αAgR + 20µM LPA B 4 CD8+ CD4+ 2.5 B220+ 3 3 2

2Indo ratio 2 Indo ratio Indo ratio 1.5 1 1 1 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400

Indo ratio Time (s) LPA +/+ LPA +/ LPA / 5 5 5

–/– Figure 3.6 LPA5 T and B cells are resistant to LPA +/– inhibition whereas LPA5 lymphocytes exhibit an intermediate phenotype +/+ (A) Intracellular calcium mobilization was assessed in LPA5 (left column), +/– –/– LPA5 (center column), and LPA5 (right column) splenocytes as gated on CD8+ T cells (upper row), CD4+ T cells (center row), and B cells (B220+, bottom row), in response to AgR stimulation (20 µg/mL anti-IgM, 10 µg/mL anti-CD3) in the presence (bold line) or absence (thin line) of 20 µM LPA. Representative of 2 independent experiments. +/+ (B) Intracellular calcium mobilization was compared between LPA5 (red +/– –/– line), LPA5 (blue line), and LPA5 (green line) splenocytes as gated on CD8+ T cells (left), CD4+ T cells (center), and B cells (B220+, right), in response to AgR stimulation in the presence of 20 µM LPA. Representative of 2 independent experiments.

42 LPA does not influence chemotaxis in naïve T cells

We had previously shown that LPA induced chemorepulsion in splenic B cells (Figure 1.4) and LPA has been shown to influence chemokinesis in T cells

(132, 149, 151). We questioned whether LPA had an effect on chemotaxis of splenic T cells, similar to previously published. To address this question, we performed a Transwell assay with C57BL/6 splenocytes and tested whether the presence of LPA in the top chamber, bottom chamber, or both chambers influenced migration of naïve T cells (Figure 3.7). In contrast to our findings with

B cells, the presence of 1 µM LPA in either the top or bottom chambers did not significantly influence T cell migration. This indicates that a low concentration of

LPA does not cause chemotaxis or chemorepulsion in naïve splenic T cells.

However we did not discriminate between CD4+ and CD8+ T cells in these experiments and cannot exclude that there may be a differential effect between T cell subsets. These data likely represent primarily CD4+ T cells, as CD4+ T cells outnumber CD8+ T cells in the spleen of 8-12 week old C57BL/6 mice (165). In addition, we cannot exclude that T cells may migrate differently to LPA when activated or differentiated to an effector T cell phenotype, or that increased levels of LPA may differently influence T cell migration. Altered chemotaxis to LPA as determined by activation state has been shown in other cell types, as immature

DC are responsive to LPA chemotaxis however mature DC downregulate the

LPA receptor and are no longer responsive (133, 134).

43 15

1.0 (% of T cells) Migrated cells 0.5

0.0 Upper LPA LPA CXCL12 CXCL12

Lower LPA LPA CXCL12 CXCL12

Figure 3.7 LPA does not influence chemotaxis of naïve splenic T cells Transwell migration assay (5 µm pore size) of erythrocyte-lysed C57BL/6 splenocytes added to the upper chamber with 1 µM LPA added either to the same chamber (to induce chemokinesis/chemorepulsion), to the lower chamber (to induce chemotaxis) or to both chambers (blue bars, n=3). After a 3h incubation at 37°C, T cells that had migrated to the bottom chamber were quantified by flow cytometry (CD3+). Right, CXCL12 (50 ng/ml) serves as a lymphocyte chemoattractant positive control (green bars).

Increased LPA signaling inhibits proliferation, accumulation, and activation of

OT-I T cells in vitro

Intracellular calcium is a critical downstream effector of antigen receptor signaling that leads to the activation of distinct transcriptional programs important for T cell activation and function (166, 167). Thus, we next wanted to assess if

LPA dampening of TCR-induced calcium mobilization would lead to diminished

CD8+ T cell activation in vitro. However, it has been reported that LPA is readily

44 degraded both in vitro and in vivo with a reported half-life as short as 3 minutes in the blood (66, 168), thus before using LPA in culture, we wanted to determine if our various media additives would contribute to LPA degradation. To test this, we cultured LPA at a high concentration (420 µM) with various media supplements such as fatty acid-free BSA (faf-BSA) and FBS, for 6 hours in the presence or absence of splenocytes. We then tested LPA for degradation by using the supernatant as the source of LPA at a calculated concentration of 20 µM in a calcium mobilization assay with splenic B cells and compared inhibition by cultured LPA versus stock LPA (Figure 3.8). We also compared the stability of the more commonly used ester-LPA with ether-LPA, as ether-LPA is reported to be more stable and LPA5 is preferentially engaged by ether over ester moieties of LPA (110, 169). We found that both LPA species incubated in faf-BSA maintained the ability to inhibit calcium mobilization, comparably to freshly thawed LPA (Figure 3.8). However, LPA incubated in FBS media for 6 hours was sufficient to ablate inhibition of calcium mobilization. These results indicated that a factor (likely LPP) is present in FBS that degrades LPA. In follow-up studies, we found that similar to faf-BSA, BSA did not affect the ability of LPA to inhibit calcium. However a common additive for in vitro mouse cell culture, β- mercaptoethanol, degraded the signaling ability of LPA similar to FBS (data not shown), thus we elected to not supplement media with β-mercaptoethanol for future in vitro experiments. Additionally, we did not see a difference when LPA was incubated with splenocytes, indicating that LPA degradation or preservation was not affected by splenocytes for this duration of experiment (data not shown).

3.5 3 No culture 2.5 2

Indo ratio 1.5 1 0 100 200 300 Time (s) 3.5 fafBSA culture 3 2.5 0 µM LPA 2 20 µM etherLPA Indo ratio 1.5 20 µM esterLPA 1 0 100 200 300 Time (s)

FBS culture 3

2 Indo ratio 1 0 100 200 300 Time (s)

Figure 3.8 The ability of LPA to inhibit calcium mobilization is ablated by culture in FBS-media Ester- or ether-LPA was used fresh (top panel) or cultured at 420 µM in culture media (IMDM) supplemented with fatty-acid free BSA (faf-BSA, center panel) or FBS (lower panel) for 6 hours. The supernatant was then collected and aliquots of the supernatant were used as a source of LPA at a final concentration of 20 µM LPA (based on original LPA concentration) for calcium mobilization assays with splenic B cells stimulated with anti- IgM (indicated by arrow). As a control, cells were treated with media cultured without LPA (red lines). Representative of 2 independent experiments.

With these results, we first examined proliferation of splenic CD8+ T cells in BSA-media, as we had found that BSA-media did not readily degrade LPA. We stimulated CFSE-labeled CD8+ T cells and assessed proliferation after 3 days.

However, we found that BSA-media was insufficient to support CD8+ T cell 46 proliferation in response to anti-CD3 stimulation, whereas CD8+ cells cultured in

FBS-media were able to proliferate (Figure 3.9). Although BSA-media was insufficient to support T cell proliferation, we found that it was sufficient to support

CD8+ T cell activation for 1 day of culture, thus we next compared CD8+ T cell activation in the presence or absence of LPA after a day of stimulation.

!"#%!"$ !"#!"#$ 60

40 Cells Cells # #

20 20

0 0 0 102 103 104 105 0 102 103 104 105 : CFSE : CFSE $%&'#

Figure 3.9 Media supplemented with BSA is insufficient to support T cell proliferation in vitro C75BL/6 splenic CD8+ T cells were purified by magnetic bead separation and labeled with CFSE before stimulation with 10 µg/mL plate-bound anti- CD3 for 3 days in media supplemented with 5% FBS (left panel) or 5% BSA (right panel).

To determine how CD8+ T cell activation was affected by LPA signaling in vitro, we stimulated splenic CD8+ T cells with anti-CD3 in the presence or absence of LPA, and examined CD62L (L-selectin) expression, which is found at high levels on naïve and memory T cells, but is downregulated upon activation in effector T cells (170). We found that LPA treatment caused a slight inhibition of 47 activation, but only with suboptimal levels of activation (Figure 3.10). Specifically, while LPA had no effect on CD62L expression in the absence of stimulation (left histogram), LPA treatment inhibited downregulation of CD62L upon activation with 1 µg/mL anti-CD3 (center histogram), but not at 10 µg/mL anti-CD3 (right histogram). This suggested to us that the negative regulatory function of LPA may operate more efficiently with relatively weak stimulation.

0 µg/mL antiCD3 1 µg/mL antiCD3 10 µg/mL antiCD3 100 100 100

80 80 80

60 60 60

% of Max % of Max % of Max 40 40 40 % of max 20 20 20

0 0 0 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 CD62L No treatment 0 µM LPA 1 µM LPA 10 µM LPA

Figure 3.10 LPA treatment inhibits activation with suboptimal stimulation C57BL/6 splenocytes were cultured in BSA-media for 24 hours with (center and right histograms) or without (left histogram) indicated plate- bound anti-CD3 and soluble anti-CD28, in the presence (blue and green lines) or absence of LPA (red line, gray filled). Data shown is gated on CD8+ T cells. Representative of 2 independent experiments.

As we had encountered difficulties using LPA in vitro, we decided to continue experiments with the metabolically stable LPA analog, octodecenyl thiophosphate (OTP) to induce LPA signaling, similar to that achieved previously in vivo (68, 69). OTP is not degraded by LPP1 and the absence of the glycerol backbone prevents acylation by lysophosphatidate transacetylases (68).

48 Importantly, OTP is preferentially recognized by LPA5 compared to LPA and engages LPA5 at a lower EC50 than other LPA receptors (68, 171). Additionally, since we found that inhibition by LPA was most evident at lower levels of stimulation, we elected to use a T cell receptor transgenic mouse to allow a more physiological stimulation with a specific peptide instead of monoclonal antibody treatment. The OT-I T cell receptor (TCR) transgenic mouse is a well- characterized model that features CD8+ T cells that express a TCR specific for the chicken ovalbumin peptide, SIINFEKL. Through use of this model, we were able to modulate the strength of stimulation by use of different peptides (23).

Surface plasmon resonance studies have revealed that wild-type TCR affinities range from 1-100 µM (172). The OT-I TCR has an affinity for the SIINFEKL peptide of KD=6.5 µM but also recognizes altered peptide ligands, such as

SIIGFEKL (G4) (172) with weaker affinity (KD ~10 µM).

Using this in vitro system we monitored CD8+ T cell proliferation and expression of activation markers after peptide stimulation in the absence or presence of OTP. To achieve this, C57BL/6 splenocytes were isolated and labeled with the intracellular fluorescent dye, carboxyfluorescein succinimidyl ester (CFSE) and T cell proliferation was monitored by the dilution of CFSE in dividing CD8+ T cells. When CFSE-labeled OT-I splenocytes were stimulated with G4 peptide, CD8+ T cells were acutely inhibited in proliferation in the presence of OTP (Figure 3.11). By day 3 control cells had typically started to divide with the majority having undergone at least 2 divisions by day 7 while the majority of OTP-treated cells remained undivided in the same period.

Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7

5 99.8 5 89.6 5 49 5 14.7 5 7.39 5 4.37 5 3.7 10 Control 10 10 10 10 10 10

104 104 104 104 104 104 104

103 103 103 103 103 103 103

102 102 102 102 102 102 102

0 0 0 0 0 0 0

0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105

5 99.7 5 97.6 5 93.4 5 84.3 5 82.4 5 78.6 5 78.9 10 OTP 10 10 10 10 10 10 104 104 104 104 104 104 104

103 103 103 103 103 103 103

102 102 102 102 102 102 102

0 0 0 0 0 0 0

2 3 4 5 2 3 4 5 2 3 4 5 2 3 4 5 2 3 4 5 2 3 4 5 2 3 4 5

CD25 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 CFSE 100 Control 100 100 100 100 100 100 80 OTP 80 80 80 80 80 80 60 Unstim. 60 60 60 60 60 60 40 40 40 40 40 40 40

20 20 20 20 20 20 20

0 0 0 0 0 0 0 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 0 102 103 104 105 CFSE

Figure 3.11 LPA signaling ablates OT-I proliferation to G4 peptide stimulation OT-I splenocytes were labeled with CFSE, stimulated with G4 peptide and cultured for 1-7 days. At each indicated timepoint, cells were harvested, stained and data shown is from gated viable CD8+ T cells. CD8+ T cells were analyzed for the expression of CD25 (Y-axis) and CFSE fluorescence (X-axis) at the indicated times. CD8+ T cells were left untreated (upper row) or OTP treated (middle row). Overlay of CFSE histograms by day are shown in bottom row in the absence (thin line) or presence of OTP (bold line) or when left unstimulated (gray filled). Representative of 3 independent experiments.

Furthermore, and as indicated in Figure 3.11, expression of the activation marker, CD25, was also depressed in OTP-treated cells. Direct comparison of the T cell activation markers, CD25 and CD44, after in vitro TCR stimulation revealed inhibited expression with OTP treatment (Figure 3.12A). As a consequence of this inhibition of activation and proliferation, OTP treatment also resulted in reduced accumulation of CD8+ T cells (Figure 3.12B) without a significant effect on viability as compared to unstimulated controls (Figure 3.12C).

50 A Control OTP Unstimulated ** 100 100 *** 100 80 80

60 60

% of Max % of Max 40 40 50 % of max % of max

20 20 % of control

0 0 0 102 103 104 105 0 102 103 104 105 0 CD25 CD44 Control CD25 CD44 MFI MFI B 20 C Viability Control )

3 ns 15 Unstimulated 100 OTP 10 50 5 Cell no. (x10 % of control 0 0 1 2 3 4 5 6 7 d d d d d d d Unstim. OTP Time

Figure 3.12 Increased LPA signaling inhibits activation and prevents accumulation of CD8+ T cells in vitro (A) CD25 and CD44 expression on OT-I CD8+ T cells is shown in histograms after 3 or 4 days, respectively, of in vitro culture in the absence of stimulation (gray filled) or after G4 peptide stimulation alone (thin line), or in the presence of OTP (bold line). Histograms are representative of 3 independent experiments and are summarized in bar chart (Mean + SEM). (B) Enumeration of in vitro cultured OT-I CD8+ T cells at the indicated times in the absence of G4 (grey circles) or after G4 stimulation alone (open circles) or in the presence of 20 µM OTP (black circles). (C) Cell viability was determined by 7-AAD staining at the peak of CD25 expression and are representative of 3 independent experiments (Mean + SEM).

Given that proliferation and CD25 (high affinity IL-2 receptor chain) expression are affected by IL-2 signaling (173), we questioned if LPA or OTP treatment was inhibiting the CD8+ T cell response independently of TCR and LPA receptor signaling. We wanted to confirm that LPA or OTP was not affecting proliferation and activation merely by preventing IL-2 binding to the IL-2 receptor or otherwise impacting the IL-2 signaling pathway. To test this, we used the HT-2 51 cell line, which is a T cell line that is dependent on IL-2 but independent of TCR signaling for survival, and thus this cell line can be used to measure IL-2 activity in a sensitive manner in vitro (174). We cultured HT-2 cells with titrated concentrations of IL-2 (0.06-1000 ng/mL [0.6-10,000 U/mL]) in the presence or absence of LPA or OTP to determine if these treatments interfered with IL-2 signaling in this cell line. We found that LPA or OTP treatment did not affect HT-2 survival, as cell death was observed at the same concentration of IL-2, regardless of the presence of LPA or OTP. This indicates that LPA or OTP does not interfere with IL-2 binding the IL-2R or IL-2 signaling independently of TCR signaling (data not shown).

OTP treatment does not affect APC function and inhibits activation and proliferation to a lesser degree to a high affinity antigen

In this in vitro system, activation of CD8+ T cells is dependent on presentation of the peptide in MHCI by splenic APC. Therefore, we wanted to determine if increased LPA signaling was affecting CD8+ T cell activation indirectly by interfering with peptide presentation by APC. By using a peptide for stimulation instead of whole antigen, the peptide can load externally into the binding groove of the MHC molecules expressed on the surface of the APC, rather than requiring internalization, processing, and internal peptide loading onto

MHC as is required for whole antigen (175). However we wanted to confirm that

LPA signaling was not interfering with external peptide loading into the MHCI molecule and thus altering the antigen presentation. To address this possibility, we pre-treated T cell-deficient splenocytes (TCRα–/–) with OTP while pulsing with

52 SIINFEKL peptide. We then washed away excess peptide and OTP and used these peptide-loaded APC to stimulate purified CD8+ OT-I T cells in the presence or absence of OTP. If OTP treatment did negatively impact peptide presentation by APC, we would expect a lower level of activation in CD8+ T cells stimulated with APC that had been pretreated with OTP. We found that CD8+ T cell activation was not affected by OTP pretreatment of APC, as shown by similar

CD25 expression after co-culture of OT-I CD8+ T cells and SIINFEKL-pulsed

TCRα–/– splenocytes (Figure 3.13). This indicates that OTP treatment does not interfere with peptide presentation by APC, however does not address if OTP treatment could interfere with APC processing of whole antigen.

30000 25000 APC pretreatment: 20000

I 15000 F

M OTP 10000 5

2 500 D

C 400 300 200 100 0 Culture condition: OTP OTP Unstimulated Stimulated

Figure 3.13 OTP treatment does not affect peptide presentation Splenocytes from a T cell-deficient mouse (TCRα–/–) were pulsed with SIINFEKL peptide in the presence (black) or absence (white) of OTP, washed, and used to stimulate purified CD8+ OT-I T cells in culture in the presence or absence of OTP (n=1).

Additionally, we wanted to compare the ability of OTP to dampen proliferation and activation to different affinity antigens, as we had previously

53 found that LPA inhibition was greatest with suboptimal stimulation (Figure 3.10).

Interestingly, we found that when we stimulated at a 100-fold lower concentration

(10 nM) of a higher affinity peptide (SIINFEKL), we again observed inhibition of activation by OTP treatment but at a diminished capacity as compared to the inhibition of activation to the G4 peptide (Figure 3.11). Proliferation with OTP treatment appeared to delay proliferation by 1 division as compared to untreated

OT-I CD8+ T cells and CD25 expression was dampened to a lesser degree than to G4 stimulation. This indicated that LPA inhibitory regulation of TCR signaling may operate more strongly over an optimal (weaker) affinity for specific antigen.

100 100 Control 80 80 OTP 60 60 Unstimulated

% of Max % of Max 40 40 % of max % of max

20 20

0 0 2 3 4 5 0 102 103 104 105 0 10 10 10 10 CFSE CD25

Figure 3.14 OTP treatment dampens OT-I activation and proliferation to the higher affinity peptide, SIINFEKL OT-I splenocytes were labeled with CFSE, stimulated with SIINFEKL peptide and cultured for 3 days. Viable CD8+ T cells were analyzed for the expression of CD25 and CFSE fluorescence in the absence (thin line) or presence of OTP (bold line) or when left unstimulated (gray filled) and is representative of 2 independent experiments.

LPA signaling modestly inhibits effector function in vitro

As we had found that increased LPA signaling dampened activation, we questioned if LPA signaling also impacted the effector function of CD8+ T cells, including the ability to produce effector cytokines and exert cytotoxic activity

54 towards a cell presenting specific peptide in the context of MHCI. In earlier experiments, we modulated LPA signaling during stimulation to determine how

LPA signaling affected the transition of naïve to activated CD8+ T cells. In contrast, for these experiments we stimulated naïve CD8+ T cells to differentiate to effector T cells in vitro in the absence of additional LPA signaling and then tested how LPA signaling would affect subsequent re-stimulated effector functions. To generate effector CD8+ T cells in vitro, we pulsed OT-I splenocytes with SIINFEKL peptide, washed away excess peptide, and cultured the cells for 5 days in the presence of IL-2. This treatment resulted in >90% CD8+ T cells that were CD25+ (data not shown). We then restimulated the effector CD8+ T cells with G4- or SIINFEKL-pulsed target cells (EL4 cells) and monitored CD8+ T cell production of the effector cytokines, TNFα and IFNγ. We were unable to detect intracellular cytokine production in response to G4-pulsed targets (data not shown), similar to published data that found no detectable intracellular IFNγ staining after co-culture with EL4 cells pulsed with 1 µM G4 (159). In response to

SIINFEKL-pulsed cells, we found that OTP treatment had a modest effect on cytokine generation as determined by intracellular staining, inhibiting the proportion of cells expressing both cytokines by approximately 9% (Figure 3.15).

This is similar to findings by another group who has found only a slight inhibition of IFNγ production with LPA treatment in human T cell lines (Paul Murray, personal communication). This indicates that LPA signaling has a negligible effect on effector cytokine generation.

55 Unstimulated Control OTP

5 105 0.26 1.78 105 3.04 50.1 10 3.21 41.6

4 104 104 10

3 103 103 10

2 102 102 10 0 0 0 α 96.5 1.43 32.3 14.5 36.5 18.7 2 3 4 5 2 3 4 5 0 102 103 104 105 0 10 10 10 10 0 10 10 10 10 TNF IFNγ

Figure 3.15 OTP treatment modestly dampens effector cytokine production OT-I effector T cells were generated in vitro by pulsing OT-I splenocytes with 1 µM SIINFEKL and culture with IL-2 for 5 days. On day 5 of culture, target cells (EL4 cells) were pulsed with SIINFEKL and cultured with OT-I T cells for 4 hours in the presence of Brefeldin A, in the presence (right panel) or absence (center panel) of OTP. Cells were permeabilized and stained for IFNγ and TNFα expression by flow cytometry. Representative of 2 independent experiments.

