• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

The function of tumor associated leukocytes in ovarian cancer growth

Dissertation
Author: S. Peter G. Bak
Abstract:
Immunosuppressive leukocytes are emerging as a critical factor in facilitating tumor progression. These leukocytes are converted by the tumor microenvironment to become tolerogenic, facilitate metastasis, and to aid in neo-vascularization. The predominant variety of suppressive leukocytes found in ovarian cancer are referred to as vascular leukocytes (VLCs), due to cell surface markers of both dendritic cells and endothelial cells. Despite the large infiltration of VLCs in ovarian ascites, their derivation, mechanisms by which they facilitate tumor progression, and effect on the global immune system are not well understood. Using the ID8 tumor cell line, and focusing on peritoneal tumor growth, we further define the phenotype and function of VLCs. We show that carrageenan-mediated depletion of peritoneal tumor-associated leukocytes inhibits ovarian tumor progression. Moreover, Scavenger Receptor-A (SR-A) is robustly and specifically expressed on VLCs within human and murine ovarian tumor ascites. Administration of anti-SR-A immunotoxin to mice challenged with peritoneal ID8 tumors substantially inhibits peritoneal tumor burden and ascites accumulation. Additionally, VLCs harvested from the peritoneum of tumor bearing mice are immunosuppressive cells. Similar to a variety of tumor infiltrating leukocytes, arginase 1 (ARG1) activity is necessary for VLCs immunosuppression. Moreover, ID8 cell growth induces in vivo immunosuppression in splenocytes, and CD11b+ cells mediate this suppression. The elimination of SR-A+ cells from the peritoneum of tumor bearing mice relives this suppression. Extending these results we demonstrate that the protective anti-tumor effect of VLC elimination does not occur when CD8+ T Cells are concomitantly depleted. Further more, using a genetic model, short-term depletion of VLCs blocks immunosuppression and allows for efficacious vaccination against model antigens in tumor bearing mice. This study further delineates the interplay between tumor growth and the immune system and has implications on the design of cancer immunotherapies. The immunosuppressive phenotype of VLCs places them among the growing number of suppressive tumor associated leukocytes. These data add to the body of literature describing the interactions between cancerous lesions and the immune system.


 vi 
 Table of Contents

Abstract…………………………………………………………………………...……….ii Acknowledgements…………..……………………………………………..…………….iv Table of Contents………………………….……………………………………………...vi List of Figures……..……………...………………………………………………………ix

Chapter 1: Introduction……………………………..……………………………………1 1.1 Introduction to Cancer and the Immune System………………………………2 1.2 Inflammation and Cancer Development………………………………………3 1.3 Myeloid Cells and Cancer……………………………………………………..4 1.4 Myeloid Derived Suppressor Cells in Cancer…………………………………8 1.5 Mechanisms of MDSC Tumor Promotion…………………………………...10 1.6 Anti-Tumor Therapies Targeting MDSCs…………………………………...15 1.7 Introduction to Ovarian Cancer……………………………………………...19 1.8 Rationale and Scope of This Work…………………………………………..21

Chapter 2: Materials and Methods……………………………………………………...25

Chapter 3: SR-A + Leukocyte Depletions Inhibits Ovarian Cancer Growth………….....35
 3.1 Abstract…………………………………………………………….………...36
 3.2 Introduction…………………………………………………………………..36
 3.3 Results………………………………………………………………………..38


 vii 
 3.4 Discussion……………………………………………………………………43

Chapter 4: Vascular Leukocytes Require arginase-1 activity for T cell Suppression…...59 4.1 Abstract………………………………………………………………………………60

4.2 Introduction…………………………………………………………………..60 4.3 Results………………………………………………………………………..63 4.4 Discussion……………………………………………………………………70

Chapter 5: VLCs Potentiate Local and Distal Immunosuppression................................93
 5.1 Abstract………………………………………………………………………94 5.2 Introduction…………………………………………………………………..95 5.3 Results………………………………………………………………………..97 5.4 Discussion…………………………………………………………………..104

Chapter 6: Discussion…………………………………………………………………124

Appendix A:

Adjuvant Activity of Calreticulin……………………………………….137 A.1 Abstract…………………………………………………………………….138 A.2 Introduction………………………………………………………………...138 A.3 Materials and Methods……………………………………………………..141 A.4 Results……………………………………………………………………...146 A.5 Discussion……………………………………………………………….…151


 viii 
 References.......................................................................................................................168


 ix 
 List of Figures

Model 1 ID8 Ovarian Ascites and VLCs……………………………………………...21 Figure 1 Targeting of Leukocytes through Phagocytosis……………………………..48 Figure 2 Murine VLCs Express Scavenger Receptor…………………………………50 Figure 3 Scavenger Receptor Expression in Human Ovarian Cancer………………...52 Figure 4 Tissue Specific SR-A + Depletion with SRA-Zap……………………………54 Figure 5 Peritoneal SR-A Depletion Inhibits ID8 Ascites…………………………….56 Figure 6 SR-A Depletion Blocks ID8 Tumor Progression……………………………58 Figure 7 Murine Ovarian Ascites Contain CD11b + cells……...……..…………...…..76 Figure 8 VLCS exhibit cell surface markers of MDSCs……………..……………….78 Figure 9 Human Ovarian Carcinomas contain CD11c + CD11b + Cells…………...……80 Figure 10 VLCs Bind Folate………………………………………………………...….82 Figure 11 VLCs Suppress CD8 + and CD4 + T cell Responses……………..……..…….84 Figure 12 VLCs Inhibit Splenocyte IFNγ Production…..………..…………………….86 Figure 13 Characterization of Cells from Untransduced Tumors……..………...……...88 Figure 14 VLCs express ARG1 but not iNOS……………..…………………………...90 Figure 15 Arginase 1 Activity is Required for VLC Suppression……………..…..…...92 Figure 16. ID8 Tumor Growth Suppresses Splenic Responses……...………………...109 Figure 17 VLC Depletion Reverses Splenic Immunosuppression……………………111 Figure 18 SR-A Depletion Alters Local Cytokine Environment…………………...…113 Figure 19 SR-A depletion in RAG -/- 
Background is
Not
Effective………………………….115 Figure 20 Depletion of SR-A + cells Requires CD8 T Cells for Effects…………….....117


 x 
 Figure 21 Survival Mediated by VLC Depletion Requires CD8 T Cells…….…….....119 Figure 22 Short-term VLC Depletion Relieves Immunosuppression……………..…..121 Figure 23 VLC Depletion Potentiates the Induction of a Peptide Specific Immune Response………………………………………………………………...…123 Model 2 Immunosuppression in Ovarian Ascites…………………………………...136 Figure 24 Endotoxin-free Calreticulin does not Mature DCs……………………...…157 Figure 25 Analysis of Cytokines Elicited from DCs by CRTΔKDEL…………….…159 Figure 26 Full-length Calreticulin does not Induce in vitro DC Maturation…….…...161 Figure 27 Endotoxin-decontaminated Calreticulin Binds Peptide and DCs……………………………………………………………………..….163 Figure 28 Calreticulin does not Induce DC Maturation in vivo………………………165 Figure 29 CRT cannot Elicit an in vivo CTL Response without Ancillary Adjuvant…………………………………………………………167

1

Chapter 1

Introduction

2

1.1 Introduction to Cancer and the Immune System Cancer is defined as the uncontrolled growth and spread of transformed tissue within a host organism. Genetic alterations, the result of environmental insults or inherited genetic abnormalities, cause cells to become unresponsive to the normal physiological cues that regulate tissue homeostasis. However, much of the morbidity and mortality associated with cancer is a function of its ability to spread from tissue of origin and colonize distal sites, a process known as metastasis. Traditionally, much of cancer research focused on the cell intrinsic effects that cause cancer, however it is now clear that the immune system influences the development of cancer. The vertebrate immune system evolved to defend the host from infectious agents such as viruses, microbes, and parasites. The immune system is comprised of two arms, the “innate” and “adaptive” immune system. The innate immune system responds to pathogens through recognition by germ line encoded receptors, named pattern recognition receptors (PRR). The adaptive immune system recognizes and responds to variety of unique molecules, known as antigens, through reorganization of germline encoded transcripts of T and B cells receptors. This receptor diversity allows the immune system to respond to an array of molecules expressed by pathogens without previous exposure to said agent. During the development of B and T cells, cells harboring receptors that recognize self-molecules are deleted. This process protects the host against an attack by the immune system directed against self-antigens. However, breakdown of self-tolerance results in pathological autoimmune diseases such as diabetes, lupus, and multiple sclerosis.