We next tested if LPA signaling affected the killing ability of effector OT-I

CD8+ T cells. Similar to the cytokine production experiment, we generated effector OT-I CD8+ T cells in vitro and then co-cultured the effector cells with

SIINFEKL-pulsed target cells (EL4 cells) and monitored killing of the target cells.

To approach this question in congruence with our findings that inhibition by LPA signaling was most potent with suboptimal stimulation, we used low concentrations (10 or 100 nM) of SIINFEKL to pulse target cells prior to culture with effector OT-I CD8+ T cells (Figure 3.16). We further modulated killing by titrating the ratio of CD8+ effector T cells to target cells pulsed (Figure 3.16). We first examined the ability of OTP to inhibit killing of fluorescently labeled target cells (CFSEhigh) that were pulsed with SIINFEKL peptide as compared to differently labeled control target cells (CFSElow) that are not pulsed with 56 SIINFEKL. By flow cytometry, we were able to compare the two target cell populations and observe specific killing by the reduction in frequency of

SIINFEKL-pulsed target cells as compared to control cells. With this assay, we determined that OTP treatment caused a slight decrease of specific killing by approximately ten percent (Figure 3.16A).

We additionally attempted to determine differences in killing with G4- pulsed target cells, however we were unable to achieve killing of said targets and were thus concerned that this assay may not be sensitive enough to detect small amounts of specific killing. Therefore we used a cytotoxicity detection kit that measures cell death by the release of lactate dehydrogenase (LDH) from lysed cells and has been used previously as a more sensitive assay to detect killing of up to a third of 1 µM G4-labeled targets by OT-I CD8+ effector T cells (159).

However in our hands, we were unable to detect killing of G4-labeled targets. We were able to monitor killing of SIINFEKL-pulsed target cells and we again observed a slight decrease in cytotoxic activity by approximately ten percent over cytotoxicity measured from OT-I CD8+ T cells that were not treated with OTP

(Figure 3.16B). Combined with the intracellular cytokine production experiments, we concluded that LPA signaling has a minor inhibitory effect on OT-I CD8+ T cell effector function in response to SIINFEKL peptide restimulation.

57 A B 90 50

40 80 30 70 20

60 10 % specific lysis % specific lysis 0 50 10 20 30 0 5 10 10 E:T ratio E:T ratio

Figure 3.16 LPA signaling induces a minor inhibition of the killing ability of CD8+ T cells OT-I effector T cells were generated in vitro by pulsing OT-I splenocytes with 1 µM SIINFEKL and culture with IL-2 for 5 days. (A) Target EL4 cells were fluorescently labeled CFSEhigh and pulsed with 100 nM SIINFEKL. Control EL4 cells were not pulsed with peptide and fluorescently labeled CFSElow. CFSEhigh and CFSElow EL4 cells were mixed at a 1:1 ratio and combined with OT-I effector T cells at the indicated effector to target (E:T) ratio. Killing was determined by the loss of CFSEhigh cells relative to CFSElow cells. Representative of 2 experiments. (B) Target EL4 cells were pulsed with 10 nM SIINFEKL and combined with OT-I effector T cells at the indicated effector to target (E:T) ratio. Cell death was assessed by Non-Radioactive Cytotoxicity Assay (LDH ELISA) and adjusted for background lysis of OT-I T cells co-cultured with EL4 cells without SIINFEKL peptide. Representative of 2 experiments.

Increased LPA signaling dampens proliferation and activation of CD8+ T cells in response to SIINFEKL-BMDC stimulation in vivo

Thus far, our in vitro data demonstrate that LPA signaling via LPA5 on

CD8+ T cells inhibits TCR signaling, activation, and proliferation but only modestly suppressed cytotoxic activity. We next addressed whether this LPA regulatory pathway similarly operated on T cells in vivo. To accomplish this, OT-I

CD8+ T cells were purified to >95% purity, CFSE-labeled, and transferred to

C57BL/6 recipients. One day after T cell transfer, 106 bone marrow-derived dendritic cells (BMDC) were induced to undergo maturation with LPS, pulsed 58 with 1 µM SIINFEKL or G4 peptide, washed, and transferred by sub-cutaneous

(s.c.) injection to the scruff of recipient mice (Figure 3.18A). Transfer of peptide- pulsed BMDC has been used by others to transiently introduce antigen in in vivo experiments (176) and we were able to measure activation by CFSE dilution in response to SIINFEKL-BMDC stimulation (Figure 3.18B). In contrast, transfer of

BMDC pulsed with G4 peptide did not stimulate previously transferred OT-I CD8+

T cells as determined by the absence of increased activation antigen expression or proliferation (data not shown). In further attempts to activate transferred OT-I

CD8+ T cells with G4-pulsed BMDC, we varied G4 peptide concentrations, the duration of BMDC peptide pulse, and tested different LPS treatment intervals to achieve maximal DC maturation as determined by CD86 and MHCII expression.

Additionally, we confirmed G4 peptide loading of MHCI on the BMDC by flow cytometry with an antibody that recognizes SIINFEKL peptides within MHCI H2-

Kb (25-D1.16). We found similar expression with G4- and SIINFEKL-pulsed

BMDC after 1-4 hours of peptide pulse and lower expression after overnight incubation with peptide, showing that increased incubation with peptide (over 1 hour) did not improve peptide presentation and that presentation of G4 peptide was similar to SIINFEKL peptide (Figure 3.17). Nevertheless, with none of these conditions did we find that transfer of G4-pulsed BMDC resulted in activation of

OT-I CD8+ T cells in vivo.

59 1 µM G4 10 µM G4 1 µM SIINFEKL 100 100 100

80 80 80

60 60 60

% of Max % of Max 40 40 40 % of max % of max % of max 20 20 20

0 0 0 0 10 2 10 3 10 4 10 5 0 102 10 3 10 4 10 5 0 102 10 3 10 4 10 5 25D1.16 Duration of peptide pulse: No peptide 1 hr 2 hr 4 hr Overnight

Figure 3.17 Peptide pulse of 1-4 hours results in equivalent expression of OVA peptide in H2-Kb BMDC were generated and pulsed with peptide for 1-4 hours or overnight, with LPS maturation for the final hour of incubation. BMDC were then stained with 25- D1.16 antibody and analyzed for surface peptide presentation.

In contrast, when 1 µM SIINFEKL-pulsed BMDC were transferred to recipient mice, OT-I CD8+ T cell proliferation was clearly observed 3 days later

(Figure 3.18B) in the draining lymph nodes (dLN) of subcutaneous (s.c) transfer to the scruff of the mouse (axillary, brachial, cervical lymph nodes). This activation could be titrated by transfer of varying numbers of dendritic cells, with

106 dendritic cells producing consistent, measurable proliferation in the transferred OT-I CD8+ T cells (Figure 3.18B). Interestingly, s.c. transfer of

SIINFEKL-BMDC resulted in OT-I CD8+ T cell activation isolated to the dLN, as

CFSE-diluted cells were not observed in non-draining lymph nodes (ndLN: inguinal, mesenteric lymph nodes) nor in the spleen (data not shown), indicating

60 that this was a local, but not systemic, method of activating transferred OT-I

CD8+ T cells. Therefore the in vivo experiments of T cell activation relied on

SIINFEKL-pulsed BMDC transfer to stimulate OT-I CD8+ T cells in the local lymph nodes.

A d1 d0, t1h d0, t0h d3

Transfer 106 Inject OTP s.c. Inject Analyze OTI CD8+ T cells (every 8h) SIINFEKLBMDC in dLN

CD45 disparate OTI

B 100

s .fc 1 _DCs A D

SO 4

80 10 DC

s .fc 2 _DCs A D

SO 5

x 60 10 DC s .fc 3 _DCs A D Ma SO f o

% 10 DC s .fc 4 _DCs A D 4SO 0

0 0 102 103 104 105 : CFSE CFSE

Figure 3.18 Schematic of peptide-BMDC transfer to induce proliferation of transferred OT-I CD8+ T cells (A) Schematic diagram of adoptive transfer of OT-I CD8+ T cells and in vivo OTP treatment. CD8+ OT-I T cells were isolated and 106 CFSE- labeled cells were adoptively transferred intravenously to C57BL/6 recipients one day prior to immunization. One hour before s.c. transfer of 106 SIINFEKL-pulsed BMDC, OTP (5 mg/kg) was administered s.c. and this dose was repeated every 8 hours. On day 3 after transfer of peptide- loaded BMDC, mice were sacrificed and draining lymph nodes harvested for analysis. (B) Activation of transferred OT-I CD8+ T cells as in Fig. 3.14A is titrated by transfer of 104 (red line), 105 (blue line), or 106 (green line) SIINFEKL- BMDC (n=1).

To enhance LPA signaling in vivo, we used OTP as previously reported

(68, 69). Initial attempts with daily dosing of 1 mg/kg OTP did not affect OT-I

CD8+ T cell activation (data not shown). However, we later learned that subcutaneous administration of OTP at 5 mg/kg results in detectable levels in the blood after 1 hour with an approximate half-life of 5.5 hours (Gabor Tigyi, personal communication). Thus, daily dosing of OTP was likely not adequately frequent to increase LPA receptor signaling. Therefore, in all future experiments, mice were treated with 5 mg/kg OTP every 8 hours, with the first dose preceding adoptive transfer of SIINFEKL-BMDC by 1 hour (Figure 3.18A). As a result, OT-I

CD8+ T cells recovered from the dLN of OTP-treated animals 3 days after stimulation had proliferated to a significantly lesser extent than vehicle-treated animals (Figure 3.19A). Furthermore, in the dLN of OTP-treated mice, OT-I CD8+

T cells were present at decreased numbers relative to vehicle-treated animals treated with OTP, likely as a result of the inhibited proliferation. In addition, in the

OTP-treated recipients, OT-I CD8+ T cells were less activated compared to vehicle-treated mice as determined by CD25 expression (Figure 3.19C). Taken together, our data demonstrate that increased LPA signaling using OTP inhibits

OT-I CD8+ T cell proliferation and activation to antigen-specific stimulation both in vitro and in vivo.

62 A B 300 ** 100 ) 80 3 200 60 % of Max40 100 # OTI (x10

% of max 20

0 2 3 4 5 0 10 10 10 10 0 CFSE Vehicle OTP C 100 Control 3 ) 80 OTP 4 2 % of Max60 Unstimulated

40 1 % of max 20 CD25 MFI (x10 0 0 0 102 103 104 105 Vehicle OTP CD25

3.19 LPA signaling inhibits TCR-mediated CD8+ T cell activation in vivo Experiment was conducted as diagrammed in Fig. 3.18. Data shown are of OT-I CD8+ T cells in the dLN and are representative of 2 independent experiments (n=4/group, mean + SEM). (A) Histograms of CFSE dilution in vehicle-treated (thin line), OTP-treated (bold line), and unstimulated control (gray filled) OT-I CD8+ T cells. (B) Number of OT-I CD8+ T cells in dLN. (C) Representative histograms of CD25 expression by OT-I CD8+ T cells after vehicle-treatment (thin line), OTP-treatment (bold line) or unstimulated (grey shaded). Right panel shows expression of CD25 as mean fluorescence intensity (MFI) by OT-I CD8+ T cells in the draining lymph nodes from individual mice after vehicle-treatment (closed circles) or OTP-treatment (open circles)

63 We next wanted to test if this inhibition of CD8+ T cell activation was mediated by LPA5, as we had shown in calcium mobilization assays (Figure 3.5).

Therefore we repeated the experiment as diagrammed in Figure 3.18A, however

+ we also transferred LPA5-deficient OT-I CD8 T cells into different recipient mice and treated with OTP as before. In contrast to our previous findings, we were

+ surprised to find that the proliferation of LPA5-deficient OT-I CD8 T cells was also inhibited by OTP treatment (Figure 3.20A). This could indicate that the increased LPA signaling was acting on other LPA receptors on the CD8+ T cells and causing inhibition, however in consideration with our calcium mobilization assay results, this would be independent of TCR signaling. Alternatively, by dosing with OTP we were increasing LPA signaling to all cells, not just CD8+ T cells, and other cells, such as antigen-presenting cells (APC), could also be negatively regulated by OTP. However, we were intrigued to find that in the vehicle-treated controls, there was an increased accumulation of LPA5-deficient

+ + OT-I CD8 T cells over LPA5-suffcient OT-I CD8 T cells in the dLN (Figure

3.20B). This suggested that normal levels of LPA were sufficient to inhibit LPA5-

+ sufficient CD8 T cells whereas LPA5-deficient T cells were unresponsive to this negative regulation. In addition, this experiment was CD8+ T cell-intrinsic, as only

+ the transferred OT-I CD8 T cells lacked LPA5 and the absence of OTP dosing did not increase LPA signaling to endogenous cells (including APC). We decided to pursue this finding in the following chapter.

64 A B 10 100 ) 80 5 8 60 6 % of Max40 4 % of max 20 # OTI (x10 2 0 0 102 103 104 105 0 CFSE WT LPA KO 5

Figure 3.20 OTP treatment inhibits proliferation of LPA5- –/– deficient OT-I; vehicle-treated LPA5 T cells accumulate to a greater extent (A) Representative histograms of CFSE dilution in vehicle-treated (thin line), OTP-treated (bold line), and unstimulated control (gray filled) LPA5- deficient OT-I CD8+ T cells (n=1). (B) Number of LPA5-sufficient (black dots) and LPA5-deficient (white dots) OT-I CD8+ T cells in dLN of vehicle-treated recipients (n=1).

Conclusion

TCR signaling is the critical first step in T cell activation (10). Previous research from our lab demonstrated that LPA inhibits B lymphocyte antigen receptor-mediated signaling that ultimately impairs intracellular calcium release from intracellular stores and indicated that this occurs via LPA5 signaling (152).

We have found that this is also true in CD4+ and CD8+ T cells and confirmed that

LPA5 is required for LPA inhibition in both B and T cells. Further, we demonstrate that inhibition of calcium mobilization is exquisitely sensitive in CD8+ T cells, with a dampened response with as low as 1 µM LPA, thus approximating physiological LPA levels (150, 177). This inhibition is ablated in LPA5-deficient T cells, and LPA5-heterozygous T cells exhibit an intermediate phenotype,

65 indicating that a single gene copy of Lpar5 is sufficient for LPA signaling to occur and also possible intermediate cell surface expression of the LPA5 receptor.

LPA analog (68), to more stably induce LPA signaling. In vitro OTP treatment strongly reduced antigen-specific CD8+ T cell activation and proliferation to a relatively weak affinity peptide but was less effective against a higher affinity peptide, suggesting this LPA negative regulatory function may operate more effectively with weaker stimulation. This may serve as a safety mechanism to protect against autoimmunity, as weakly reactive T cells may be cross-reactive to self-tissue and LPA signaling could function to inhibit these cells from becoming activated and inducing damage to the host.

We next questioned if LPA signaling also had an inhibitory effect on the function of in vitro-generated CD8+ T effector cells. We found that OTP treatment caused a slight decrease in effector cytokine generation and in cytolytic ability in response to SIINFEKL stimulation. We were unable to induce cytokine generation to G4 peptide stimulation and although others had measured OT-I

CD8+ T cell killing of 20-40% of EL4 targets pulsed with 1 µM G4 peptide (159), we were unable to achieve killing with the same experimental parameters.

However to provide optimal conditions for OTP treatment, BSA was used instead of FBS as the media additive, therefore the effector function was likely dampened.

66 The ability of G4 to activate naïve OT-I CD8+ T cells but not stimulate effector function may indicate that the affinity threshold for cytotoxic function is higher than that required for activation and proliferation. However we cannot exclude that in contrast to the activation studies, EL4 cells were used as the APC for the effector function studies and they may have diminished antigen presentation abilities as compared to primary APC (159, 178, 179).

By transfer of peptide-pulsed BMDC, we wanted to test the results of the in vitro experiments by comparing LPA inhibition of activation and proliferation to different affinity peptides. We questioned if we would see a greater inhibition of proliferation by OTP treatment if we stimulated OT-I CD8+ T cells with the lower affinity peptide, G4, as we had observed more potent inhibition by LPA signaling to G4 over SIINFEKL in vitro. However in contrast to the results of the in vitro experiments, we were unable to achieve OT-I CD8+ T cell activation to G4-BMDC in vivo. We conducted various experiments testing different peptide pulse and

LPS maturation variables but were unable to achieve activation of transferred

OT-I to G4-BMDC in vivo, similar to another researcher (Matt Burchill, personal communication). This indicated to us that G4 may not be a physiologically relevant peptide to stimulate OT-I CD8+ T cells in vivo when presented in the context of BMDC. Although other researchers have been able to activate transferred OT-I CD8+ T cells to lower affinity peptides, this occurred with the use of Listeria monocytogenes (Lm) modified to express different OT-I peptides (23), which results in a more robust, systemic infection and activation of T cells than activation from peptide-loaded dendritic cells which induces a transient presence

67 of antigen (176). Regardless, we were able to more potently inhibit activation to

SIINFEKL peptide in vivo with OTP administration than what we observed with in vitro stimulation with SIINFEKL peptide, therefore we estimate that the strength of activation with SIINFEKL peptide is diminished in vivo. However we cannot exclude the possibility that the effective dose of OTP may have been greater in vivo than in culture. Taken together, we conclude that LPA signaling is capable of inhibiting TCR signaling, which results in reduced activation and proliferation both in vitro and in vivo.

Although peptide loading was not affected by OTP treatment, other cells that influence T cell activation could have been affected by OTP treatment, such as APC that engulfed the transferred SIINFEKL-BMDC or CD4+ T helper cells.

+ We also cannot exclude the fact that transferred LPA5-deficient CD8 T cells may have been directly inhibited by OTP treatment in an LPA5-independent manner, however this is less likely as our calcium data suggest that this inhibition does not affect activation in the absence of LPA5.

From these studies, we had preliminary data to suggest that endogenous

+ levels of LPA inhibited normal CD8 T cell activation, whereas LPA5-deficient

CD8+ T cells featured a greater expansion in response to stimulation. Therefore,

+ we wanted to further test how CD8 T cells are affected by the absence of LPA5 signaling. By using LPA5-deficient mice in the next chapter, we addressed this question and additionally, focused the effect of LPA signaling to CD8+ T cells instead of increasing LPA signaling to all cells and LPA receptors by LPA or OTP treatment.

CHAPTER IV

+ LPA5-DEFICIENT CD8 T CELLS EXHIBIT ENHANCED IMMUNE

RESPONSES AND ANTI-TUMOR ACTIVITY

Introduction

In the previous chapter, we demonstrated that increased LPA signaling via

+ LPA5 inhibited the CD8 T cell immune response. This is significant as LPA has received considerable attention for its association with cancer and is produced by a number of different malignant cell types (84-87). The increased production of

LPA has been shown to enhance tumor progression (116-118) by stimulating migration, survival, metastases, angiogenesis, drug resistance, and proliferation of cancer cells (119-125). One mechanism of increased LPA production is by upregulation of the enzyme that produces LPA, autotaxin (ATX), resulting in LPA concentrations up to 10 µM in individuals with certain tumors (84-87, 124).

Various studies in mouse models have demonstrated that the immune system is able to identify and specifically react to tumor cells (33-35). This cancer immunosurveillance is chiefly mediated by tumor-specific cytotoxic CD8+ T cells that eliminate tumor cells expressing tumor antigens. The presence of tumor- infiltrating T cells, and CD8+ T cells in particular, have been shown to be important for tumor rejection (36, 37). However multiple inhibitory mechanisms within the tumor microenvironment have been described that impede tumor rejection by T cells (39, 40) including low affinity reactivity for these tumor

69 antigens (44-47).

+ In this chapter, we show that LPA5-deficient CD8 T cells exhibit enhanced proliferation and accumulation in response to peptide stimulation. We further

+ demonstrate that in the absence of LPA5, tumor-specific CD8 T cells are more efficient in controlling tumor progression and we examine the mechanism of this enhanced control. Finally, we discuss our findings as the first report of lysophospholipid-mediated protection of tumor from adaptive immunity and LPA5 as a potential target for cancer treatment.

Results

Characterization of LPA5-deficient mice

Although we had used the LPA5-deficient mice to confirm the dependence of LPA inhibition in the previous chapter, we wanted to further characterize these mice before use in activation studies as LPA5-deficient mice had not been described in the literature. LPA5-deficient mice were generated at Lexicon

Pharmaceuticals on a mixed 129/SvEvBrd and C57BL/6 background, thus we used gender-matched littermates as controls as we backcrossed the strain to

–/– C57BL/6. We found that LPA5 mice were phenotypically unremarkable as compared to wild-type littermates, and heterozygous crosses resulted in

+/+ –/– Mendelian ratios of pups (Figure 2.1D). Comparison of LPA5 and LPA5 mice revealed comparable numbers and frequencies of splenic B and T cell populations and CD4+ and CD8+ T cell subsets (Figure 4.1A). Importantly for calcium mobilization assays, the expression of surface CD3 was similar between

+/+ +/– –/– + LPA5 , LPA5 , and LPA5 CD8 T cells, when normalized for size by gating on

70 cells of similar forward scatter (FSC slices, Figure 4.1B). Another group has

independently generated LPA5-deficient mice and also found similar lymphocyte

populations and a lack of phenotypic differences (115).

250K 100 100 A B FSC slices Slice Slice 80 80 x (% of total) x 1 2 80 ns 60 60 Ma Ma f f o o

+/+ 200K % LPA % 5 40 40 / LPA Slice 1 Slice 2 Slice 3 20 20 60 5

150K 0 0 A ns 2 3 4 5 2 3 4 5 18.5 25 12.7 0 10 10 10 10 0 10 10 10 10 : CD3e C : CD3e

100

40 ns SS 6 . 59 s .fc 1 _DCs A D Slice SO

ns 100K 80 +/+

LPA5 % of viable 0 5 s .fc 2 _DCs A D x 3 SO 60 Ma f

o +/

% LPA

40 5 5 . 64 s .fc 3 _DCs A D

50K SO /

20 LPA 3 . 58 s .fc 4 _DCs A D 0 SO 5 0 + + 0 102 103 104 105 B cells T cells CD4 CD8 0 : CD3e 0 50K 100K 150K 200K 250K CD3 FSCA

250K 100 100 A B FSC slices Slice Slice 80 80 x (% of total) x 1 2 80 ns 60 60 Ma Ma f f o o

+/+ 200K % LPA % 5 40 40 / LPA Slice 1 Slice 2 Slice 3 20 20 60 5

150K 0 0 A ns 2 3 4 5 2 3 4 5 18.5 25 12.7 0 10 10 10 10 0 10 10 10 10 : CD3e C : CD3e

100

40 ns SS 6 . 59 s .fc 1 _DCs A D Slice SO

ns 100K 80 +/+

LPA5 % of viable 0 5 s .fc 2 _DCs A D x 3 SO 60 Ma f

o +/

% LPA

40 5 5 . 64 s .fc 3 _DCs A D

50K SO /

20 LPA 3 . 58 s .fc 4 _DCs A D 0 SO 5 0 + + 0 102 103 104 105 B cells T cells CD4 CD8 0 : CD3e 0 50K 100K 150K 200K 250K CD3 FSCA

–/– Figure 4.1 Characterization of lymphocytes in LPA5 mice (A) Proportions of B cells and T cells, and CD4+ and CD8+ T cell subsets +/+ –/– in spleens of LPA5 (closed circles) and LPA5 mice (open circles) (n=6/group, mean +/- SEM). +/+ (B) Surface expression of CD3 (right panels) was assessed in LPA5 (red +/- –/– line), LPA5 (blue line) and LPA5 (green line) after gating on viable (7- AAD-) CD8+ T cells of similar size (FSC, left panel).

71 LPA5-deficient T cells show enhanced proliferation to SIINFEKL-BMDC in vivo

We had previously shown that LPA signaling inhibited calcium mobilization

+ + in CD8 T cells, however calcium mobilization in LPA5-deficient CD8 T cells was not affected by LPA treatment (Figures 3.2, 3.5), confirming that LPA5 was required for inhibition by LPA. Thus we next wanted to determine if LPA5- deficient CD8+ T cells were refractory to LPA inhibition of downstream biological events and would exhibited an enhanced response to stimulation in vivo as compared to wild-type CD8+ T cells. We decided to test this at physiological

–/– + levels of LPA, as we had preliminary data that LPA5 OT-I CD8 had increased

+/+ + accumulation in the dLN relative to LPA5 OT-I CD8 T cells in the absence of

OTP treatment (Figure 3.20B). The concentration of LPA in the plasma of mice is approximately 1 µM (150, 168), which we have determined is sufficient to dampen calcium mobilization in CD8+ T cells (Figure 3.2). However, since 1 µM

LPA does not potently inhibit calcium mobilization, we were unsure if the low level of inhibition observed in vitro would translate to a significant defect in proliferation and activation in vivo with systemic levels of LPA. To test this, we

transferred congenically mismatched (by CD45.1 expression) LPA5-deficient or wild-type OT-I CD8+ T cells to C57BL/6 recipients and again stimulated by adoptive transfer of SIINFEKL-BMDC, in the absence of additional LPA signaling

(Figure 4.2). By using this approach, we were able to discriminate transferred

CD8+ T cells from endogenous CD8+ T cells by CD45.1 expression and isolate

+ the deficiency of LPA5 signaling to the transferred CD8 T cells, as the recipient

72 mice and SIINFEKL-BMDC were LPA5-sufficient. This is in contrast to the in vivo experiments in the previous chapter that utilized OTP treatment to increase LPA signaling systemically, thus likely also affecting cells other than CD8+ T cells.

d1 d0 d3

Transfer 106 Inject Analyze OTI CD8+ T cells SIINFEKLBMDC in dLN

CD45 mismatched LPA +/+ or LPA / OTI 5 5

Figure 4.2 Schematic of adoptive transfer of OT-I CD8+ T cells and subsequent activation to SIINFEKL-BMDC + CD8 T cells were isolated from CD45-mismatched wild-type and LPA5- deficient OT-I spleen and pooled lymph nodes and labeled with CFSE. 106 CFSE-labeled cells were adoptively transferred intravenously to C57BL/6 recipients a day prior to immunization. On day 0, 106 SIINFEKL-pulsed BMDC were transferred by subcutaneous injection to the scruff. On day 3 after transfer of peptide-loaded BMDC, mice were sacrificed and transferred OT-I T cells in the draining lymph nodes were analyzed.

We found that, as we had previously shown (Chapter 3), some of the transferred wild-type CD8+ OT-I T cells had proliferated in response to transfer of

SIINFEKL-DC, although a portion of the cells had not proliferated and thus

high remained CFSE (Figure 4.3A). In contrast, during the same period, LPA5- deficient OT-I T cells featured increased cell division with fewer cells that had not proliferated and remained CFSEhigh (Figure 4.3A). Consistent with an increased

–/– + frequency of LPA5 OT-I CD8 T cells proliferation, a relatively higher number of

73 mutant OT-I CD8+ T cells had accumulated in the dLN relative to wild-type OT-I

CD8+ T cells (Figure 4.3B). Together these data suggest that in the absence of

+ negative regulation by LPA5, antigen-specific stimulation of OT-I CD8 T cells trends to an increased percentage of cells that are stimulated to proliferate and, accordingly, higher numbers of antigen-specific CD8+ T cells in the dLN.

A LPA +/+ LPA / Unstim. B 5 5 p= 0.053 100 10 +/+ LPA5 80 ) 5 LPA / 60 5 % of Max40 5 20 # OTI (x10 0 0 102 103 104 105 0 CFSE Unstim. SIINFEKL

+ Figure 4.3 LPA5-deficient CD8 T cells stimulated by peptide in vivo exhibit enhanced proliferation and accumulation +/+ –/– + 6 Purified LPA5 or LPA5 OT-I CD8 T cells were CFSE labeled and 10 cells transferred to C57BL/6 recipients, as described in Fig. 4.2. Recipients were injected one day later with SIINFEKL-pulsed BMDC. 3 days after BMDC transfer, mice were sacrificed and draining lymph nodes harvested for analysis. +/+ –/– (A) Proliferation of LPA5 (thin line), LPA5 (bold line) or unstimulated (filled gray) OT-I CD8+ T cells as indicated by CFSE dilution. Representative of 2 independent experiments. (B) Number of OT-I CD8+ T cells harvested from the draining lymph nodes +/+ –/– + of LPA5 (closed circles) and LPA5 (open circles) OT-I CD8 T cell recipients in the absence of stimulation or 3 days after peptide stimulation. Cumulative of 2 independent experiments.

Given the enhanced proliferation, we next wanted to determine if LPA5- deficient CD8+ T cells were also more activated by analyzing expression of surface markers that are upregulated upon activation by flow cytometry. However,

+ despite the increased antigen-specific proliferation, LPA5-deficient OT-I CD8 T

74 cells did not exhibit enhanced activation as evaluated by expression of the

+ activation markers, CD25 and CD44. Surprisingly, LPA5-deficient OT-I CD8 T cells showed a delay in CD25 upregulation (Figure 4.4A). Wild-type OT-I CD8+ T cells upregulated CD25 and CD44 upon proliferation, starting with early divisions and later divisions having increased CD25 expression as compared to early

+ divisions. However LPA5-deficient CD8 T cells were able to proliferate a few divisions before upregulating CD25. When we analyzed the CD25 and CD44 expression of transferred OT-I CD8+ T cells by mean fluorescent intensity (MFI), there was a significant decrease in population expression (Figure 4.4B), confirming decreased upregulation of activation markers.

Together these data suggest that in the absence of negative regulation by

+ LPA5, antigen-specific stimulation of OT-I CD8 T cells leads to an increased percentage of cells that are stimulated to proliferate, resulting in higher numbers of antigen-specific CD8+ T cells in the dLN. However, the reduced level of activation was perplexing, as we would have predicted that LPA5-deficient OT-I

CD8+ T cells would be less inhibited from activation and would thus express either higher levels of activation markers overall or upregulate CD25 at an earlier round of proliferation. Thus, we next sought to determine why there was delayed

+ activation marker upregulation in proliferating LPA5-deficient CD8 T cells.

75 A +/+ –/– LPA5 LPA5

105 105

104 104

103 103

102 102

0 0

2 3 4 5 2 3 4 5 CD25 0 10 10 10 10 0 10 10 10 10 CFSE B * *

103 Norm. CD25 MFI 102 Norm. CD44 MFI 103 LPA +/+ LPA –/– LPA +/+ LPA –/– 5 5 5 5

+ Figure 4.4 LPA5-deficient CD8 T cells exhibit reduced activation +/+ –/– + 6 Purified LPA5 or LPA5 OT-I CD8 T cells were CFSE labeled and 10 cells transferred to C57BL/6 recipients, as described in Fig. 4.2. Recipients were injected one day later with SIINFEKL-pulsed BMDC. Three days after BMDC transfer, mice were sacrificed and draining lymph nodes harvested for analysis. +/+ –/– (A) CFSE and CD25 expression of LPA5 (left plot) or LPA5 (right plot) OT-I CD8+ T cells. Representative of 2 independent experiments. (B) The normalized MFI of OT-I CD8+ T cells 3 days after peptide +/+ –/– stimulation is shown from individual LPA5 (closed circles) and LPA5 (open circles) OT-I CD8+ T cell recipients for CD25 (left panel) and CD44 (right panel) expression. Cumulative of 2 independent experiments (*p < 0.05).

+ LPA5-deficient CD8 T cells do not produce nor consume IL-2 differently

+ In the previous section, we showed that LPA5-deficient CD8 T cells exhibited enhanced proliferation but reduced CD25 expression as compared to 76 + LPA5-sufficent CD8 T cells. CD25 is the α chain of the high affinity IL-2 receptor trimer. The IL-2 receptor can bind IL-2 with a lower affinity when it is a dimer composed of the common γ chain and CD122. However when the IL-2 receptor is expressed as a trimer of the common γ chain, CD122, and CD25, the addition of

CD25 increases the affinity for IL-2 by a magnitude of 10- to 100-fold. Thus,

CD25 confers high affinity to the IL-2 receptor (173). Naïve CD8+ T cells are

CD25– and TCR signaling induces IL-2 production, which leads to autocrine IL-2 signaling resulting in internalization and degradation of IL-2, and increased CD25 expression (173, 180). As we had also seen a delay of CD25 upregulation in

+ LPA5-deficient CD8 T cells (Figure 4.4A), we questioned if the lack of LPA5 expression resulted in delayed IL-2 production or defective IL-2 receptor binding by IL-2.

+ To determine if LPA5-deficient CD8 T cells produced less IL-2, we

–/– +/+ + stimulated purified LPA5 and LPA5 OT-I CD8 T cells with SIINFEKL-BMDC in vitro, then collected supernatants and assayed for IL-2 production by ELISA.

We found no significant difference in IL-2 production at any timepoint tested

(Figure 4.5A). This showed that the LPA5 deficiency did not result in reduced IL-2 production to explain the delay in CD25 expression. We also wanted to test if

IL-2 consumption was different, as upon IL-2R binding, IL-2 is internalized and degraded, depleting the culture of exogenous IL-2. Thus we added IL-2 to the culture and measured internalization indirectly by the loss of IL-2 in the culture supernatant over time by IL-2 ELISA (181). Over a 3 day timecourse, we again

–/– +/+ did not find a difference in consumption of IL-2 by stimulated LPA5 and LPA5

77 OT-I CD8+ T cells (Figure 4.5B). Additionally, we did not find a difference in IL-2

consumption in the absence of stimulation (data not shown). This indicates that

–/– +/+ neither IL-2 production nor consumption are different in LPA5 and LPA5 OT-I

CD8+ T cells in vitro. However, we also did not see a delay in CD25 upregulation

–/– + in LPA5 OT-I CD8 T cells in response to in vitro stimulation as we did in vivo

(data not shown), which indicates differential CD25 upregulation between our in

vitro and in vivo experiments.

A B 10

WT + DC ) ) L L

m KO + DC m / / g g n n ( (

2 2 10 WT + low DC L L I I KO + low DC 1

1 2 3 1 2 3 d d d d d d Time Time

Figure 4.5 IL-2 production and consumption is similar by +/+ –/– + LPA5 and LPA5 CD8 OT-I T cells CD8+ T cells were purified from pooled splenocytes and lymph nodes from +/+ –/– 5 + LPA5 or LPA5 OT-I mice. 10 CD8 T cells were cultured at a 2:1 (“DC”, squares) or 10:1 (“low DC”, circles) ratio with SIINFEKL-BMDC in the absence (A) or presence (B) of 15 ng/mL rmIL-2. Supernatants were collected at the indicated timepoints and assayed for IL-2 by ELISA. Cumulative of 2 independent experiments.

–/– + Transfer of LPA5 OT-I CD8 T cells controls tumor growth in vivo

+ Our data thus far show that LPA signaling via LPA5 on CD8 T cells

impairs T cell activation and proliferation. Strikingly, a number of different tumor

and malignant cell types have been reported to produce elevated levels of LPA

that, in turn, promotes survival, growth and tumorigenesis (84-87). However the

effect of this elevated tumor-generated LPA on the adaptive immune response 78 has not been addressed. Our findings suggest that this increased LPA production may contribute to the immune suppression observed in the tumor

+ microenvironment. Thus, we next asked if LPA5-deficient tumor-specific CD8 T cells would be unaffected by the increased LPA concentration and thus be able

+ to better control tumor progression over LPA5-sufficient CD8 T cells.

Autotaxin was originally isolated from the ascites fluid of a human melanoma (70) and the B16 C57BL/6 melanoma cell line expresses functional autotaxin that can convert LPC to LPA (Gabor Tigyi, personal communication).

We elected to use a derivative of the B16 line: the B16.cOVA line expresses chicken ovalbumin constitutively and is thus able to present “tumor antigen” directly to OT-I CD8+ T cells. This allowed us to directly compare tumor-specific

+ responses of LPA5-deficient versus LPA5-sufficient CD8 T cells in the context of a melanoma. We decided to implant the tumor cells by subcutaneous injection, so that tumor progression could be tracked over time by measuring the diameter of the tumor. Preliminary experiments revealed that subcutaneous injection of

B16.cOVA cells in the leg led to consistent tumor growth that was palpable and measurable. Thus, B16.cOVA cells were implanted in the rear leg of C57BL/6 mice and allowed to establish for 5 days whereupon either naïve LPA5-deficient

+ or LPA5-sufficient tumor-specific CD8 T cells were transferred by intravenous injection (Figure 4.6).

79 d5 d0 d5 or d8

C57BL/6

Implant B16.cOVA SC Analyze tumor + Transfer CD8 IV OTI

LPA5/ or LPA5+/+ OTI (CD45 mismatched)

Figure 4.6 Schematic of tumor implantation and OT-I CD8+ T cell adoptive transfer 105 B16.cOVA cells were implanted s.c. in the leg of C57BL/6 recipients. 6 +/+ –/– Five days post-implantation, 0.5 x 10 LPA5 or LPA5 tumor-specific (OT-I) CD8+ T cells were transferred i.v. and tumor progression evaluated at the specified timepoints.

With this experimental design, we measured tumor diameter every 2 days

–/– + to determine if the transfer of LPA5 tumor-specific CD8 T cells could control tumor progression better than wild-type tumor-specific CD8+ T cells. We were excited to find that while tumors grew similarly in the absence or presence of wild-type tumor-specific CD8+ T cell transfer, tumor growth was abated in mice

+ that received LPA5-deficient tumor-specific CD8 T cells (Figure 4.7). Averaging tumor size across a treatment group revealed that while tumors were similar in size at the time of T cell transfer, a significant reduction in tumor size was

–/– + observed at 6 and 8 days post-transfer of LPA5 tumor-specific CD8 T cells compared to wild-type T cells (Figure 4.7B). These findings indicate that transfer

–/– + of LPA5 tumor-specific CD8 T cells is sufficient to cause abatement of tumor growth. 80 A No T cell transfer LPA5+/+ OTI LPA5/ OTI 15 15 15

10 10 10

5 5 5

0 0 0 umor diameter (mm)

T 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 Day posttransfer of CD8+ T cells

10 No transfer +/+ ** 8 LPA5 ** LPA / * 6 5

umor diameter (mm) 0

T 0 2 4 6 8 Day posttransfer of CD8+ T cells

+ Figure 4.7 LPA5-deficient OT-I CD8 T cells control B16.cOVA growth (A) B16.cOVA cells were implanted and tumor size is shown for individual mice at the indicated timepoints in the absence of CD8+ T cell transfer +/+ (gray circles; n=9) or after adoptive transfer of LPA5 (black circles, n=10) –/– + or LPA5 (open, n=11) OT-I CD8 T cells. Data are cumulative of 3 independent experiments. (B) Average combined tumor burden of Fig. 4.7A. Data are cumulative of 3 independent experiments (mean +/- SEM). Statistical significance was determined by Student’s T test at the indicated timepoint (*p<0.05, **p<0.01).

When we excised these tumors, we found that the tumor diameter measurements correlated with less vascularized, more compact tumors (Figure

+/+ + 4.8A). In contrast, tumors harvested from LPA5 tumor-specific CD8 T cell recipients and control mice were often penetrant into the local tissue and well vascularized. We did not observe metastases in the lungs by visual inspection in

–/– + any treatment group within this timeframe. Finally, the ability of LPA5 CD8 T 81 cells to better control tumor progression compared to wild-type CD8+ T cells was

further shown by the significantly smaller tumors harvested from mice that

+ received LPA5-deficient tumor-specific CD8 T cells as compared to the other

treatment groups (Figure 4.8B). Thus, these data show that tumor growth is

+ inhibited when LPA5 signaling is prevented on tumor-specific CD8 T cells.

A +/+ / B OTI: None LPA5 LPA5

2.0 ** 1.5 **

1.0

0.5 umor mass (g) T 0.0 No LPA5+/+ LPA5/ transfer A +/+ / B OTI: None LPA5 LPA5

2.0 ** 1.5 **

1.0

0.5 umor mass (g) T 0.0 No LPA5+/+ LPA5/ transfer

+ Figure 4.8 Transfer of LPA5-deficient CD8 T cells results in smaller tumors (A) Representative tumors in situ and ex vivo from C57BL/6 mice after 8 +/+ days in the absence of T cell transfer or post-adoptive transfer of LPA5 –/– + or LPA5 tumor-specific CD8 T cells. +/+ (B) Tumor mass at time of sacrifice after adoptive transfer of LPA5 OT-I –/– (black, n=10), LPA5 OT-I (white, n=11) or untreated controls (gray, n=9). Data are cumulative of 3 independent experiments (mean +/- SEM, **p<0.01). 82 Single transfer of LPA5-deficient OT-I T cells is not sufficient to ablate B16.cOVA tumor

–/– + Our initial results showed that injection of LPA5 tumor-specific CD8 T cells could control tumor growth within a week of transfer and we next questioned if these T cells could reject the tumor completely. Therefore, we repeated the experiment (Figure 4.6) except we continued the experiment beyond 8 days as a survival study and euthanized mice once the tumor reached a diameter greater

–/– than 1 cm. We found that in this experiment, transfer of neither LPA5 or wild- type tumor-specific CD8+ T cells conferred long-term protection from tumor burden (Figure 4.9A). Mice were euthanized due to tumor burden at similar timepoints as compared to mice that did not receive OT-I T cells. In one of the two experiments, excised tumors were analyzed for the presence of transferred

OT-I CD8+ T cells within the tumor, dLN, and ndLN. We did not detect remaining

OT-I CD8+ T cells in mice that were euthanized at later timepoints, regardless of

LPA5 expression (data not shown). This indicated to us that a single transfer of naïve T cells may be insufficient to clear a tumor and serial transfers may induce better tumor regression. This has been shown in clinical trials where transfer of tumor-specific T cell lines to patients resulted in detectable levels of T cells for only 14 days and only with IL-2 infusions to boost T cell survival (182).