3

Since cancer arises from host tissue, it was historically thought that the immune system remained ignorant of developing neoplasia. Recent research demonstrates that the immune system responds to, and shapes, the growth of cancer (1). Since genomic instability is a hallmark of cancer, many cancerous tissues express one or more mutated self-antigens, known as tumor associated antigens (TAAs) (2). In some instances, the immune system can recognize TAAs and eradicate transformed cells (3, 4). However, tumors orchestrate an environment that both inhibits immune responses directed against it and converts leukocytes into tumor promoting cells (5, 6). Leukocytes promote the development and continued growth of tumors by creating a permissive inflammatory environment.

1.2 Inflammation and Cancer Development Inflammation is critical to the clearance of pathogens by the immune system. Upon disruption of tissue, innate immune cells release soluble factors, such as cytokines and chemokines, that initiate the activation and infiltration of leukocytes. This leads to a resolution of infection through stimulation of an adaptive immune response. After clearance of the pathogen, the acute inflammation is resolved and the tissue returns to its normal homeostatic state. However, when inflammatory insults are not cleared, the chronic activation of the immune system leads to numerous pathological conditions, including cancer. Over 100 years ago Rudolph Virchow first described the infiltration of leukocytes into cancerous tissues (7). There is now an accepted link between inflammation and cancer growth, as sustained inflammation promotes the development of, and retards, an

4

immune response against an established cancer (8). Indeed, chronic inflammation associated with infection enhances incidence of gastric (H. pylori) and hepatocellular (Hepatitis Virus) cancer (9). Additionally, chronic tissue specific inflammatory diseases, such as colitis, increase the risk of cancer in the affected tissue (10). While the immune system is designed to recognize and eradicate damaged or transformed cells, it is paradoxical that inflammation would potentiate rather than resolve cancer (11). It has been proposed that the inflammation in the cancer environment is skewed toward a chronic state, rather than an acute response (12). Thus, chronically activated tumor associated leukocytes, specifically; myeloid cells take on a tumor promoting phenotype.

1.3 Myeloid Cells and Cancer Myeloid precursors emigrate from the bone marrow, populate peripheral tissues, and differentiate into a variety of cells distinguished by morphology, cell surface markers, and function. Among these cell types are myeloid dendritic cells (DCs), granulocytes, and macrophages. These cells are immunological sentinels, acting as first responders to disruption in homeostasis, and initiating signals to stimulate the innate and adaptive arms of the immune system. A distinction between DCs and macrophages can be made on the basis of cell surface markers and functional attributes, however some argue their common origin and function make the distinction less clear (13, 14). Macrophages are professional phagocytes, able to engulf and degrade cellular debris, release cytokines, and function as cytotoxic cells. In mice and humans, macrophages express the cell surface markers colony stimulating factor-1 receptor (CSFR-1 or c-fms), F4/80 antigen, CD68, and CD11b in varying intensity depending on

5

tissue location (15). In murine systems, and additional marker, the F4/80 antigen is expressed on macrophages (16). In addition to their ability to phagocytose pathogens and dying cells, macrophages release cytokines upon recognition of pathogen associated molecular patterns (PAMPs) by PRRs named Toll-like Receptors (TLRs) (17). Macrophages promote adaptive immune responses through the release of cytokines, and orchestrate cytotoxic responses through the release of reactive oxygen species (ROS) generated by enzymes such as inducible nitric oxide synthase (iNOS) (18). Depending on the cytokine milieu, macrophages are “polarized” toward different phenotypes. Similar to the Th1 and Th2 designation of T cells, Macrophages can be divided into M1 (“classically activated”) and M2 (“alternatively activated”) cells (19). M1 macrophages secrete IFNγ, IL-12, produce nitric oxide (NO) under the control of iNOS, and present antigen (20). M2 macrophages secrete IL-10, IL-4, express the enzyme arginase-1 (ARG1), and contribute to angiogenesis (21). While this designation is useful, some suggest that the states of macrophage activation should be defined solely upon their functions in homeostasis, wound healing, or immune regulation (22). DCs potently active naive T cells through their acquisition and processing of antigens. Murine myeloid DCs express the cell surface markers CD11c, DEC205, and MHC II (23). Cell surface markers on human DCs are heterogeneous and less well characterized (24). Human DCs arise from CD34 + progenitors and exists as CD14 + CD11c + CD1 - precursors in the blood (25). Myeloid DCs harvested from the spleen display a heterogeneous CD11c + CD11b + phenotype (24, 26). DCs express the specific cellular machinery necessary to degrade peptides, load them on to major histocompatability molecules (MHC), and stably present the peptide MHC complexes