We questioned if we could achieve better protection if we transferred

CD8+ T cells on the same day as tumor implantation, rather than allowing the tumor to establish before transferring CD8+ T cells. In this preliminary experiment, we did see extended protection with OT-I transfer, however there was no

83 –/– +/+ appreciable difference between the LPA5 and LPA5 OT-I recipients (Figure

4.9B). The survival curves of the OT-I recipients were influenced by one mouse per group that survived beyond 30 days post-T cell transfer (Figure 4.9C).

+ Interestingly, the LPA5-sufficient OT-I CD8 T cell recipient had a measurable tumor that remained at a consistent size for approximately 10 days, was undetectable on days 21-25, then resurfaced on day 27 and ultimately grew

+ greater than 1 cm. In contrast, one LPA5-deficient OT-I CD8 T cell recipient did not exhibit a measurable tumor until 31 days post-transfer of T cells. Although these data represent one independent experiment, they suggest that transfer of

OT-I CD8+ T cells simultaneous to the tumor allows better control of tumor progression, regardless of LPA5 expression.

A B C 100 100 15 80 80 10 60 60 40 40 5 20 20 Percent survival Percent survival

0 0 umor diameter (mm) 0 0 10 20 30 0 10 20 30 T 7 11 16 19 23 27 31 Day posttransfer of CD8+ T cells No transfer LPA5+/+ LPA5/

+ Figure 4.9 A single transfer of LPA5-deficient CD8 T cells is insufficient to protect mice from B16.cOVA tumors +/+ (A) Survival curve after adoptive transfer of LPA5 OT-I (black, n=6), –/– LPA5 OT-I (white, n=6) or untreated controls (gray, n=5) 5 days after tumor transfer. Data are cumulative of 2 independent experiments (p=ns, Mantel-Cox log rank test). +/+ (B) Survival curve after adoptive transfer of LPA5 OT-I (black, n=3), –/– LPA5 OT-I (white, n=3) or untreated controls (gray, n=3) simultaneous to tumor transfer. 1 experiment (p=ns, Mantel-Cox log rank test). (C) Tumor size is shown at the indicated timepoints for individual mice in Fig. 4.9B.

84 –/– + LPA5 OT-I CD8 T cells accumulate within the tumor

+ Thus far, we have shown that LPA5-deficient CD8 T cells control

+ established tumor progression better than LPA5-sufficient CD8 T cells in a short- term experiment. We next questioned if this enhanced tumor control was due to

–/– increased proliferation and accumulation of transferred LPA5 tumor-specific T cells, as we had observed in previous experiments.

We first performed various experiments to examine T cell proliferation within the tumor and dLN at 3, 5, and 8 days post-T cell transfer, by transferring purified CFSE-labeled OT-I CD8+ T cells to tumor-bearing recipients, as before described (Figure 4.6). From these studies, we found that there was a trend of

+ increased proliferation of LPA5-deficient OT-I CD8 T cells relative to LPA5- sufficient OT-I CD8+ T cells in the tumor and dLN by 3 days post-transfer, though this difference was not significant (Figure 4.10). By day 5 post-transfer, roughly

–/– +/+ + the same number of LPA5 and LPA5 OT-I CD8 T cells had proliferated in the tumor and dLN (Figure 4.10) and by day 8, both populations of OT-I CD8+ T cells had divided completely in the tumor and were CFSE negative (data not shown).

–/– + This suggested that perhaps LPA5 OT-I CD8 T cells were capable of more rapid proliferation after transfer to tumor-bearing recipients, however by day 5,

+/+ + LPA5 OT-I CD8 T cells had proliferated to a similar degree.

A d3 posttransfer d5 posttransfer 60 80

60 40 40 20 % divided % divided 20

0 0 +/+ / LPA +/+ LPA / LPA5 LPA5 5 5 B 20 25

15 20 15 10 10 % divided % divided 5 5 0 0 LPA +/+ LPA / LPA +/+ LPA / 5 5 5 5

Figure 4.10 Proliferation of LPA5-deficient OT-I T cells trends to be increased at day 3 105 B16.cOVA cells were implanted s.c. in the leg of C57BL/6 recipients 6 + +/+ as in Fig. 4.6. Five days post-implantation, 0.5 x 10 CD45.1 LPA5 –/– + (closed circles) or LPA5 (open circles) OT-I CD8 T cells were purified, CFSE labeled, and transferred by intravenous injection to tumor-bearing recipients. Mice were euthanized at 3 or 5 days post-transfer and tumors were assessed for proliferation of OT-I CD8+ T cells by CFSE dilution. (A) Proliferation of OT-I T cells in the tumor at 3 (left panel, p=0.3) or 5 (right panel, p=0.2) days post-transfer. Data are cumulative of 2 independent experiments. (B) Proliferation of OT-I T cells in the dLN at 3 (left panel, p=0.2) or 5 (right panel, p=0.5) days post-transfer. Data are cumulative of 3 (d3) or 2 (d5) independent experiments.

86 We next assessed if this modest difference in proliferation would result in differential accumulation of OT-I CD8+ T cells in the tumor. When we assessed

OT-I CD8+ T cell accumulation as a proportion of infiltrating CD8+ T cells, we

+ found a slight increased presence of LPA5-deficient OT-I CD8 T cells at earlier timepoints (d3, d5) but not at 8 days post-transfer (Figure 4.11A). When we quantified the concentration of OT-I CD8+ T cells per gram of tumor, we again found a trend of increased number of LPA-deficient OT-I CD8+ T cells in the tumor at 3 and 5 (p=0.15) days post-transfer (Figure 4.11B). However in contrast,

+ by day 8 the concentration of LPA5-deficient OT-I CD8 T cells was roughly

+ equivalent to the concentration of LPA5-sufficient OT-I CD8 T cells in the tumor,

–/– + though LPA5 OT-I CD8 T cells trended to be a lesser proportion of the infiltrating CD8+ T cells. Taken together, this indicates that although transferred

+/+ –/– + tumor-specific LPA5 and LPA5 CD8 T cells are present at similar

–/– + concentrations by day 8, transferred LPA5 tumor-specific CD8 T cells represent a smaller fraction of the total infiltrating population of CD8+ T cells.

Thus, an increase of endogenous CD8+ T cells in the tumor is also observed in

–/– + LPA5 OT-I CD8 T cells recipients, resulting in a significantly increased percentage of total CD8+ cells within the tumor (Figure 4.11C). This suggests that the presence of transferred LPA5-deficient tumor-specific T cells alters the tumor microenvironment to allow infiltration of endogenous CD8+ T cells. This phenomenon could be further probed by examination of the endogenous CD8+ T cells within the tumor for tumor specificity and anti-tumor function by cytokine staining, to determine the potential contribution of host CD8+ T cells to anti-tumor

87 immunity. In addition, the contribution of the transferred OT-I CD8+ T cells versus the endogenous T cell response could be further probed by transfer of OT-I CD8+

T cells into tumor-bearing T cell-deficient or TCR non-specific recipients, to prevent an endogenous anti-tumor CD8+ T cell response. However, we already know that any potential contribution of endogenous CD8+ cells is insufficient to

–/– + clear the tumor, even when combined with a single transfer of LPA5 OT-I CD8

T cells (Figure 4.9).

When we examined the activation status of the transferred OT-I CD8+ T cells, we found that CD25 was not upregulated and CD44 upregulation was only

+/+ –/– + modestly detectable with no difference between LPA5 and LPA5 OT-I CD8 T cells in the tumor and the dLN by day 5 (data not shown). We returned to in vitro

+ experiments to compare additional activation markers in LPA5-deficient CD8 T cells as compared to wild-type CD8+ T cells. To determine differences in activation and homing marker expression in vitro, we pulsed C57BL/6

–/– splenocytes with SIINFEKL and used them to stimulate purified LPA5 or

+/+ + LPA5 OT-I CD8 T cells in vitro. In a preliminary experiment, we did not see a difference in PD-1, CCR7, nor CD127 expression between the two groups (data not shown), similar to previous findings (149). We thus next questioned if the

+/+ –/– + effector function was different between LPA5 and LPA5 tumor-specific CD8

T cells.

88 d3 posttransfer d5 posttransfer d8 posttransfer A 0.8 2.0 5 0.6 1.5 4 3 0.4 1.0 2 0.2 0.5 1 % OTI of CD8+ 0.0 % OTI of CD8+ 0.0 % OTI of CD8+ 0 +/+ / LPA +/+ LPA / LPA +/+ LPA / LPA5 LPA5 5 5 5 5 B ) ) ) 5 5 5 1.5 20 10 15 1 10 5 0.5 5 0 0 0 # OTI / g of tumor (x10 # OTI / g of tumor (x10 # OTI / g of tumor (x10 +/+ / LPA +/+ LPA / LPA +/+ LPA / LPA5 LPA5 5 5 5 5

d8 posttransfer

C 25 * 20 15 10 % CD8+ 5 0 OTI T cell transfer: None LPA +/+ LPA / 5 5

Figure 4.11 Trend of slightly increased concentration of + LPA5-deficient OT-I CD8 T cells in the tumor at 3 and 5 days post-transfer C57BL/6 recipients were implanted s.c. in the leg with 105 B16.cOVA cells +/+ –/– as in Fig. 4.6. Data shown are from LPA5 (closed circles) or LPA5 (open circles) OT-I CD8+ T cell recipients. (A) Percent OT-I T cells of CD8+ T cells in the tumor was assessed at 3 (left panel, p=1.0), 5 (center panel, p=0.3), or 8 (right panel, p=0.09) days post-transfer. Data are cumulative of 4 (d3) or 2 (d5, d8) independent experiments (mean +/- SEM). (B) Number of OT-I per gram of tumor was assessed at 3 (left panel, p=0.4), 5 (center panel, p=0.15), or 8 (right panel, p=0.9) days post- transfer. Data are cumulative of 3 (d3) or 2 (d5, d8) independent experiments (mean +/- SEM). (C) Percentage of tumor cells that are CD8+ in tumor-bearing recipients 6 + +/+ that received no OT-I T cells (gray circles), or 0.5 x 10 CD45.1 LPA5 –/– + (closed circles) or LPA5 (open circles) OT-I CD8 T cells. Cumulative of 2 independent experiments (*p<0.05, mean +/- SEM).

89 –/– LPA5 T cells kill target cells and produce intracellular cytokines similarly to wild- type T cells

Although we were unable to detect a difference in activation, we next

–/– + asked if LPA5 CD8 T cells were more adept at generating an effector response. We first addressed this question by conducting an in vivo cytotoxicity assay, by transfer of control and SIINFEKL-pulsed target cells into tumor bearing

+ recipients that had received either LPA5-deficient or –sufficient OT-I CD8 T cells

(Figure 4.12A). Target cells were distinguished from host cells by congenic marker expression (CD45.1) and from each other by CFSE staining (CFSEhigh versus CFSElow). Killing was assessed by the loss of SIINFEKL-pulsed target cells relative to control cells that did not express SIINFEKL. Having observed

+ smaller tumors in LPA5-deficient OT-I CD8 T cells recipients, we predicted that

–/– + LPA5 OT-I CD8 T cells recipients would feature a higher level of killing of target cells, however we were surprised to see that there was no difference in killing, in either the tumor (Figure 4.12C) or the dLN (Figure 4.12B). A higher proportion of SIINFEKL-pulsed targets were killed in the dLN, likely due to the increased number of CD8+ T cells in the dLN. This shows that SIINFEKL-labeled

+/+ –/– + target cells are equally cleared in tumor-bearing LPA5 and LPA5 OT-I CD8 T cell recipients by day 5 after T cell transfer.

90 A d-5 d0 d4 d5

C57BL/6

Implant B16.cOVA SC Transfer 5-10x106 target cells i.v.

200 60 Killing 150

40 Cells

Cells 100 # LPA5-/- or # 20 50 33.3 LPA5+/+ OT-I 49.1 SIINFEKL 0 0 2 3 4 5 2 3 4 5 0 10 10 10 10 0 10 10 10 10 : CFSE : CFSE pulsed CFSE CFSE

B C dLN Tumor

200 30 300 20

150 15 20 200 s s s +/+ s Cell Cell Cell

100 Cell

# #

LPA5 #

10 #

100 10 50 39.7 33.3 5 46.8 45.7

0 0 0 0 0 102 103 104 105 0 102 103 104 105 2 3 4 5 Control 0 10 10 10 10 0 102 103 104 105 : CFSE : CFSE : CFSE 400 40 : CFSE 300 80

60 300 30 60 200 s s s s 40 / s Cell

Cell LPA5 Cell Cell Cell 20

200

# 40 # # # #

100 20 100 10 49.1 37 33.5 44.6 20 48.1

0 0 0 0 0 2 3 4 5 0 10 10 10 10 2 3 4 5 2 3 4 5 2 3 4 5 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 0 102 103 104 105 : CFSE : CFSE : CFSE : CFSE CFSE

Figure 4.12 Killing ability is similar in the dLN and tumor of –/– +/+ LPA5 and LPA5 OT-I T cell recipients 5 days after T cell transfer (A) Schematic of in vivo cytotoxicity assay. C57BL/6 recipients were implanted s.c. in the leg with 0.5 x 105 B16.cOVA cells. Five days post- 6 +/+ –/– + implantation, 10 CFSE-labeled LPA5 or LPA5 OT-I CD8 T cells were transferred i.v. C57BL/6 splenocytes were divided to two populations and one population was labeled CFSEhigh and pulsed with SIINFEKL and the second population was labeled CFSElow and not pulsed with peptide. The CFSEhigh and CFSElow cells were combined at a 1:1 ratio and injected i.v. 1 day before harvest. +/+ –/– (B-C) The ability of LPA5 (upper row) and LPA5 OT-I (lower row) to kill target cells (CFSEhigh) is shown in the dLN (B) and tumor (C). Each histogram represents one C57BL/6 tumor-bearing recipient and control histogram is from a naïve C57BL/6 recipient. Representative of 2 experiments. 91

Although we were unable to detect a difference in the killing ability of

+ LPA5-deficient and LPA5-sufficient OT-I CD8 T cells in vivo, we wanted to test if the effector cytokine generation profiles were affected by signaling via LPA5. To address this question, we pulsed purified OT-I CD8+ T cells and C57BL/6 splenocytes with SIINFEKL peptide and then removed unbound peptide and cultured the cells for 3 days. We then further cultured cells with brefeldin A for 5 hours to induce intracellular cytokine accumulation. By this method, we did not detect a significant difference in intracellular TNFα production between LPA5-

+ deficient and LPA5-sufficient OT-I CD8 T cells but interestingly, found a slight decrease in intracellular IFNγ staining, both by percentage and MFI of the entire

CD8+ OT-I CD8+ T cell population (Figure 4.13). Although we would have

+ predicted that LPA5-deficient CD8 T cells would produce a greater level of

+/+ –/– + effector cytokines, the similarities between LPA5 and LPA5 CD8 cytokine production was congruous with our previous in vitro findings that LPA signaling does not potently affect IFNγ and TNFα production (Figure 3.12). Taken together with our findings from activation studies, this suggests that LPA inhibition may operate over initial antigen receptor signaling, activation, and proliferation of

CD8+ T cells, however once CD8+ T cells are activated, the effector function is not largely impacted by LPA signaling.

92 Unstim. control LPA +/+ LPA –/– A 5 5

105 105 105 0.054 22.2 15.1 104 104 104

103 103 103

102 102 102

0 0 0 γ

2 3 4 5 2 3 4 5 2 3 4 5

IFN 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 CD8 B

105 0.27 105 44.9 105 43.7

104 104 104

103 103 103

102 102 102

α 0 0 0

2 3 4 5 2 3 4 5 2 3 4 5 TNF 0 10 10 10 10 0 10 10 10 10 0 10 10 10 10 CD8

C D IFNγ MFI TNFα MFI 440 400 420 300 400 380 200 360 100 340 320 0 LPA +/+ LPA –/– LPA +/+ LPA –/– 5 5 5 5

–/– Figure 4.13 LPA5 OT-I T cells produce slightly less IFNγ but +/+ equivalent TNFα as LPA5 OT-I T cells + +/+ –/– CD8 T cells were purified from LPA5 or LPA5 OT-I splenocytes and lymph nodes, and combined at a 1:2 ratio with C57BL/6 splenocytes before pulse with SIINFEKL peptide. Cells were washed and plated at 2.5 x 106 cells/mL. On day 3, cells were transferred to a round-bottom plate and brefeldin A was added for 5 hours. Cells were then fixed, permeablized, and stained for IFNγ or TNFα and were analyzed by flow cytometry. Data shown is from gated OT-I CD8+ T cells. (A-B) Representative staining of IFNγ (A) or TNFα (B) staining with % postitive indicated. Representative of 2 independent experiments. +/+ –/– (C-D) MFI of IFNγ (A) or TNFα (B) staining of LPA5 or LPA5 OT-I CD8+ T cells (n=2). 93 Conclusion

In the previous chapter, we found that increased LPA signaling via LPA5 was sufficient to inhibit CD8+ T cell activation and proliferation. We acquired

LPA5-deficient mice and found them to be phenotypically unremarkable with similar lymphocyte populations to wild-type littermates. Two other groups have

–/– recently published studies with independently generated LPA5 mice (115, 183) and Lin et al. also reported similar lymphocyte populations. When we compared

–/– + + the activation of LPA5 CD8 T cells with wild-type CD8 T cells by calcium

–/– + mobilization assay, we found that LPA5 CD8 T cells were not inhibited by LPA

(Figure 3.5). Thus we hypothesized that as we had shown that increased LPA

+ signaling inhibited the CD8 T cell response via LPA5, ablating signaling to that receptor would negate that inhibition and allow an enhanced CD8+ T cell response.

To test this hypothesis, we transferred LPA5-deficient or –sufficient OT-I

CD8+ T cells into C57BL/6 mice and stimulated the transferred cells by subsequent transfer of SIINFEKL-BMDC. We were initially unsure if we would see a significant difference in the CD8+ T cell response, as we were not increasing the concentration of LPA by OTP injection and we were stimulating

OT-I CD8+ T cells with the high affinity peptide, SIINFEKL, which we had previously shown to be less affected by LPA signaling. However we found that as

–/– + predicted from in vitro experiments, LPA5 CD8 T cells proliferated and accumulated to a greater degree than their wild-type counterparts. This indicates that the low concentration of LPA found in the mouse (1 µM in plasma) is

94 sufficient to dampen, though not ablate proliferation if OT-I CD8+ T cells in response to SIINFEKL peptide stimulation. However the concentration of LPA in lymph nodes is unknown and we cannot exclude the possibility that transfer of mature SIINFEKL-BMDC may alter the concentration of LPA.

We were surprised to observe that activation marker upregulation was not

+ increased in LPA5-deficient OT-I CD8 T cells but in contrast, was delayed in

+ –/– comparison to wild-type OT-I CD8 T cells. Further, it was curious that LPA5

OT-I T cells could proliferate for a few rounds of division in the absence of expression of the high affinity chain of the IL-2 receptor, CD25 (Figure 4.4). We tested if IL-2 production or consumption was different in vitro but we could not find any dissimilarities (Figure 4.6). However when we examined the CD8+ T cells in this experiment, we also did not see a difference in CD25 expression, which indicated that the in vitro experiment did not recapitulate what we had previously observed in vivo. Thus the reason for the reduced CD25 expression

–/– observed in LPA5 OT-I T cells in response to SIINFEKL-BMDC stimulation remains to be elucidated.

+ With the finding that LPA5-deficient CD8 T cells featured enhanced proliferation and accumulation, we wanted to test how LPA signaling would affect the CD8+ T cell response to tumor. It has been well reported that a variety of tumor and malignant cell types produce elevated levels of LPA that promotes tumorigenesis by enhancing survival, growth, and angiogenesis (84-87). As B16

+ melanomas express functional ATX, we tested how LPA5 signaling by OT-I CD8

T cells influenced control of tumor growth of SIINFEKL-expressing melanoma,

95 –/– + with the prediction that LPA5 OT-I CD8 T cells would have better anti-tumor capabilities in the presence of pathogenic levels of LPA.

To test this, we implanted B16.cOVA cells s.c. in mice and allowed the

–/– +/+ tumor to establish for five days before transfer of naïve LPA5 or LPA5 OT-I

CD8+ T cells. As we had previously shown that LPA inhibition was more potent with a less robust stimulation, we were unsure if we would detect a difference

–/– + with LPA5 OT-I CD8 T cells in response to a high affinity peptide with tumor stimulation. Impressively, tumor-specific CD8+ T cells transferred to recipients with established subcutaneous melanoma tumors were able to significantly control tumor growth when unable to signal via the LPA5 receptor. This resulted in much smaller tumors when excised that trended to have a higher concentration of OT-I CD8+ T cells. This indicates that with stimulation to a high affinity antigen in the context of a tumor, LPA signaling is potent enough to

+ impact the CD8 anti-tumor response, and ablating LPA5 signaling allows greater

T cell accumulation and control of tumor progression.

+/+ –/– + When we examined LPA5 and LPA5 OT-I CD8 T cells, we did not find significant differences in activation or effector cytokine production. However we did see a trend of increased proliferation and concentration of LPA5-deficient

OT-I CD8+ T cells in the tumor at earlier timepoints. This could indicate that

+ tumor-specific CD8 T cells lacking LPA5 are able to more quickly expand and control tumor growth than wild-type OT-I CD8+ T cells.

–/– While we observed a decrease in tumor growth with transfer of LPA5

OT-I T cells, we did not see an effect with transfer of wild-type OT-I T cells.

96 Although we were surprised that wild-type OT-I CD8+ T cells did not affect tumor growth, published studies showing a decrease in B16.cOVA tumor burden with wild-type OT-I T cell transfer utilized in vitro generated effector cells and/or transferred OT-I T cells earlier than 5 days post-transfer of tumor (184, 185).

When we transferred OT-I CD8+ T cells on the same day as tumor, we achieved

+/+ –/– + similar tumor control by both LPA5 and LPA5 OT-I CD8 T cells. This indicates that the timing and activation state of transferred CD8+ T cells can have a profound effect of the outcome of this experiment. As our initial experimental design more closely mimics a tumor therapeutic strategy, this suggests that

+ dampening of LPA5 signaling on CD8 T cells could have a beneficial anti-tumor effect with established tumors. Consistent with these findings, our collaborators

–/– have found that B16 melanomas metastasize in wild-type mice, but not in LPA5 mice (Gabor Tigyi, personal communication). Although in their experiments the

+ LPA5 deficiency is not restricted to CD8 T cells, the correlation with our results is tantalizing to suggest exploration of a targeted therapeutic.

CHAPTER V

DISCUSSION AND FUTURE DIRECTIONS

A model for LPA inhibition of CD8+ T cells

Adaptive immunity is the specialized arm of the immune response that locates, targets, and facilitates the removal of foreign invaders in a specific manner. CD8+ T cells in particular are important for their cytotoxic ability to identify and kill infected and altered self cells, such as cancer cells. With such a powerful subset of cells, a variety of mechanisms exist to modulate the T cell response and temper activation of T cells to minimize or prevent damage to the host. It has been well documented that the lysophospholipid, LPA, has a variety of effects on different cell types and that LPA receptors are present on immune cells (57, 87), however how LPA affects the adaptive immune response has not been thoroughly addressed. We have previously shown that LPA functions to inhibit B cell activation and antibody production. Given the signaling similarities between B and T cells, we questioned if LPA would also inhibit the CD8+ T cell response, with the hypothesis that increased LPA signaling would decrease

CD8+ T cell activation and effector function. Finally, we wanted to test if this inhibition would affect an immune response to a known condition that increases

LPA concentrations within individuals; cancer.

In chapter III, we demonstrate that, similar to B cells, LPA potently dampens T cell signaling via inhibition of intracellular calcium release and LPA5.

98 We questioned if this defect in TCR signaling would translate to decreased activation and proliferation in response to stimulation. Initial experiments revealed that LPA inhibition was most potent with sub-optimal levels of anti-CD3 stimulation, which activates T cells by cross-linking the CD3 subunit of the T cell signaling complex. To use a more physiological stimulus we elected to use peptides to stimulate CD8+ T cells from OT-I mice, a well-characterized TCR transgenic model. By use of this model, we used different affinity peptides to show that LPA more potently inhibited stimulation by a lower affinity peptide (G4) than a high affinity peptide (SIINFEKL) (Figures 3.8, 3.9, 3.11), which coincides with a previous finding of increased impact of LPA induction of IL-13 production by CD4+ T cells with sub-optimal stimulation (147). Considering these findings, we propose a model where the potency of LPA inhibition is dependent on the strength of stimulation (Figure 5.1). A strongly activated CD8+ T cell, such as an

OT-I CD8+ T cell activated by SIINFEKL peptide, will proliferate and accumulate robustly in the absence of LPA. In the presence of LPA, CD8+ T cell proliferation is dampened but not ablated, resulting in decreased accumulation. In contrast, a weaker activation, such as OT-I CD8+ T cells activated by SIIGFEKL (G4) peptide, will result in reduced proliferation as compared to activation by a strongly activating peptide. In the presence of LPA, proliferation of CD8+ T cells as a result of weaker activation is ablated, preventing accumulation and expansion of the CD8+ T cell population.

99 #+,-.(/01"(( 2345(0(6.77(

02:(;<=>+7(;?@.>=?A(

)*( !""#$%&'(

5('89( 1('89( 5('89( 1('89(

Figure 5.1 A model for LPA regulation of CD8+ T cell activation to different affinity peptides We propose that when a naïve CD8+ T cell is activated via strong TCR signal, such as a high affinity peptide, proliferation is slightly dampened by the presence of LPA. In contrast, when a naïve CD8+ T cell is activated by a weaker TCR signal, LPA signaling potently inhibits proliferation.

This concept of reduced influence of a TCR co-receptor with strong TCR stimulation is not entirely new, as it has been shown that strong TCR stimulation can negate the need for positive co-stimulation by the canonical T cell co- receptor, CD28, whereas weak TCR stimulation requires CD28 signaling for T cell activation (13, 14). However in the case of inhibitory receptors, increased strength of TCR signaling does not always equate with reduced inhibition, as increased TCR signaling has been shown to increase the accumulation of the T cell inhibitory receptor, CTLA-4, at the immunological synapse, such that T cells

100 receiving stronger stimuli are more sensitive to CTLA-4 signaling (186). By this mechanism, strongly activated T cells are also more susceptible to CTLA-4- mediated attenuation of the T cell response and this provides a rapid method of halting the T cell response. In contrast, we have found that strongly activated T cells are relatively resistant to LPA inhibition. This presents a novel pathway by which signaling by a lysophospholipid is an inhibitory rheostat for T cell activation and suggests that in contrast to CTLA-4 that is upregulated as an “emergency brake” to quell T cell responses, LPA signaling functions to temper responses starting from initiation of T cell signaling. This signaling could prevent weakly activated T cells, possibly from inappropriate autoreactive activation by self tissue, from a robust proliferative response. In support of this theory,

–/– collaborators have found exacerbated disease development in LPA5 mice in an induced autoimmune model (Tamas Oravecz, personal communication). As earlier described, extracellular levels of LPA are tightly regulated, and this could indicate a delicate system of immune control that has evolved to prevent unwanted immune responses.

CTLA-4–/– mice feature a lymphoproliferative disorder characterized by T cell proliferation and infiltration into non-lymphoid tissues, and become moribund

–/– by 3-4 weeks of age (187-189). We have not found evidence of this in LPA5 mice within this timeframe. This could indicate that in contrast to CTLA-4, inhibition via LPA5 is not important for regulation of T cell homeostasis. In consideration of our calcium mobilization assays demonstrating that LPA treatment in the absence of antigen receptor stimulation has no effect on calcium

101 mobilization, inhibition through LPA5 may only occur in conjunction with antigen receptor signaling. This would indicate that LPA signaling has little influence on the status of resting T cells but functions to curb the T cell response during an immune event. However TCR signaling does occur during thymic development, as T cells are tested for self-reactivity, thus it would be interesting to determine

LPA5 expression on developing thymocytes and if LPA signaling affects T cell development. If LPA5 is present on thymocytes and functions similarly to dampen

–/– TCR signaling, we would predict that LPA5 mice would feature a different T cell repertoire. Increased LPA signaling during thymic development would dampen

TCR signaling on potentially autoreactive T cells, allowing these autoreactive T cells to enter the periphery and possibly induce autoimmunity.

This model (Figure 5.1) is particularly interesting when considered in the context of cancer. High affinity tumor-specific T cells have by definition a degree of self-reactivity since they are responding to modified self tissue and thus have been reported to be deleted during thymic education. Therefore T cells specific for tumor antigens typically display low affinity for these ‘self tumor antigens’ (44-

47). Our findings would indicate that LPA signaling would thus have a more potent effect on these tumor-specific T cells that are stimulated by only lower affinity antigens. As LPA production appears to be a hallmark for a variety of cancer cell types (84-87), this would provide an additional mechanism to explain why it is difficult to generate an effective immune response to tumor.

102 LPA signaling in CD8+ T cells in vivo: implications for enhancing the

CD8+ T cell response

We next tested if the in vitro findings of LPA inhibition of CD8+ T cell activation would translate to in vivo experiments. We continued to use the OT-I model by transferring purified OT-I CD8+ T cells into C57BL/6 mice and stimulating these cells with transfer of peptide-pulsed bone marrow-derived dendritic cells (BMDC). By use of congenic markers, we were able to distinguish transferred OT-I T cells from the endogenous T cells and determine the effects of increased LPA signaling by treating with a stable LPA analog, OTP. By this method, we were surprised that we were unable to achieve activation to the lower affinity G4 peptide in the context of BMDC, in contrast to our in vitro results.

In a similarly designed study, Zehn et al. achieved in vivo activation of transferred

OT-I to a variety of peptides by the generation of recombinant Listeria monocytogenes (Lm) strains containing SIINFEKL or altered SIINFEKL-derived peptides (23). The panel of peptides included lower affinity peptides as low as

680-fold less potent over SIINFEKL (SIIVFEKL) and yet they observed a degree of expansion to each lower affinity peptide tested. We found that in contrast to

Lm-OVA peptide infection that induced activation of T cells in various lymphoid organs, our SIINFEKL-BMDC transfer resulted in localized activation isolated to the draining lymph node. This limited activation by peptide-BMDC transfer has been exploited by others as a transient introduction of antigen (176). Therefore we estimate that the peptide presentation “vector” of BMDC was insufficient to activate transferred OT-I CD8+ T cells to lower affinity peptides but Lm is a

103 stronger vector that can be used to compare lower affinity peptides in vivo. We predict that immunization with different Lm-OVA peptide strains would reveal increased inhibition by LPA signaling over weaker affinity peptide stimulation.

However we also expect that immunization with Lm-OVA peptides would shift activation such that a moderately activating peptide when presented by BMDC would be strongly activating within the Lm construct. As a result, a lower affinity

Lm-peptide would induce proliferation in transferred OT-I CD8+ T cells and this proliferation may be ablated or at least strongly inhibited by OTP treatment, as we observed in vitro with G4 peptide stimulation (Figure 3.8).

In chapter IV, we explored how the CD8+ T cell response is affected by genetic deletion of LPA5. We predicted that without the inhibition via LPA5, these

T cells would exhibit an enhanced response to stimulation. We repeated the experimental design from the in vivo experiments in chapter III, except that the

+ +/+ –/– transferred OT-I CD8 T cells were either LPA5 or LPA5 . In contrast to the experiments in chapter III when we increased LPA signaling to all cells by OTP administration, we relied on endogenous sources of LPA, which have been reported to be in the high nanomolar to low micromolar range (63, 117, 163). In addition, we focused the effect of LPA signaling to the transferred OT-I CD8+ T cells by modulating expression of LPA5, allowing us to make cell intrinsic comparisons. The results from these experiments revealed that, as predicted,

+ LPA5-deficient OT-I CD8 T cells trended to have a higher proliferative capacity and thus a near significant increase in OT-I CD8+ T cell accumulation. We were

+/+ not surprised that the difference in proliferation was not greater between LPA5

104 –/– and LPA5 as in vitro experiments showed that LPA inhibition of activation to

SIINFEKL was diminished as compared to a lower affinity peptide. In addition, the concentration of LPA in plasma is low and has been reported to be ~1µM in mouse (150, 168), although we do not know the concentration of LPA in lymph nodes or if it is changed by SIINFEKL-BMDC transfer. However in comparison to another major lysophospholipid, S1P, it might be expected that the concentration of LPA in tissues is reduced relative to blood (190, 191).

+ We found that LPA5-deficient OT-I CD8 T cells were able to proliferate and accumulate more robustly than wild type OT-I CD8+ T cells, resulting in few cells that had not proliferated and increased accumulation of OT-I CD8+ T cells in

+ the draining lymph node. Interestingly, although the LPA5-sufficient OT-I CD8 T cell population had more cells that had not proliferated and were CFSEhigh, the cells that had proliferated were able to divide roughly the same number of

–/– + divisions as LPA5 OT-I CD8 T cells. This indicated to us that inhibition via

LPA5 by endogenous levels of LPA was sufficient to prevent proliferation in a

high subset of the transferred cells and thus remained CFSE . However, the LPA5-

+ sufficient CD8 T cells that could be activated despite LPA inhibition via LPA5

+ were able to proliferate as efficiently as LPA5-deficient CD8 T cells. This is further evidenced by the similar shape of the proliferation profiles after the first

+/+ –/– + division between LPA5 and LPA5 OT-I CD8 T cells (Figure 4.3). This suggests that LPA functions to inhibit CD8+ T cell activation at an early event, however those cells that are activated despite the presence of LPA, are not inhibited from further proliferation.

105 +/+ –/– When we examined activation between LPA5 and LPA5 OT-I T cells,

–/– we were surprised to find that LPA5 OT-I T cells had lower levels of activation marker expression (CD25 and CD44). When we compared CD25 expression with proliferation by analyzing CFSE dilution, it appeared that there was a delay in

+ CD25 upregulation in LPA5-deficient OT-I CD8 T cells, such that earlier divisions occurred in the absence of CD25 upregulation and only later divisions had

+ upregulated CD25. This was in contrast to LPA5-sufficent OT-I CD8 T cells that featured increased CD25 expression with successive divisions. As CD25 is part of the IL-2 receptor complex and is upregulated by IL-2 signaling, we questioned

+ if LPA5-deficient OT-I CD8 T cells produced IL-2 differently from LPA5-sufficient

OT-I CD8+ T cells. Although we did not find a difference in IL-2 production or consumption in vitro, we also did not see a difference in CD25 expression as we had observed in vivo, therefore we question the relevance of these findings in comparison with the in vivo experiments. We therefore decided to further characterize the phenotype of responding LPA5-deficient versus LPA5-sufficient

CD8+ T cells in vivo and look for differences in the kinetics of the response. The differences in effector versus memory population generation in response to infection have been well characterized and include differences in surface markers and cytokine production, including KLRG1 and granzyme B (Figure 5.2).

Although we know that LPA5-sufficient OT-I T cells are capable of differentiating to both memory and effector populations (192), we questioned if the increased proliferation we observed with LPA5-deficient OT-I T cells was an indication that a

+ lack of LPA5 signaling would skew the differentiation of responding CD8 T cells

106 to an effector phenotype, which is characterized by a robust proliferative response (193).

' !"# (&(' *++,-./!' $ ' )'

%&%' !"#$

+ Figure 5.2 Does LPA5 signaling influence CD8 T cell differentiation? Upon infection, CD8+ T cells that develop into long-term memory T cells express KLRG1low and granzyme Blow, whereas terminally differentiated effector T cells express KLRG1high and granzyme Bhigh. We question if LPA5 signaling could influence the ratio of these outcomes. Figure adapted from Belz, 2010 (193).

An initial experiment to characterize differences in differentiation was not successful, as the effector versus memory phenotype markers analyzed were not upregulated in response to SIINFEKL-BMDC activation. This was also documented by another group that found that transfer of DC with Listeria monocytogenes (Lm) but not DC alone, induced KLRG-1 (194). However we did

–/– + +/+ observe an increased population of LPA5 OT-I CD8 T cells relative to LPA5

OT-I CD8+ T cells at an earlier timepoint (7 days) and this was reversed at a later timepoint (14 days, Figure 5.3). Although preliminary, this may indicate that a

107 + lack of LPA5 signaling influences CD8 T cells towards an effector phenotype.

–/– + However as we have shown that LPA5 CD8 T cells exhibit greater proliferation in response to stimulation, this may indicate that mutant T cells may also exhibit an earlier contraction as the antigen is more rapidly cleared by a larger population of specific CD8+ T cells. This could be tracked with an infection model by quantifying the infectious burden at different timepoints to determine if the

–/– + infection is cleared more rapidly by LPA5 OT-I CD8 T cells and thus followed by an earlier contraction.

Day 7 Day 14 10 10

+ 8 + 8

6 6 of CD8 of CD8

4 4 % OTI % OTI 2 2

0 0 Unstimulated SIINFEKLBMDC Unstimulated SIINFEKLBMDC LPA5+/+ / LPA5

–/– Figure 5.3 LPA5 T cells may have decreased memory population formation or an earlier contraction +/+ –/– + A preliminary experiment in which LPA5 or LPA5 CD8 OT-I T cells were isolated and 106 CFSE-labeled cells were adoptively transferred intravenously to C57BL/6 recipients a day prior to immunization. On day 0, 106 SIINFEKL-pulsed BMDC were transferred by subcutaneous injection to the scruff, similar to Fig. 4.2. On days 7 and 14 after transfer of peptide- loaded BMDC, mice were sacrificed and the percentage of OT-I of CD8+ T cells in the dLN was assessed.

108 + To further investigate the differentiation of LPA5-deficient OT-I CD8 T cells, we examined effector molecule expression, including granzyme B and

Lamp-1 (CD107a). Lamp-1 is a marker of degranulation and granzyme B causes apoptosis in target cells and is produced at high levels in effector T cells but at lower levels in memory T cells (Figure 5.2) (159, 193). When we stimulated

+ purified LPA5-deficient OT-I CD8 T cells in vitro with G4 peptide-pulsed splenocytes, we found no difference in Lamp-1 production (data not shown) but a

–/– + higher proportion of LPA5 OT-I CD8 T cells produced granzyme B in response

+/+ + to stimulation, as compared to LPA5 OT-I CD8 T cells (Figure 5.4). As granzyme B production by CD8+ T cells causes apoptosis in target cells,

–/– + increased production of granzyme B by LPA5 CD8 T cells in the tumor microenviroment may be a factor in the enhanced control of tumor progression

–/– + by LPA5 OT-I CD8 T cells. This higher expression of granzyme B may also

–/– + indicate that LPA5 CD8 T cells differentiate to an effector phenotype, in coordination with the results of the memory experiment (Figure 5.3). Future studies to pursue this possibility in the context of Lm-OVA peptide infection, as earlier described, would allow comparisons of how LPA5 signaling influences differentiation of CD8+ T cells to an effector versus memory phenotype in response to different affinity peptides.

109 +/+ / LPA5 LPA5

105 105

104 35.5 104 52.2

103 103

102 102

0 0

0 102 103 104 105 0 102 103 104 105 Granzyme B CD8

–/– Figure 5.4 A higher proportion of LPA5 OT-I T cells produce granzyme B + +/+ –/– CD8 T cells were purified from LPA5 (left plot) or LPA5 (right plot) OT-I splenocytes and lymph nodes, and combined at a 1:2 ratio with C57BL/6 splenocytes before pulse with G4 peptide. Cells were washed and plated at 2.5 x 106 cells/mL. On day 3, cells were transferred to a round-bottom plate and brefeldin A was added for 5 hours. Cells were fixed, permeablized, and stained for flow cytometry (n=1).

Implications for tumor studies and drug development

LPA5-deficient T cells more effectively abate tumor progression

In chapter IV, we also wanted to investigate the role of LPA signaling in the CD8+ T cell response to tumor, knowing that various tumors have been reported to feature elevated LPA levels (63, 79, 118, 195, 196). Given the enhanced proliferation but reduced expression of CD25 and CD44 activation

–/– + –/– markers of LPA5 OT-I CD8 T cells, we were not sure if transfer of LPA5 OT-I

+ +/+ + CD8 T cells would better control tumor growth over LPA5 OT-I CD8 T cells.

However we were excited to find that transfer of LPA5-deficient OT-I T cells caused abatement of established B16 melanoma growth and significantly smaller

110 tumors. This indicated that even with the increased concentration of LPA generated from the tumor, LPA5-deficient tumor-specific T cells were not inhibited from controlling tumor growth within an 8-day period. When we extended the tumor experiment beyond 8 days as a survival experiment, we found that a single

–/– transfer of LPA5 OT-I T cells was not sufficient to enhance long-term survival from tumor challenge. This may suggest that multiple transfers of LPA5-deficient

OT-I CD8+ T cells may be required to provide long-term protection from tumor.