6

(pMHC) on their cell surface (27). Upregulation of costimulatory molecules, such as CD80 and CD86, allow DCs to effectively stimulate T cells (28). DCs migrate to peripheral lymphoid tissues to initiate adaptive immune response, a phenotype that other myeloid cells lack (29). However, under non-inflammatory conditions, DCs capture and present antigens in the periphery to induce tolerance (30). Thus, DCs initiate cytotoxic responses against invading pathogens, and induce anergy against naïve T cells directed against self. While macrophages and DCs maintain tissue homeostasis, tolerance to self, and initiate inflammation in response to infection, their dysregulation can lead to instances of chronic infection, autoimmunity, and cancer. In the context of neoplasia, DCs and macrophages exposed to tumor derived factors (TDFs) take on a pro-tumoral role (31, 32). Indeed, the presence of DCs and macrophages in the tumor environment correlates with a poor clinical outcome. For example, in a number of tumor types, the presence of macrophages is an indicator of poorer survival (33), and DCs coordinate a tolerogenic environment that promotes cancer growth (34). The specific mechanisms used by DCs and macrophages to propagate these effects are varied. TAMs regulate tumor growth through the promotion of immunosuppression, matrix remodeling, and angiogenesis (35). TAMs suppress the immune system at the site of tumor growth through secretion of IL-10, which dampens anti tumor cytotoxic T lymphocytes (CTLs) (36). Furthermore, TAMs express TGF-β that induce T regulatory cells (T regs ), inhibit T cell responses, and promote de novo carcinogenesis (21). In order to metastasize, cancerous cells require the breakdown of the surrounding extracellular matrix (ECM) to facilitate extravasation through the basement membrane into the blood

7

stream (37). TAMs mediate this process through the expression of chemokines and matrix metalloprotinases (MMPs) such as MMP-2 and –9 (38). The development of new blood vessels, a process known as angiogenesis, to supply the tumor with an adequate supply of oxygen is a necessary step in the sustained growth of cancer. TAMs secrete pro angiogenic molecules such as vascular endothelial growth factor (VEGF), and can mediate the release of such molecules from the ECM through MMPs (9). Like macrophages, tumor associated DCs take on a distinctive phenotype that promotes the growth of cancer. Firstly, in many types of cancers, there are a reduced number of DCs compared to healthy tissues, negating the ability to raise a peptide specific response (39). While the immunogenicity of DCs partially relies on their ability to up-regulate co-stimulatory molecules, such as CD80 and CD86, DCs in tumors show a marked reduction in their stimulatory capacity (40, 41). The lack of co-stimulatory molecules on tumor associated DCs not only prevents the stimulation of an immune response against cancer antigens, but also induces bona fide tolerance to TAAs (42). Furthermore, tumor associated DCs remain at tumor lesions preventing the priming of T cells in lymph nodes (43). In addition to the failure to express stimulatory molecules, tumor associated DCs express inhibitory molecules, such as B7-H1, that engage receptors on T cells inducing anergy (44). In ovarian cancer the most numerous tumor-infiltrating leukocyte is the vascular leukocyte. First described in tumors generated from the ID8 cell line transduced with VEGF-A and β-defensin-29 (Defb29), VLCs express a unique set of cell surface molecules comprising myeloid (CD45, CD11c, DEC205, MHC Class II) and vascular (CD31, VE-Cadherin) markers (45). VLCs co-localize with functional vasculature in

8

solid tumors, and infiltrate ascites of mice injected intra peritoneally (ip) with ID8 cells (45). Subsequent human studies found defensins present in ovarian cancer tissues, and VLCs present around tumor capillaries. Moreover, VLCs implanted into immuno- deficient mice form functional blood vessels (46). Thus, presence of VLCs in solid tumor tissue promotes neovasculature allowing for rapid tumor growth. In this thesis, we report that peritoneal VLCs are immunosuppressive cells. Since VLCs arise from immature myeloid cells, VLCs function similarly to the broadly defined, heterogeneous class of cells known as myeloid derived suppressor cells (MDSCs).