+/+ Although intracellular IFNγ and TNFα production was similar between LPA5

–/– + and LPA5 CD8 T cells, LPA5-deficient tumor-specific T cells trended to have increased proliferation and accumulation at early timepoints in the tumor, again possibly indicating a proliferative advantage and differentiation to an effector phenotype. The previous findings that LPA influences chemokinesis in T cells

(149, 197) fits well with our tumor study findings, as we find increased numbers of transferred LPA5-deficient T cells in tumor as compared to LPA5-sufficient T cells. Although we found that the presence of LPA alone did not influence naïve

T cell migration, we did not test the effect of LPA on activated T cell chemotaxis and additionally, it has been shown that LPA has a synergistic effect on chemotaxis when combined with the chemoattractant, CXCL12 (149). It would be useful to identify the LPA receptor responsible for the LPA trafficking phenotype and determine if LPA signaling in conjunction with activation affects chemokine receptor expression. Given the differential effect of LPA on DC dependent on maturation status (133, 134), it would be interesting to compare LPA receptor expression and function of naïve versus activated CD8+ T cells, however this

111 would best be analyzed at the protein level by a monoclonal antibody in comparison with other surface markers to indicate activation and differentiation status, yet flow cytometric antibodies to some LPA receptors (including LPA5) are not yet available (104).

+ Given our findings, we believe that LPA5-sufficient CD8 T cells are inhibited during activation by cognate peptide, resulting in ablation of proliferation in a subset of cells and a resultant decrease in accumulation. It has yet to be shown if this inhibition of activation occurs primarily in the tumor microenvironment or in the lymph nodes, and whether this occurs with steady state levels of LPA or elevated levels of LPA systemically or within the tumor microenvironment due to LPA generation by the tumor. Additionally, tumor- generated LPA may result in altered homing of tumor-specific CD8+ T cells as has been shown by other research groups (151, 197, 198), also contributing to a reduced concentration of CD8+ T cells within the tumor. Our data suggests that this inhibition of activation and possible altered homing results in fewer CD8+ T cells in the tumor and reduced anti-tumor adaptive immunity (Figure 5.5A). In

+ contrast, when LPA5 signaling is prevented on CD8 T cells, either by genetic ablation as we have shown or as we would predict by blockade of LPA5 via receptor-specific antagonists, CD8+ T cells are no longer inhibited to proliferate and accumulate (Figure 5.5B). In addition, we have preliminary evidence that a

+ higher proportion of LPA5-deficient CD8 T cells produce granzyme B (Figure

5.4). This results in an increased intratumoral population of tumor-specific CD8+

T cells, with enhanced anti-tumor cytokine production and thus abated tumor

112 progression. We believe that the inhibition of CD8+ T cell activation by LPA represents a new mechanism of tumor protection from the adaptive immune response.

Figure 5.5 A model: interference with LPA5 signaling improves CD8+ T cell anti-tumor immunity + (A) CD8 T cells that express LPA5 are inhibited by LPA upon activation, resulting in reduced proliferation and accumulation. This reduction in CD8+ T cell expansion results in fewer CD8+ T cells in the tumor to exert ant- tumor effects. + (B) CD8 T cells that do not express LPA5 or are treated with LPA5- blocking treatment accumulate are not inhibited from activation and subsequent proliferation. This allows increased expansion and CD8+ T cell concentration within the tumor, resulting in smaller tumors.

113 + Considerations for enhancing the CD8 T cell antitumor response

Various therapies have been explored to enhance the adaptive immune response to tumor, such as enriching and activating tumor-specific T cells ex vivo as a T cell therapy, immunizing with a tumor-derived peptide to stimulate tumor- specific T cells, or enhancing the endogenous T cell response by administration of drugs to induce T cell stimulation by activating T cells or antagonizing inhibitory receptors (32, 33, 182). Of note, two drugs that are currently in clinical trials or have already been approved for use, function by antagonizing inhibitory receptors on T cells. These drugs include Yervoy (ipilimumab), which was approved by the U.S. Food and Drug Administration in 2011 to treat advanced metastatic melanoma and enhances the T cell response by blocking the inhibitory receptor, CTLA-4 (199) and BMS-936558, an anti-PD-1 antibody that is in Phase I trials to treat various cancers, including melanoma, kidney, and non- small cell lung cancers (200). The success of these two drugs represents an exciting new therapeutic tool to enhance the T cell response to tumor and provides compelling evidence that activating T cells in the absence of other treatments (chemotherapy, etc.) is sufficient to provide therapeutic antitumoral benefits (32).

G-protein coupled receptors (GPCRs) in particular represent important drug targets (201) as over 1% of human genes are postulated to encode GPCRs

(61) and approximately one third of currently marketed drugs target GPCRs (202,

203). A prominent GPCR-targeted drug, FTY720 (GILENYA, fingolimod), targets another lysophospholipid, S1P and was approved by the U.S. Food and Drug

114 Association in 2010 to treat relapsing-remitting multiple sclerosis (96). In contrast to receptor antagonists, FTY720 acts as an agonist to the GPCR, S1P1,3-5.

However continued agonism results in functional antagonism, as the S1P receptors are internalized and irreversibly degraded (61). This downregulation of

S1P1, in particular, results in sequestration of lymphocytes in lymph nodes, thereby preventing their contribution to autoimmunity. FTY720 is also being tested in type I diabetes, systemic lupus erythematosus, rheumatoid arthritis, and other autoimmune diseases, and FTY720-related agonists are also being developed to more selectively target S1P1. The concept of receptor agonism to induce functional antagonism could be explored to reduce LPA5 signaling, as

LPA5 is also internalized upon ligand binding (108), although it is not yet clear if it is subsequently irreversibly degraded.

LPA receptor-targeted drugs have been developed, however these compounds tend to lack in vivo validation and receptor selectivity (61). VPC-

12449 is a LPA1,3 antagonist and has been shown to protect against renal ischemia-reperfusion injury via LPA3 in mice (204). Ki16425 is a also a LPA1,3 antagonist that has been shown to inhibit breast cancer cell proliferation and metastasis in mice (205). BrP-LPA is an LPA analog that antagonizes LPA1-4, inhibits ATX function, and has been shown to inhibit blood vessel density and tumor size in a mouse breast cancer model (206) and non-small cell lung cancer

(NSCLC) (207). Although BrP-LPA was designed to target cancer cells, this compound antagonizes LPA2, which we have found to be expressed on T and B cells (Figure 3.4)(152). Thus it would be important to understand the impact of

115 LPA2 and other LPA receptor antagonism on immune cells. To the author’s knowledge, an LPA5-specific antagonist has not been developed although it would be interesting to test the effects of such a compound in a tumor model.

Given that we know that LPA has been shown to be beneficial to the tumor, an

LPA-depleting treatment would theoretically be detrimental to the tumor while simultaneously enhancing the CD8+ T cell response. Anti-cancer therapies that function to reduce LPA concentration by inhibiting ATX function are being developed (67, 73, 75, 208-210). The side effects of such a treatment have yet to be reported, however detrimental complications could prevent long-term use of

ATX inhibitors, as LPA is important for vascular development and osteogenesis

(71, 211).

Success with metastatic melanoma has resulted in 47 new clinical trials to test the efficacy of Yervoy against other cancers, such as prostate cancer, non- small cell lung cancer, and even graft versus host disease (212). Although

Yervoy has been widely celebrated as a breakthrough in cancer immunotherapy, this treatment has caveats. It is not surprising that global blockade of a T cell inhibitory receptor has resulted in reports of increased autoimmunity, such as autoimmune hypophysitis (212) and colitis (32). Given the lymphoproliferative disease observed in CTLA-4–/– mice as we described earlier in this chapter, it follows that blockade of this receptor causes normally quiescent T cells to respond in the absence of infection or other stimulation. We have not seen

–/– lymphoproliferative disease in LPA5 mice and given that we have shown inhibition is only in conjunction with TCR stimulation, we propose that blocking

116 LPA5 signaling to enhance the T cell response could represent a preferable approach to cancer immunotherapy. We have found that ablation of LPA5 signaling only enhances the response of T cells upon activation, leaving unactivated and possibly autoreactive cells unresponsive to LPA5 blockade.

Although this may exclude the possibility of stimulating unactivated tumor- specific T cells into action, this may avoid autoimmune complications experienced by roughly a quarter of Yervoy patients (32).

In addition, it is important to consider combination therapies when developing cancer immunotherapies. It is known that cancer patients have tumor- specific T cells that can be elicited to respond to the tumor by determination of immunogenic tumor-derived peptides and immunizing patients with those peptides (32). Additionally, approaches to enhance stimulation of T cells indirectly by increased APC function have been attempted (32). DC-based vaccines have been tested in the clinic to induce T cell responses to tumor by transfer of antigen-loaded DC, though the results have not been impressive (213).

Further efficiency of targeting to CD8+ T cells has been achieved by targeting

CD8+ DC in vivo, which are specialized in cross presentation of antigens in MHC class I (214). By this method, others have successfully bolstered CD8+ T cell activation by conjugating the immunizing peptide to antibodies to target CD8+ DC, in combination with a DC agonist to induce maturation (215, 216). This approach has been examined to enhance the CD8+ T cell response to HIV (216-218), as well as some cancers (219). This approach may be further strengthened in

117 combination with a blocking LPA5 treatment, to enhance presentation of tumor antigen and also reduce inhibition of T cell activation.

118 REFERENCES

1. Zhang N, Bevan MJ. 2011. CD8(+) T cells: foot soldiers of the immune system. Immunity 35: 161-8

2. Masopust D, Vezys V, Wherry EJ, Ahmed R. 2007. A brief history of CD8 T cells. European journal of immunology 37 Suppl 1: S103-10

3. Murphy JB. 1914. Studies in Tissue Specificity : Ii. The Ultimate Fate of Mammalian Tissue Implanted in the Chick Embryo. The Journal of experimental medicine 19: 181-6

4. Murphy JB. 1914. Factors of Resistance to Heteroplastic Tissue-Grafting : Studies in Tissue Specificity. Iii. The Journal of experimental medicine 19: 513-22

5. Zinkernagel RM, Doherty PC. 1974. Restriction of in vitro T cell-mediated cytotoxicity in lymphocytic choriomeningitis within a syngeneic or semiallogeneic system. Nature 248: 701-2

6. Kappler JW, Skidmore B, White J, Marrack P. 1981. Antigen-inducible, H- 2-restricted, interleukin-2-producing T cell hybridomas. Lack of independent antigen and H-2 recognition. The Journal of experimental medicine 153: 1198-214

7. Allison JP, McIntyre BW, Bloch D. 1982. Tumor-specific antigen of murine T-lymphoma defined with monoclonal antibody. Journal of immunology 129: 2293-300

8. Meuer SC, Fitzgerald KA, Hussey RE, Hodgdon JC, Schlossman SF, Reinherz EL. 1983. Clonotypic structures involved in antigen-specific human T cell function. Relationship to the T3 molecular complex. The Journal of experimental medicine 157: 705-19

9. Haskins K, Kubo R, White J, Pigeon M, Kappler J, Marrack P. 1983. The major histocompatibility complex-restricted antigen receptor on T cells. I. Isolation with a monoclonal antibody. The Journal of experimental medicine 157: 1149-69

10. Dustin ML. 2009. The cellular context of T cell signaling. Immunity 30: 482-92

11. Smith-Garvin JE, Koretzky GA, Jordan MS. 2009. T cell activation. Annual review of immunology 27: 591-619

12. Reis e Sousa C. 2006. Dendritic cells in a mature age. Nature reviews. Immunology 6: 476-83 119 13. Sharpe AH, Freeman GJ. 2002. The B7-CD28 superfamily. Nature reviews. Immunology 2: 116-26

14. Riley JL, June CH. 2005. The CD28 family: a T-cell rheostat for therapeutic control of T-cell activation. Blood 105: 13-21

15. Hermiston ML, Xu Z, Weiss A. 2003. CD45: a critical regulator of signaling thresholds in immune cells. Annual review of immunology 21: 107-37

16. Shuford WW, Klussman K, Tritchler DD, Loo DT, Chalupny J, Siadak AW, Brown TJ, Emswiler J, Raecho H, Larsen CP, Pearson TC, Ledbetter JA, Aruffo A, Mittler RS. 1997. 4-1BB costimulatory signals preferentially induce CD8+ T cell proliferation and lead to the amplification in vivo of cytotoxic T cell responses. The Journal of experimental medicine 186: 47- 55

17. Horejsi V, Zhang W, Schraven B. 2004. Transmembrane adaptor proteins: organizers of immunoreceptor signalling. Nature reviews. Immunology 4: 603-16

18. Keir ME, Butte MJ, Freeman GJ, Sharpe AH. 2008. PD-1 and its ligands in tolerance and immunity. Annual review of immunology 26: 677-704

19. Baitsch L, Legat A, Barba L, Fuertes Marraco SA, Rivals JP, Baumgaertner P, Christiansen-Jucht C, Bouzourene H, Rimoldi D, Pircher H, Rufer N, Matter M, Michielin O, Speiser DE. 2012. Extended co- expression of inhibitory receptors by human CD8 T-cells depending on differentiation, antigen-specificity and anatomical localization. PLoS One 7: e30852

20. Bevan MJ. 2004. Helping the CD8(+) T-cell response. Nat Rev Immunol 4: 595-602

21. Sewell AK. 2012. Why must T cells be cross-reactive? Nature reviews. Immunology 12: 669-77

22. Mason D. 1998. A very high level of crossreactivity is an essential feature of the T-cell receptor. Immunology today 19: 395-404

23. Zehn D, Lee SY, Bevan MJ. 2009. Complete but curtailed T-cell response to very low-affinity antigen. Nature 458: 211-4

24. Gottschalk RA, Hathorn MM, Beuneu H, Corse E, Dustin ML, Altan- Bonnet G, Allison JP. 2012. Distinct influences of peptide-MHC quality and quantity on in vivo T-cell responses. Proceedings of the National Academy of Sciences of the United States of America 109: 881-6

120 25. Zehn D, Turner MJ, Lefrancois L, Bevan MJ. 2010. Lack of original antigenic sin in recall CD8(+) T cell responses. Journal of immunology 184: 6320-6

26. Carreno LJ, Riquelme EM, Gonzalez PA, Espagnolle N, Riedel CA, Valitutti S, Kalergis AM. 2010. T-cell antagonism by short half-life pMHC ligands can be mediated by an efficient trapping of T-cell polarization toward the APC. Proc Natl Acad Sci U S A 107: 210-5

27. Riquelme E, Carreno LJ, Gonzalez PA, Kalergis AM. 2009. The duration of TCR/pMHC interactions regulates CTL effector function and tumor- killing capacity. European journal of immunology 39: 2259-69

28. Corse E, Gottschalk RA, Allison JP. 2011. Strength of TCR-peptide/MHC interactions and in vivo T cell responses. Journal of immunology 186: 5039-45

29. Walter U, Santamaria P. 2005. CD8+ T cells in autoimmunity. Current opinion in immunology 17: 624-31

30. Curran MA, Kim M, Montalvo W, Al-Shamkhani A, Allison JP. 2011. Combination CTLA-4 blockade and 4-1BB activation enhances tumor rejection by increasing T-cell infiltration, proliferation, and cytokine production. PLoS One 6: e19499

31. Leach DR, Krummel MF, Allison JP. 1996. Enhancement of antitumor immunity by CTLA-4 blockade. Science 271: 1734-6

32. Mellman I, Coukos G, Dranoff G. 2011. Cancer immunotherapy comes of age. Nature 480: 480-9

33. Vesely MD, Kershaw MH, Schreiber RD, Smyth MJ. 2011. Natural innate and adaptive immunity to cancer. Annual review of immunology 29: 235- 71

34. Swann JB, Smyth MJ. 2007. Immune surveillance of tumors. The Journal of clinical investigation 117: 1137-46

35. Schreiber RD, Old LJ, Smyth MJ. 2011. Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion. Science 331: 1565-70

36. Pages F, Galon J, Dieu-Nosjean MC, Tartour E, Sautes-Fridman C, Fridman WH. 2010. Immune infiltration in human tumors: a prognostic factor that should not be ignored. Oncogene 29: 1093-102

121 37. Quezada SA, Peggs KS, Simpson TR, Allison JP. 2011. Shifting the equilibrium in cancer immunoediting: from tumor tolerance to eradication. Immunological reviews 241: 104-18

38. Thompson ED, Enriquez HL, Fu YX, Engelhard VH. 2010. Tumor masses support naive T cell infiltration, activation, and differentiation into effectors. The Journal of experimental medicine 207: 1791-804

39. Peggs KS, Quezada SA, Allison JP. 2008. Cell intrinsic mechanisms of T- cell inhibition and application to cancer therapy. Immunological reviews 224: 141-65

40. Pardoll DM. 2012. The blockade of immune checkpoints in cancer immunotherapy. Nature reviews. Cancer 12: 252-64

41. Lund AW, Duraes FV, Hirosue S, Raghavan VR, Nembrini C, Thomas SN, Issa A, Hugues S, Swartz MA. 2012. VEGF-C promotes immune tolerance in B16 melanomas and cross-presentation of tumor antigen by lymph node lymphatics. Cell reports 1: 191-9

42. Meng Y, Graves L, Do TV, So J, Fishman DA. 2004. Upregulation of FasL by LPA on ovarian cancer cell surface leads to apoptosis of activated lymphocytes. Gynecologic oncology 95: 488-95

43. Ichinose M, Masuoka J, Shiraishi T, Mineta T, Tabuchi K. 2001. Fas ligand expression and depletion of T-cell infiltration in astrocytic tumors. Brain tumor pathology 18: 37-42

44. Slingluff CL, Jr., Hunt DF, Engelhard VH. 1994. Direct analysis of tumor- associated peptide antigens. Current opinion in immunology 6: 733-40

45. Gervois N, Guilloux Y, Diez E, Jotereau F. 1996. Suboptimal activation of melanoma infiltrating lymphocytes (TIL) due to low avidity of TCR/MHC- tumor peptide interactions. The Journal of experimental medicine 183: 2403-7

46. Colella TA, Bullock TN, Russell LB, Mullins DW, Overwijk WW, Luckey CJ, Pierce RA, Restifo NP, Engelhard VH. 2000. Self-tolerance to the murine homologue of a tyrosinase-derived melanoma antigen: implications for tumor immunotherapy. The Journal of experimental medicine 191: 1221- 32

47. Slansky JE, Rattis FM, Boyd LF, Fahmy T, Jaffee EM, Schneck JP, Margulies DH, Pardoll DM. 2000. Enhanced antigen-specific antitumor immunity with altered peptide ligands that stabilize the MHC-peptide-TCR complex. Immunity 13: 529-38

122 48. Whiteside TL. 2006. Immune suppression in cancer: effects on immune cells, mechanisms and future therapeutic intervention. Seminars in cancer biology 16: 3-15

49. McMahan RH, Slansky JE. 2007. Mobilizing the low-avidity T cell repertoire to kill tumors. Seminars in cancer biology 17: 317-29

50. Lewis JJ. 2004. Therapeutic cancer vaccines: using unique antigens. Proceedings of the National Academy of Sciences of the United States of America 101 Suppl 2: 14653-6

51. Teshima T, Mach N, Hill GR, Pan L, Gillessen S, Dranoff G, Ferrara JL. 2001. Tumor cell vaccine elicits potent antitumor immunity after allogeneic T-cell-depleted bone marrow transplantation. Cancer research 61: 162-71

52. Dhodapkar MV, Dhodapkar KM, Palucka AK. 2008. Interactions of tumor cells with dendritic cells: balancing immunity and tolerance. Cell death and differentiation 15: 39-50

53. Palucka K, Banchereau J, Mellman I. 2010. Designing vaccines based on biology of human dendritic cell subsets. Immunity 33: 464-78

54. Trinh VA, Hagen B. 2012. Ipilimumab for advanced melanoma: A pharmacologic perspective. Journal of oncology pharmacy practice : official publication of the International Society of Oncology Pharmacy Practitioners PMID: 23047236

55. Chen DS, Irving BA, Hodi FS. 2012. Molecular Pathways: Next Generation Immunotherapy: Inhibiting Programmed Death-Ligand 1 and Programmed Death-1. Clinical cancer research : an official journal of the American Association for Cancer Research 18: 6580-7

56. Shablak A, Sikand K, Shanks JH, Thistlethwaite F, Spencer-Shaw A, Hawkins RE. 2011. High-dose interleukin-2 can produce a high rate of response and durable remissions in appropriately selected patients with metastatic renal cancer. Journal of immunotherapy 34: 107-12

57. Choi JW, Herr DR, Noguchi K, Yung YC, Lee CW, Mutoh T, Lin ME, Teo ST, Park KE, Mosley AN, Chun J. 2010. LPA receptors: subtypes and biological actions. Annual review of pharmacology and toxicology 50: 157- 86

58. Spiegel S, Milstien S. 2003. Sphingosine-1-phosphate: an enigmatic signalling lipid. Nature reviews. Molecular cell biology 4: 397-407

59. Yanagida K, Kurikawa Y, Shimizu T, Ishii S. 2012. Current progress in non-Edg family LPA receptor research. Biochimica et biophysica acta 1831: 33-41 123 60. Yanagida K, Ishii S. 2011. Non-Edg family LPA receptors: the cutting edge of LPA research. Journal of biochemistry 150: 223-32

61. Mutoh T, Rivera R, Chun J. 2012. Insights into the pharmacological relevance of lysophospholipid receptors. British journal of pharmacology 165: 829-44

62. Smyth SS, Cheng HY, Miriyala S, Panchatcharam M, Morris AJ. 2008. Roles of lysophosphatidic acid in cardiovascular physiology and disease. Biochimica et biophysica acta 1781: 563-70

63. Tigyi G. 2010. Aiming drug discovery at lysophosphatidic acid targets. British journal of pharmacology 161: 241-70

64. Shan L, Jaffe K, Li S, Davis L. 2008. Quantitative determination of lysophosphatidic acid by LC/ESI/MS/MS employing a reversed phase HPLC column. Journal of chromatography. B, Analytical technologies in the biomedical and life sciences 864: 22-8

65. Sano T, Baker D, Virag T, Wada A, Yatomi Y, Kobayashi T, Igarashi Y, Tigyi G. 2002. Multiple mechanisms linked to platelet activation result in lysophosphatidic acid and sphingosine 1-phosphate generation in blood. The Journal of biological chemistry 277: 21197-206

66. Tomsig JL, Snyder AH, Berdyshev EV, Skobeleva A, Mataya C, Natarajan V, Brindley DN, Lynch KR. 2009. Lipid phosphate phosphohydrolase type 1 (LPP1) degrades extracellular lysophosphatidic acid in vivo. The Biochemical journal 419: 611-8

67. Baker DL, Fujiwara Y, Pigg KR, Tsukahara R, Kobayashi S, Murofushi H, Uchiyama A, Murakami-Murofushi K, Koh E, Bandle RW, Byun HS, Bittman R, Fan D, Murph M, Mills GB, Tigyi G. 2006. Carba analogs of cyclic phosphatidic acid are selective inhibitors of autotaxin and cancer cell invasion and metastasis. The Journal of biological chemistry 281: 22786-93

68. Deng W, Shuyu E, Tsukahara R, Valentine WJ, Durgam G, Gududuru V, Balazs L, Manickam V, Arsura M, VanMiddlesworth L, Johnson LR, Parrill AL, Miller DD, Tigyi G. 2007. The lysophosphatidic acid type 2 receptor is required for protection against radiation-induced intestinal injury. Gastroenterology 132: 1834-51

69. Kosanam H, Ma F, He H, Ramagiri S, Gududuru V, Tigyi GJ, Van Rompay K, Miller DD, Yates CR. 2010. Development of an LC-MS/MS assay to determine plasma pharmacokinetics of the radioprotectant octadecenyl thiophosphate (OTP) in monkeys. Journal of chromatography. B, Analytical technologies in the biomedical and life sciences 878: 2379-83

124 70. Stracke ML, Krutzsch HC, Unsworth EJ, Arestad A, Cioce V, Schiffmann E, Liotta LA. 1992. Identification, purification, and partial sequence analysis of autotaxin, a novel motility-stimulating protein. The Journal of biological chemistry 267: 2524-9

71. Tanaka M, Okudaira S, Kishi Y, Ohkawa R, Iseki S, Ota M, Noji S, Yatomi Y, Aoki J, Arai H. 2006. Autotaxin stabilizes blood vessels and is required for embryonic vasculature by producing lysophosphatidic acid. The Journal of biological chemistry 281: 25822-30

72. van Meeteren LA, Ruurs P, Stortelers C, Bouwman P, van Rooijen MA, Pradere JP, Pettit TR, Wakelam MJ, Saulnier-Blache JS, Mummery CL, Moolenaar WH, Jonkers J. 2006. Autotaxin, a secreted lysophospholipase D, is essential for blood vessel formation during development. Molecular and cellular biology 26: 5015-22

73. Hoeglund AB, Howard AL, Wanjala IW, Pham TC, Parrill AL, Baker DL. 2010. Characterization of non-lipid autotaxin inhibitors. Bioorganic & medicinal chemistry 18: 769-76

74. North EJ, Howard AL, Wanjala IW, Pham TC, Baker DL, Parrill AL. 2010. Pharmacophore development and application toward the identification of novel, small-molecule autotaxin inhibitors. Journal of medicinal chemistry 53: 3095-105

75. Saunders LP, Ouellette A, Bandle R, Chang WC, Zhou H, Misra RN, De La Cruz EM, Braddock DT. 2008. Identification of small-molecule inhibitors of autotaxin that inhibit melanoma cell migration and invasion. Molecular cancer therapeutics 7: 3352-62

76. Ishii S, Noguchi K, Yanagida K. 2009. Non-Edg family lysophosphatidic acid (LPA) receptors. Prostaglandins & other lipid mediators 89: 57-65

77. Bolen AL, Naren AP, Yarlagadda S, Beranova-Giorgianni S, Chen L, Norman D, Baker DL, Rowland MM, Best MD, Sano T, Tsukahara T, Liliom K, Igarashi Y, Tigyi G. 2011. The phospholipase A1 activity of lysophospholipase A-I links platelet activation to LPA production during blood coagulation. Journal of lipid research 52: 958-70

78. Aoki J, Inoue A, Okudaira S. 2008. Two pathways for lysophosphatidic acid production. Biochimica et biophysica acta 1781: 513-8

79. Mills GB, Moolenaar WH. 2003. The emerging role of lysophosphatidic acid in cancer. Nat Rev Cancer 3: 582-91

125 80. Taylor LA, Arends J, Hodina AK, Unger C, Massing U. 2007. Plasma lysophosphatidylcholine concentration is decreased in cancer patients with weight loss and activated inflammatory status. Lipids in health and disease 6: 17

81. Pilquil C, Singh I, Zhang QX, Ling ZC, Buri K, Stromberg LM, Dewald J, Brindley DN. 2001. Lipid phosphate phosphatase-1 dephosphorylates exogenous lysophosphatidate and thereby attenuates its effects on cell signalling. Prostaglandins & other lipid mediators 64: 83-92

82. Umezu-Goto M, Kishi Y, Taira A, Hama K, Dohmae N, Takio K, Yamori T, Mills GB, Inoue K, Aoki J, Arai H. 2002. Autotaxin has lysophospholipase D activity leading to tumor cell growth and motility by lysophosphatidic acid production. J Cell Biol 158: 227-33

83. Brindley DN, Pilquil C. 2009. Lipid phosphate phosphatases and signaling. Journal of lipid research 50 Suppl: S225-30

84. Xu Y, Shen Z, Wiper DW, Wu M, Morton RE, Elson P, Kennedy AW, Belinson J, Markman M, Casey G. 1998. Lysophosphatidic acid as a potential biomarker for ovarian and other gynecologic cancers. JAMA 280: 719-23

85. Sasagawa T, Okita M, Murakami J, Kato T, Watanabe A. 1999. Abnormal serum lysophospholipids in multiple myeloma patients. Lipids 34: 17-21

86. Baker DL, Morrison P, Miller B, Riely CA, Tolley B, Westermann AM, Bonfrer JM, Bais E, Moolenaar WH, Tigyi G. 2002. Plasma lysophosphatidic acid concentration and ovarian cancer. JAMA 287: 3081- 2

87. Goetzl EJ, Rosen H. 2004. Regulation of immunity by lysosphingolipids and their G protein-coupled receptors. J Clin Invest 114: 1531-7

88. Tokumura A, Carbone LD, Yoshioka Y, Morishige J, Kikuchi M, Postlethwaite A, Watsky MA. 2009. Elevated serum levels of arachidonoyl-lysophosphatidic acid and sphingosine 1-phosphate in systemic sclerosis. International journal of medical sciences 6: 168-76

89. Watanabe N, Ikeda H, Nakamura K, Ohkawa R, Kume Y, Tomiya T, Tejima K, Nishikawa T, Arai M, Yanase M, Aoki J, Arai H, Omata M, Fujiwara K, Yatomi Y. 2007. Plasma lysophosphatidic acid level and serum autotaxin activity are increased in liver injury in rats in relation to its severity. Life sciences 81: 1009-15

126 90. Yamada T, Sato K, Komachi M, Malchinkhuu E, Tobo M, Kimura T, Kuwabara A, Yanagita Y, Ikeya T, Tanahashi Y, Ogawa T, Ohwada S, Morishita Y, Ohta H, Im DS, Tamoto K, Tomura H, Okajima F. 2004. Lysophosphatidic acid (LPA) in malignant ascites stimulates motility of human pancreatic cancer cells through LPA1. The Journal of biological chemistry 279: 6595-605

91. Hausmann J, Kamtekar S, Christodoulou E, Day JE, Wu T, Fulkerson Z, Albers HM, van Meeteren LA, Houben AJ, van Zeijl L, Jansen S, Andries M, Hall T, Pegg LE, Benson TE, Kasiem M, Harlos K, Kooi CW, Smyth SS, Ovaa H, Bollen M, Morris AJ, Moolenaar WH, Perrakis A. 2011. Structural basis of substrate discrimination and integrin binding by autotaxin. Nature structural & molecular biology 18: 198-204

92. Nishimasu H, Okudaira S, Hama K, Mihara E, Dohmae N, Inoue A, Ishitani R, Takagi J, Aoki J, Nureki O. 2011. Crystal structure of autotaxin and insight into GPCR activation by lipid mediators. Nature structural & molecular biology 18: 205-12

93. Fulkerson Z, Wu T, Sunkara M, Kooi CV, Morris AJ, Smyth SS. 2011. Binding of autotaxin to integrins localizes lysophosphatidic acid production to platelets and mammalian cells. The Journal of biological chemistry 286: 34654-63

94. Moolenaar WH, Perrakis A. 2011. Insights into autotaxin: how to produce and present a lipid mediator. Nature reviews. Molecular cell biology 12: 674-9

95. Moolenaar WH, van Meeteren LA, Giepmans BN. 2004. The ins and outs of lysophosphatidic acid signaling. BioEssays : news and reviews in molecular, cellular and developmental biology 26: 870-81

96. Chun J, Hla T, Lynch KR, Spiegel S, Moolenaar WH. 2010. International Union of Basic and Clinical Pharmacology. LXXVIII. Lysophospholipid receptor nomenclature. Pharmacological reviews 62: 579-87

97. Hecht JH, Weiner JA, Post SR, Chun J. 1996. Ventricular zone gene-1 (vzg-1) encodes a lysophosphatidic acid receptor expressed in neurogenic regions of the developing cerebral cortex. The Journal of cell biology 135: 1071-83

98. An S, Bleu T, Hallmark OG, Goetzl EJ. 1998. Characterization of a novel subtype of human G protein-coupled receptor for lysophosphatidic acid. The Journal of biological chemistry 273: 7906-10

127 99. Bandoh K, Aoki J, Hosono H, Kobayashi S, Kobayashi T, Murakami- Murofushi K, Tsujimoto M, Arai H, Inoue K. 1999. Molecular cloning and characterization of a novel human G-protein-coupled receptor, EDG7, for lysophosphatidic acid. The Journal of biological chemistry 274: 27776-85

100. McIntyre TM, Pontsler AV, Silva AR, St Hilaire A, Xu Y, Hinshaw JC, Zimmerman GA, Hama K, Aoki J, Arai H, Prestwich GD. 2003. Identification of an intracellular receptor for lysophosphatidic acid (LPA): LPA is a transcellular PPARgamma agonist. Proceedings of the National Academy of Sciences of the United States of America 100: 131-6

101. Stapleton CM, Mashek DG, Wang S, Nagle CA, Cline GW, Thuillier P, Leesnitzer LM, Li LO, Stimmel JB, Shulman GI, Coleman RA. 2011. Lysophosphatidic acid activates peroxisome proliferator activated receptor-gamma in CHO cells that over-express glycerol 3-phosphate acyltransferase-1. PLoS One 6: e18932

102. Cheng Y, Makarova N, Tsukahara R, Guo H, Shuyu E, Farrar P, Balazs L, Zhang C, Tigyi G. 2009. Lysophosphatidic acid-induced arterial wall remodeling: requirement of PPARgamma but not LPA1 or LPA2 GPCR. Cellular signalling 21: 1874-84

103. Blaho VA, Hla T. 2011. Regulation of mammalian physiology, development, and disease by the sphingosine 1-phosphate and lysophosphatidic acid receptors. Chemical reviews 111: 6299-320

104. Georas SN. 2009. Lysophosphatidic acid and autotaxin: emerging roles in innate and adaptive immunity. Immunologic research 45: 229-38

105. Lin FT, Lai YJ, Makarova N, Tigyi G, Lin WC. 2007. The lysophosphatidic acid 2 receptor mediates down-regulation of Siva-1 to promote cell survival. The Journal of biological chemistry 282: 37759-69

106. Jonkers J, Moolenaar WH. 2009. Mammary tumorigenesis through LPA receptor signaling. Cancer Cell 15: 457-9

107. Kotarsky K, Boketoft A, Bristulf J, Nilsson NE, Norberg A, Hansson S, Owman C, Sillard R, Leeb-Lundberg LM, Olde B. 2006. Lysophosphatidic acid binds to and activates GPR92, a G protein-coupled receptor highly expressed in gastrointestinal lymphocytes. The Journal of pharmacology and experimental therapeutics 318: 619-28

108. Lee CW, Rivera R, Gardell S, Dubin AE, Chun J. 2006. GPR92 as a new G12/13- and Gq-coupled lysophosphatidic acid receptor that increases cAMP, LPA5. The Journal of biological chemistry 281: 23589-97

128 109. Oh DY, Yoon JM, Moon MJ, Hwang JI, Choe H, Lee JY, Kim JI, Kim S, Rhim H, O'Dell DK, Walker JM, Na HS, Lee MG, Kwon HB, Kim K, Seong JY. 2008. Identification of farnesyl pyrophosphate and N- arachidonylglycine as endogenous ligands for GPR92. The Journal of biological chemistry 283: 21054-64

110. Williams JR, Khandoga AL, Goyal P, Fells JI, Perygin DH, Siess W, Parrill AL, Tigyi G, Fujiwara Y. 2009. Unique ligand selectivity of the GPR92/LPA5 lysophosphatidate receptor indicates role in human platelet activation. The Journal of biological chemistry 284: 17304-19

111. Yin H, Chu A, Li W, Wang B, Shelton F, Otero F, Nguyen DG, Caldwell JS, Chen YA. 2009. Lipid G protein-coupled receptor ligand identification using beta-arrestin PathHunter assay. The Journal of biological chemistry 284: 12328-38

112. Khandoga AL, Pandey D, Welsch U, Brandl R, Siess W. 2011. GPR92/LPA(5) lysophosphatidate receptor mediates megakaryocytic cell shape change induced by human atherosclerotic plaques. Cardiovascular research 90: 157-64

113. Kozian DH, Evers A, Florian P, Wonerow P, Joho S, Nazare M. 2012. Selective non-lipid modulator of LPA5 activity in human platelets. Bioorganic & medicinal chemistry letters 22: 5239-43

114. Lundequist A, Boyce JA. 2011. LPA5 is abundantly expressed by human mast cells and important for lysophosphatidic acid induced MIP-1beta release. PLoS One 6: e18192

115. Lin ME, Rivera RR, Chun J. 2012. Targeted deletion of LPA5 identifies novel roles for lysophosphatidic acid signaling in development of neuropathic pain. The Journal of biological chemistry 287: 17608-17

116. Houben AJ, Moolenaar WH. 2011. Autotaxin and LPA receptor signaling in cancer. Cancer metastasis reviews 30: 557-65

117. Samadi N, Bekele R, Capatos D, Venkatraman G, Sariahmetoglu M, Brindley DN. 2011. Regulation of lysophosphatidate signaling by autotaxin and lipid phosphate phosphatases with respect to tumor progression, angiogenesis, metastasis and chemo-resistance. Biochimie 93: 61-70

118. Gotoh M, Fujiwara Y, Yue J, Liu J, Lee S, Fells J, Uchiyama A, Murakami- Murofushi K, Kennel S, Wall J, Patil R, Gupte R, Balazs L, Miller DD, Tigyi GJ. 2012. Controlling cancer through the autotaxin-lysophosphatidic acid receptor axis. Biochemical Society transactions 40: 31-6

129 119. Shida D, Kitayama J, Yamaguchi H, Okaji Y, Tsuno NH, Watanabe T, Takuwa Y, Nagawa H. 2003. Lysophosphatidic acid (LPA) enhances the metastatic potential of human colon carcinoma DLD1 cells through LPA1. Cancer research 63: 1706-11

120. Rivera-Lopez CM, Tucker AL, Lynch KR. 2008. Lysophosphatidic acid (LPA) and angiogenesis. Angiogenesis 11: 301-10

121. Bian D, Su S, Mahanivong C, Cheng RK, Han Q, Pan ZK, Sun P, Huang S. 2004. Lysophosphatidic Acid Stimulates Ovarian Cancer Cell Migration via a Ras-MEK Kinase 1 Pathway. Cancer research 64: 4209-17

122. Ishdorj G, Graham BA, Hu X, Chen J, Johnston JB, Fang X, Gibson SB. 2008. Lysophosphatidic acid protects cancer cells from histone deacetylase (HDAC) inhibitor-induced apoptosis through activation of HDAC. The Journal of biological chemistry 283: 16818-29

123. Du J, Sun C, Hu Z, Yang Y, Zhu Y, Zheng D, Gu L, Lu X. 2010. Lysophosphatidic acid induces MDA-MB-231 breast cancer cells migration through activation of PI3K/PAK1/ERK signaling. PLoS One 5: e15940

124. Yamada T, Yano S, Ogino H, Ikuta K, Kakiuchi S, Hanibuchi M, Kanematsu T, Taniguchi T, Sekido Y, Sone S. 2008. Lysophosphatidic acid stimulates the proliferation and motility of malignant pleural mesothelioma cells through lysophosphatidic acid receptors, LPA1 and LPA2. Cancer science 99: 1603-10

125. Vidot S, Witham J, Agarwal R, Greenhough S, Bamrah HS, Tigyi GJ, Kaye SB, Richardson A. 2010. Autotaxin delays apoptosis induced by carboplatin in ovarian cancer cells. Cellular signalling 22: 926-35

126. Liu S, Umezu-Goto M, Murph M, Lu Y, Liu W, Zhang F, Yu S, Stephens LC, Cui X, Murrow G, Coombes K, Muller W, Hung MC, Perou CM, Lee AV, Fang X, Mills GB. 2009. Expression of autotaxin and lysophosphatidic acid receptors increases mammary tumorigenesis, invasion, and metastases. Cancer Cell 15: 539-50

127. Fukui R, Kato K, Okabe K, Kitayoshi M, Tanabe E, Fukushima N, Tsujiuchi T. 2012. Enhancement of Drug Resistance by Lysophosphatidic Acid Receptor-3 in Mouse Mammary Tumor FM3A Cells. Journal of toxicologic pathology 25: 225-8

128. Hao F, Tan M, Xu X, Han J, Miller DD, Tigyi G, Cui MZ. 2007. Lysophosphatidic acid induces prostate cancer PC3 cell migration via activation of LPA(1), p42 and p38alpha. Biochim Biophys Acta 1771: 883- 92

130 129. Stahle M, Veit C, Bachfischer U, Schierling K, Skripczynski B, Hall A, Gierschik P, Giehl K. 2003. Mechanisms in LPA-induced tumor cell migration: critical role of phosphorylated ERK. Journal of cell science 116: 3835-46

130. Jongsma M, Matas-Rico E, Rzadkowski A, Jalink K, Moolenaar WH. 2011. LPA is a chemorepellent for B16 melanoma cells: action through the cAMP-elevating LPA5 receptor. PLoS One 6: e29260

131. Lagadari M, Truta-Feles K, Lehmann K, Berod L, Ziemer M, Idzko M, Barz D, Kamradt T, Maghazachi AA, Norgauer J. 2009. Lysophosphatidic acid inhibits the cytotoxic activity of NK cells: involvement of Gs protein- mediated signaling. International immunology 21: 667-77

132. Li S, Zhang J. 2009. Lipopolysaccharide induces autotaxin expression in human monocytic THP-1 cells. Biochemical and biophysical research communications 378: 264-8

133. Panther E, Idzko M, Corinti S, Ferrari D, Herouy Y, Mockenhaupt M, Dichmann S, Gebicke-Haerter P, Di Virgilio F, Girolomoni G, Norgauer J. 2002. The influence of lysophosphatidic acid on the functions of human dendritic cells. Journal of immunology 169: 4129-35

134. Chan LC, Peters W, Xu Y, Chun J, Farese RV, Jr., Cases S. 2007. LPA3 receptor mediates chemotaxis of immature murine dendritic cells to unsaturated lysophosphatidic acid (LPA). Journal of leukocyte biology 82: 1193-200

135. Leslie DS, Dascher CC, Cembrola K, Townes MA, Hava DL, Hugendubler LC, Mueller E, Fox L, Roura-Mir C, Moody DB, Vincent MS, Gumperz JE, Illarionov PA, Besra GS, Reynolds CG, Brenner MB. 2008. Serum lipids regulate dendritic cell CD1 expression and function. Immunology 125: 289-301