1.4 Myeloid Derived Suppressor Cells in Cancer MDSCs encompass immature DCs, macrophages, and granulocytes, and have recently garnered much attention in the field of tumor immunology (47). Whereas immature myeloid cells emigrate from the bone marrow and populate peripheral tissues under homeostatic conditions, MDSCs are induced under conditions of infection, trauma, and cancer growth (48). In murine systems MDSCs are classically defined by the presence of the CD11b and Gr-1 cell surface markers (49). Additional MDSC markers in different models include: CD115 (50-52), F4/80 (53), CD11c (53), IL-4Rα (50), and CD80 (54). Human MDSC have a heterogeneous cell surface phenotype but are generally considered to be CD11b + CD33 - (55). The expression of CD14 is reported on some human MDSCs but not others (56-62). However, a recent study by Zanovello and colleagues defines IL-4Rα as a marker of MDSCs from a variety of human cancers (63).

9

In addition to tumor tissue, MDSCs localize to secondary lymphoid tissues. Stimulation of the immune system through viral infection or tumor inoculation induces MDSCs in the spleen and bone marrow (BM) in a GM-CSF dependent mechanism (64). During primary tumor growth, MDSCs accumulate in lymphoid tissues in numerous mouse models (50, 52, 65). Furthermore, on a per cell basis, Gr-1 + /CD11b + cells from spleens of tumor bearing mice are more immunosuppressive than their counterparts from non-tumor bearing animals (52). The expansion of MDSCs in tumor bearing hosts is regulated by TDFs that control myeloid cell function. Cytokines secreted by tumors, such as GM-CSF and stem cell factor (SCF), are important for homeostatic myelopoesis, but induce suppressive myeloid cells within the tumor environment (66, 67). In addition to cytokines involved in myelopoesis, cytokines that modulate myeloid cell differentiation are important in MDSC development, such as IL-6 and VEGF (68, 69). Inflammatory cytokines, frequently present in tumor tissue, promote MDSCs. IL- 1β, a pro-inflammatory mediator, recruits MDSCs to tumor lesions (70). Furthermore, MDSCs express the IL-1 receptor (IL-1R), and mice with a genetic deficiency for IL-1R contain a reduced number of MDSCs (71). In addition to proteins, the inflammatory lipid Prostaglandin E2 (PGE2) is necessary for MDSC expansion (72). Indeed, Rodriguez et al report that tumor cells express cyclooxygenase (COX)-2, an enzyme responsible for PGE2 synthesis (73). These TDFs promote specific intracellular signaling programs in MDSCs. Signal transducer and activator of transcription 3 (STAT3) is a master regulator of immunosuppressive networks in the tumor environment (74). TDFs that expand MDSCs,

10

such as VEGF and IL-10, induce the expression of STAT3 (75, 76). Indeed, inhibition of STAT3 within the myeloid compartment reverses the suppressive phenotype of MDSCs (77). These networks of tumor derived factors and intracellular signaling programs generate a specific tumor promoting phenotype in MDSCs.

1.5 Mechanisms of MDSC Tumor Promotion Upon expansion at, or recruitment to, the site of carcinogenesis MDSCs promote tumor growth. The function of MDSCs in the tumor microenvironment is multifaceted. MDSCs express enzymes, cell surface receptors, and secreted factors that inhibit the immune system. Through the production of cytokines and promotion of neovasculature, MDSCs act as pro-angiogenic agents for growing tumors. The ability of MDSCs to alter the tumor environment also allows tumors to metastasize. MDSCs exert their suppressive function through a number of distinct pathways. One such pathway is the expression of ROS. MDSCs harvested from numerous types of tumors produce high levels of ROS (52, 78). Indeed, MDSCs express specific intra cellular machinery to produce ROS, including NADPH oxidase (NOX2), iNOS, and ARG1 (73, 79). The presence of ROS has a two-fold effect in MDSCs. Firstly; ROS expression by MDSCs maintains their immature, suppressive, phenotype since reduction of ROS leads to maturation of the myeloid cells (78). Secondly, ROS production by MDSCs inhibits T cell activity (80). A number of MDSC derived ROS mediate a suppressive pathway. Kusmartsev et al demonstrate that chelating H 2 0 2 through the addition of catalase inhibits the MDSC suppressive phenotype (81). The presence of peroxynitrites generated by MDSCs alters