136. Emo J, Meednu N, Chapman TJ, Rezaee F, Balys M, Randall T, Rangasamy T, Georas SN. 2012. Lpa2 is a negative regulator of both dendritic cell activation and murine models of allergic lung inflammation. Journal of immunology 188: 3784-90

137. Donovan EE, Pelanda R, Torres RM. 2010. S1P3 confers differential S1P- induced migration by autoreactive and non-autoreactive immature B cells and is required for normal B-cell development. European journal of immunology 40: 688-98

138. Pereira JP, Xu Y, Cyster JG. 2010. A role for S1P and S1P1 in immature- B cell egress from mouse bone marrow. PLoS One 5: e9277

131 139. Matloubian M, Lo CG, Cinamon G, Lesneski MJ, Xu Y, Brinkmann V, Allende ML, Proia RL, Cyster JG. 2004. Lymphocyte egress from thymus and peripheral lymphoid organs is dependent on S1P receptor 1. Nature 427: 355-60

140. Allende ML, Tuymetova G, Lee BG, Bonifacino E, Wu YP, Proia RL. 2010. S1P1 receptor directs the release of immature B cells from bone marrow into blood. The Journal of experimental medicine 207: 1113-24

141. Rosen H, Gonzalez-Cabrera P, Marsolais D, Cahalan S, Don AS, Sanna MG. 2008. Modulating tone: the overture of S1P receptor immunotherapeutics. Immunological reviews 223: 221-35

142. Green JA, Suzuki K, Cho B, Willison LD, Palmer D, Allen CD, Schmidt TH, Xu Y, Proia RL, Coughlin SR, Cyster JG. 2011. The sphingosine 1- phosphate receptor S1P(2) maintains the homeostasis of germinal center B cells and promotes niche confinement. Nature immunology 12: 672-80

143. Goetzl EJ, Kong Y, Voice JK. 2000. Cutting edge: differential constitutive expression of functional receptors for lysophosphatidic acid by human blood lymphocytes. Journal of immunology 164: 4996-9

144. Wang L, Knudsen E, Jin Y, Gessani S, Maghazachi AA. 2004. Lysophospholipids and chemokines activate distinct signal transduction pathways in T helper 1 and T helper 2 cells. Cellular signalling 16: 991- 1000

145. Takemura H, Imoto K, Sakano S, Kaneko M, Ohshika H. 1996. Lysophosphatidic acid-sensitive intracellular Ca2+ store does not regulate Ca2+ entry at plasma membrane in Jurkat human T-cells. The Biochemical journal 319 ( Pt 2): 393-7

146. Zheng Y, Voice JK, Kong Y, Goetzl EJ. 2000. Altered expression and functional profile of lysophosphatidic acid receptors in mitogen-activated human blood T lymphocytes. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 14: 2387-9

147. Rubenfeld J, Guo J, Sookrung N, Chen R, Chaicumpa W, Casolaro V, Zhao Y, Natarajan V, Georas S. 2006. Lysophosphatidic acid enhances interleukin-13 gene expression and promoter activity in T cells. American journal of physiology. Lung cellular and molecular physiology 290: L66-74

148. Sevastou I, Kaffe E, Mouratis MA, Aidinis V. 2012. Lysoglycerophospholipids in chronic inflammatory disorders: The PLA(2)/LPC and ATX/LPA axes. Biochimica et biophysica acta 1831: 42- 60

132 149. Kanda H, Newton R, Klein R, Morita Y, Gunn MD, Rosen SD. 2008. Autotaxin, an ectoenzyme that produces lysophosphatidic acid, promotes the entry of lymphocytes into secondary lymphoid organs. Nature immunology 9: 415-23

150. Pamuklar Z, Federico L, Liu S, Umezu-Goto M, Dong A, Panchatcharam M, Fulkerson Z, Berdyshev E, Natarajan V, Fang X, van Meeteren LA, Moolenaar WH, Mills GB, Morris AJ, Smyth SS. 2009. Autotaxin/lysopholipase D and lysophosphatidic acid regulate murine hemostasis and thrombosis. The Journal of biological chemistry 284: 7385-94

151. Tanikawa T, Kurohane K, Imai Y. 2010. Regulatory effect of lysophosphatidic acid on lymphocyte migration. Biological & pharmaceutical bulletin 33: 204-8

152. Hu J, Oda SK, Donovan EE, Shotts K, Strauch P, Pujanauski LM, Victorino F, Waterman PM, Al-Shami A, Fujiwara Y, Tigyi G, Oravecz T, Pelanda R, Torres RM. 2012. Lysophosphatidic Acid (LPA) Receptor 5 Inhibits B Cell Antigen Receptor Signaling and Antibody Response. In preparation

153. Hogquist KA, Jameson SC, Heath WR, Howard JL, Bevan MJ, Carbone FR. 1994. T cell receptor antagonist peptides induce positive selection. Cell 76: 17-27

154. Contos JJ, Ishii I, Fukushima N, Kingsbury MA, Ye X, Kawamura S, Brown JH, Chun J. 2002. Characterization of lpa(2) (Edg4) and lpa(1)/lpa(2) (Edg2/Edg4) lysophosphatidic acid receptor knockout mice: signaling deficits without obvious phenotypic abnormality attributable to lpa(2). Molecular and cellular biology 22: 6921-9

155. Hermiston ML, Zikherman J, Tan AL, Lam VC, Cresalia NM, Oksenberg N, Goren N, Brassat D, Oksenberg JR, Weiss A. 2009. Differential impact of the CD45 juxtamembrane wedge on central and peripheral T cell receptor responses. Proceedings of the National Academy of Sciences of the United States of America 106: 546-51

156. Khandoga AL, Fujiwara Y, Goyal P, Pandey D, Tsukahara R, Bolen A, Guo H, Wilke N, Liu J, Valentine WJ, Durgam GG, Miller DD, Jiang G, Prestwich GD, Tigyi G, Siess W. 2008. Lysophosphatidic acid-induced platelet shape change revealed through LPA(1-5) receptor-selective probes and albumin. Platelets 19: 415-27

133 157. Durgam GG, Virag T, Walker MD, Tsukahara R, Yasuda S, Liliom K, van Meeteren LA, Moolenaar WH, Wilke N, Siess W, Tigyi G, Miller DD. 2005. Synthesis, structure-activity relationships, and biological evaluation of fatty alcohol phosphates as lysophosphatidic acid receptor ligands, activators of PPARgamma, and inhibitors of autotaxin. Journal of medicinal chemistry 48: 4919-30

158. Gerlach E, Deuticke B. 1963. [a Simple Method for Microdetermination of Phosphate in Paper Chromatography]. Biochemische Zeitschrift 337: 477- 9

159. Jenkins MR, Tsun A, Stinchcombe JC, Griffiths GM. 2009. The strength of T cell receptor signal controls the polarization of cytotoxic machinery to the immunological synapse. Immunity 31: 621-31

160. Barber DL, Wherry EJ, Ahmed R. 2003. Cutting edge: rapid in vivo killing by memory CD8 T cells. Journal of immunology 171: 27-31

161. Liu S, Murph M, Panupinthu N, Mills GB. 2009. ATX-LPA receptor axis in inflammation and cancer. Cell cycle 8: 3695-701

162. Donovan EE. 2010. Lysophospholipid Regulation of Immature B Cell Development. Doctoral thesis. University of Colorado Denver, Denver

163. Pua TL, Wang FQ, Fishman DA. 2009. Roles of LPA in ovarian cancer development and progression. Future oncology 5: 1659-73

164. Heng TS, Painter MW. 2008. The Immunological Genome Project: networks of gene expression in immune cells. Nature immunology 9: 1091-4

165. Pinchuk LM, Filipov NM. 2008. Differential effects of age on circulating and splenic leukocyte populations in C57BL/6 and BALB/c male mice. Immunity & ageing : I & A 5: 1

166. Feske S, Skolnik EY, Prakriya M. 2012. Ion channels and transporters in lymphocyte function and immunity. Nature reviews. Immunology 12: 532- 47

167. Rao A. 2009. Signaling to gene expression: calcium, calcineurin and NFAT. Nature immunology 10: 3-5

168. Albers HM, Dong A, van Meeteren LA, Egan DA, Sunkara M, van Tilburg EW, Schuurman K, van Tellingen O, Morris AJ, Smyth SS, Moolenaar WH, Ovaa H. 2010. Boronic acid-based inhibitor of autotaxin reveals rapid turnover of LPA in the circulation. Proceedings of the National Academy of Sciences of the United States of America 107: 7257-62

134 169. Lu J, Xiao Yj YJ, Baudhuin LM, Hong G, Xu Y. 2002. Role of ether-linked lysophosphatidic acids in ovarian cancer cells. Journal of lipid research 43: 463-76

170. Schlub TE, Badovinac VP, Sabel JT, Harty JT, Davenport MP. 2010. Predicting CD62L expression during the CD8+ T-cell response in vivo. Immunology and cell biology 88: 157-64

171. Kiss GN, Fells JI, Gupte R, Lee SC, Liu J, Nusser N, Lim KG, Ray RM, Lin FT, Parrill AL, Sumegi B, Miller DD, Tigyi GJ. 2012. Virtual Screening for LPA2-Specific Agonists Identifies a Nonlipid Compound with Antiapoptotic Actions. Molecular pharmacology 82: 1162-73

172. Stone JD, Chervin AS, Kranz DM. 2009. T-cell receptor binding affinities and kinetics: impact on T-cell activity and specificity. Immunology 126: 165-76

173. Boyman O, Sprent J. 2012. The role of interleukin-2 during homeostasis and activation of the immune system. Nature reviews. Immunology 12: 180-90

174. Ho SN, Abraham RT, Gillis S, McKean DJ. 1987. Differential bioassay of interleukin 2 and interleukin 4. Journal of immunological methods 98: 99- 104

175. Nguyen van Binh P, Duc HT. 2006. Epitopes of the class I major histocompatibility complex (MHC-I) recognized in the syngeneic or allogeneic context predominantly linked to antigenic peptide loading to its binding groove. Clinical and experimental immunology 145: 372-9

176. Wakim LM, Woodward-Davis A, Bevan MJ. 2010. Memory T cells persisting within the brain after local infection show functional adaptations to their tissue of residence. Proceedings of the National Academy of Sciences of the United States of America 107: 17872-9

177. Baker DL, Desiderio DM, Miller DD, Tolley B, Tigyi GJ. 2001. Direct quantitative analysis of lysophosphatidic acid molecular species by stable isotope dilution electrospray ionization liquid chromatography-mass spectrometry. Analytical biochemistry 292: 287-95

178. Enders GA. 2002. Mechanism of antigen presentation after hypertonic loading of soluble antigens. Immunology 106: 511-6

179. Zhou G, Ding ZC, Fu J, Levitsky HI. 2011. Presentation of acquired peptide-MHC class II ligands by CD4+ regulatory T cells or helper cells differentially regulates antigen-specific CD4+ T cell response. Journal of immunology 186: 2148-55

135 180. Malek TR. 2008. The biology of interleukin-2. Annual review of immunology 26: 453-79

181. Subtil A, Hemar A, Dautry-Varsat A. 1994. Rapid endocytosis of interleukin 2 receptors when clathrin-coated pit endocytosis is inhibited. Journal of cell science 107 ( Pt 12): 3461-8

182. Leen AM, Rooney CM, Foster AE. 2007. Improving T cell therapy for cancer. Annual review of immunology 25: 243-65

183. Callaerts-Vegh Z, Leo S, Vermaercke B, Meert T, D'Hooge R. 2012. LPA(5) receptor plays a role in pain sensitivity, emotional exploration and reversal learning. Genes, brain, and behavior doi: 10.1111/j.1601- 183X.2012.00840.x

184. Petersen CC, Petersen MS, Agger R, Hokland ME. 2006. Accumulation in tumor tissue of adoptively transferred T cells: A comparison between intravenous and intraperitoneal injection. Journal of immunotherapy 29: 241-9

185. Garcia-Hernandez Mde L, Hamada H, Reome JB, Misra SK, Tighe MP, Dutton RW. 2010. Adoptive transfer of tumor-specific Tc17 effector T cells controls the growth of B16 melanoma in mice. Journal of immunology 184: 4215-27

186. Egen JG, Allison JP. 2002. Cytotoxic T lymphocyte antigen-4 accumulation in the immunological synapse is regulated by TCR signal strength. Immunity 16: 23-35

187. Egen JG, Kuhns MS, Allison JP. 2002. CTLA-4: new insights into its biological function and use in tumor immunotherapy. Nature immunology 3: 611-8

188. Waterhouse P, Penninger JM, Timms E, Wakeham A, Shahinian A, Lee KP, Thompson CB, Griesser H, Mak TW. 1995. Lymphoproliferative disorders with early lethality in mice deficient in Ctla-4. Science 270: 985-8

189. Chambers CA, Kuhns MS, Egen JG, Allison JP. 2001. CTLA-4-mediated inhibition in regulation of T cell responses: mechanisms and manipulation in tumor immunotherapy. Annual review of immunology 19: 565-94

190. Weber C, Krueger A, Munk A, Bode C, Van Veldhoven PP, Graler MH. 2009. Discontinued postnatal thymocyte development in sphingosine 1- phosphate-lyase-deficient mice. Journal of immunology 183: 4292-301

191. Schwab SR, Pereira JP, Matloubian M, Xu Y, Huang Y, Cyster JG. 2005. Lymphocyte sequestration through S1P lyase inhibition and disruption of S1P gradients. Science 309: 1735-9 136 192. Pedicord VA, Montalvo W, Leiner IM, Allison JP. 2011. Single dose of anti- CTLA-4 enhances CD8+ T-cell memory formation, function, and maintenance. Proceedings of the National Academy of Sciences of the United States of America 108: 266-71

193. Belz GT, Masson F. 2010. Interleukin-2 tickles T cell memory. Immunity 32: 7-9

194. Prlic M, Sacks JA, Bevan MJ. 2012. Dissociating markers of senescence and protective ability in memory T cells. PLoS One 7: e32576

195. Peyruchaud O. 2009. Novel implications for lysophospholipids, lysophosphatidic acid and sphingosine 1-phosphate, as drug targets in cancer. Anti-cancer agents in medicinal chemistry 9: 381-91

196. Goldshmit Y, Matteo R, Sztal T, Ellett F, Frisca F, Moreno K, Crombie D, Lieschke GJ, Currie PD, Sabbadini RA, Pebay A. 2012. Blockage of Lysophosphatidic Acid Signaling Improves Spinal Cord Injury Outcomes. The American journal of pathology 181: 978-92

197. Zhang Y, Chen YC, Krummel MF, Rosen SD. 2012. Autotaxin through Lysophosphatidic Acid Stimulates Polarization, Motility, and Transendothelial Migration of Naive T Cells. Journal of immunology 189: 3914-24

198. Zheng Y, Kong Y, Goetzl EJ. 2001. Lysophosphatidic acid receptor- selective effects on Jurkat T cell migration through a Matrigel model basement membrane. Journal of immunology 166: 2317-22

199. Mansh M. 2011. Ipilimumab and cancer immunotherapy: a new hope for advanced stage melanoma. The Yale journal of biology and medicine 84: 381-9

200. Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, Powderly JD, Carvajal RD, Sosman JA, Atkins MB, Leming PD, Spigel DR, Antonia SJ, Horn L, Drake CG, Pardoll DM, Chen L, Sharfman WH, Anders RA, Taube JM, McMiller TL, Xu H, Korman AJ, Jure-Kunkel M, Agrawal S, McDonald D, Kollia GD, Gupta A, Wigginton JM, Sznol M. 2012. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. The New England journal of medicine 366: 2443-54

201. Dunworth WP, Caron KM. 2009. G protein-coupled receptors as potential drug targets for lymphangiogenesis and lymphatic vascular diseases. Arteriosclerosis, thrombosis, and vascular biology 29: 650-6

202. Overington JP, Al-Lazikani B, Hopkins AL. 2006. How many drug targets are there? Nature reviews. Drug discovery 5: 993-6

137 203. Kroeze WK, Sheffler DJ, Roth BL. 2003. G-protein-coupled receptors at a glance. Journal of cell science 116: 4867-9

204. Okusa MD, Ye H, Huang L, Sigismund L, Macdonald T, Lynch KR. 2003. Selective blockade of lysophosphatidic acid LPA3 receptors reduces murine renal ischemia-reperfusion injury. American journal of physiology. Renal physiology 285: F565-74

205. Ohta H, Sato K, Murata N, Damirin A, Malchinkhuu E, Kon J, Kimura T, Tobo M, Yamazaki Y, Watanabe T, Yagi M, Sato M, Suzuki R, Murooka H, Sakai T, Nishitoba T, Im DS, Nochi H, Tamoto K, Tomura H, Okajima F. 2003. Ki16425, a subtype-selective antagonist for EDG-family lysophosphatidic acid receptors. Molecular pharmacology 64: 994-1005

206. Zhang H, Xu X, Gajewiak J, Tsukahara R, Fujiwara Y, Liu J, Fells JI, Perygin D, Parrill AL, Tigyi G, Prestwich GD. 2009. Dual activity lysophosphatidic acid receptor pan-antagonist/autotaxin inhibitor reduces breast cancer cell migration in vitro and causes tumor regression in vivo. Cancer research 69: 5441-9

207. Xu X, Prestwich GD. 2010. Inhibition of tumor growth and angiogenesis by a lysophosphatidic acid antagonist in an engineered three-dimensional lung cancer xenograft model. Cancer 116: 1739-50

208. Albers HM, Ovaa H. 2012. Chemical evolution of autotaxin inhibitors. Chemical reviews 112: 2593-603

209. Cui P, Tomsig JL, McCalmont WF, Lee S, Becker CJ, Lynch KR, Macdonald TL. 2007. Synthesis and biological evaluation of phosphonate derivatives as autotaxin (ATX) inhibitors. Bioorg Med Chem Lett 17: 1634- 40

210. East JE, Kennedy AJ, Tomsig JL, De Leon AR, Lynch KR, Macdonald TL. 2010. Synthesis and structure-activity relationships of tyrosine-based inhibitors of autotaxin (ATX). Bioorg Med Chem Lett 20: 7132-6

211. Gennero I, Laurencin-Dalicieux S, Conte-Auriol F, Briand-Mesange F, Laurencin D, Rue J, Beton N, Malet N, Mus M, Tokumura A, Bourin P, Vico L, Brunel G, Oreffo RO, Chun J, Salles JP. 2011. Absence of the lysophosphatidic acid receptor LPA1 results in abnormal bone development and decreased bone mass. Bone 49: 395-403

212. Juszczak A, Gupta A, Karavitaki N, Middleton MR, Grossman AB. 2012. Ipilimumab: a novel immunomodulating therapy causing autoimmune hypophysitis: a case report and review. European journal of endocrinology / European Federation of Endocrine Societies 167: 1-5

138 213. Caminschi I, Maraskovsky E, Heath WR. 2012. Targeting Dendritic Cells in vivo for Cancer Therapy. Frontiers in immunology 3: 13

214. den Haan JM, Lehar SM, Bevan MJ. 2000. CD8(+) but not CD8(-) dendritic cells cross-prime cytotoxic T cells in vivo. The Journal of experimental medicine 192: 1685-96

215. Caminschi I, Lahoud MH, Shortman K. 2009. Enhancing immune responses by targeting antigen to DC. European journal of immunology 39: 931-8

216. Idoyaga J, Lubkin A, Fiorese C, Lahoud MH, Caminschi I, Huang Y, Rodriguez A, Clausen BE, Park CG, Trumpfheller C, Steinman RM. 2011. Comparable T helper 1 (Th1) and CD8 T-cell immunity by targeting HIV gag p24 to CD8 dendritic cells within antibodies to Langerin, DEC205, and Clec9A. Proceedings of the National Academy of Sciences of the United States of America 108: 2384-9

217. Cheong C, Choi JH, Vitale L, He LZ, Trumpfheller C, Bozzacco L, Do Y, Nchinda G, Park SH, Dandamudi DB, Shrestha E, Pack M, Lee HW, Keler T, Steinman RM, Park CG. 2010. Improved cellular and humoral immune responses in vivo following targeting of HIV Gag to dendritic cells within human anti-human DEC205 monoclonal antibody. Blood 116: 3828-38

218. Huang XL, Fan Z, Borowski L, Rinaldo CR. 2008. Maturation of dendritic cells for enhanced activation of anti-HIV-1 CD8(+) T cell immunity. Journal of leukocyte biology 83: 1530-40

219. Rosenblatt J, Vasir B, Uhl L, Blotta S, Macnamara C, Somaiya P, Wu Z, Joyce R, Levine JD, Dombagoda D, Yuan YE, Francoeur K, Fitzgerald D, Richardson P, Weller E, Anderson K, Kufe D, Munshi N, Avigan D. 2011. Vaccination with dendritic cell/tumor fusion cells results in cellular and humoral antitumor immune responses in patients with multiple myeloma. Blood 117: 393-402

139