11

the function of tumor infiltrating CD8 T cells. Peroxynitrites react with proteins leaving nitrated tyrosine residues. Indeed, TILs from prostate cancer patients contain nitrated tyrosine residues, and this nitration correlates with T cell inactivity (82). The addition of enzymatic inhibitors of peroxynitrite formation reversed TIL anergy. A recent study defines the mechanism for peroxynitrite induced T cell unresponsiveness (83). The presence of MDSC-dependent nitrations on the TCR complex of TILs renders them unable to bind their cognate pMHC complex (83). In addition to ROS production, the breakdown of L-Arginine (L-Arg) by ARG1 and iNOS is important mediators of MDSC biology. ARG1 and iNOS catabolize L- Arginine and produce ROS as a consequence (84). Studies demonstrate their role in tumor-induced immunosuppression as a function of ROS production, the mechanisms of which are outlined above (81, 85). However, they coordinate additional pathways to exert immunosuppressive effects. The catabolism of L-Arg by iNOS produces NO. NO effects homeostatic and pathological immune function (86). In the tumor environment NO dampens T cell activity by two different mechanisms. The phosphorylation of proteins, such as STAT5 and JAK3, important in IL-2 signaling, are required for T cell activation. Elevated NO levels, commonly found in the tumor environment, perturbs the phosphorylation of STAT5 and JAK3 (87, 88). Further, exposure of T cells to NO levels comparable to those in tumor tissues induces their apoptosis (89, 90). The immunosuppressive effects of ARG1 activity are thought to be two fold. Firstly, since L-Arg is a necessary factor for cell growth and division, the local depletion of L-Arg through ARG1 activity prevents T cells from undergoing rapid cell division

12

upon antigen recognition (91). Indeed, exhausted T cells from pathological conditions, such as hepatitis B infected patients, can be “revived” through supplementation with L- Arg (92). Secondly, the activity of ARG1 directly impacts the expression of critical signaling molecules in T cells. The presence of 3 rd party ARG1 positive cells causes hyporesponsiveness of stimulated CD8 + T cells (93). This blunted T cell response is a function of the down regulation of the ζ chain subunit of the TCR (93). Upon recognition of its cognate pMHC complex, the TCR initiates a signaling cascade through adaptor proteins bound to the TCR. Without proper expression of TCR ζ chain, T cells cannot undergo the intracellular signaling program necessary for full activation. Indeed, T cells from ovarian cancer patients express lower levels of TCR ζ chain as compared to healthy controls (94). MDSCs maintain ARG1 activity in numerous tumor models (95). In mice, MDSCs harvested from such cancerous tissues such as colon, breast, lung, lymphoma, and fibrosarcomas express ARG1 (50, 52, 81, 96-99). In agreement with animal studies, MDSCs from human patients express ARG1 (60, 84). In most tumor models ARG1 activity is directly tied to the MDSC suppressive capacity, as inhibition of ARG1 with the small molecule N ω -hydroxy-nor-Arginine (nor-NOHA), abrogates the suppress effect of MDSCs (81). Moreover, ARG1 function in MDSCs has in vivo consequences for tumor growth. Rodriguez et al report that the administration of nor-NOHA leads to reduced growth the Lewis Lung Carcinoma (3LL) cell line (96). Thus, ARG1 activity in MDSCs is a critical factor in their ability to promote tumor growth through the suppression of the immune system. Interestingly, tumor derived prostaglandins, a molecule known to induce MDSCs, leads to the expression of ARG1 in human myeloid cells (73). This suggests that

13

ARG1 expression may be tied to the normal development of MDSCs under pathological conditions. However, the effect of ARG1 inhibition on in vivo TCR ζ chain expression in cancer models remains to be evaluated. In other models, a contact dependent, peptide specific, interaction between MDSCs and T cells is necessary for suppression. Using transgenic OT-1 T cells expressing a TCR specific for the SIINFEKL epitope of chicken ovalbumin (OVA), Kusmartsev et al demonstrate that MDSCs isolated from tumor bearing mice increase ROS production when cultured in the presence of T cells and SIINFEKL peptide, but not control peptide (81). The increase depends on ARG1 activity, as addition of nor-NOHA reduces ROS levels. Further studies by this group define the importance of this peptide interaction, as MDSCs suppress TCR transgenic T cells only in the presence of their cognate epitope (100). Furthermore, MDSCs isolated from tumors in H2 d mice are unable to suppress H2 b restricted T cells when transferred into H2 b /H2 d F1 mice (100). Whether this MHC/TCR interaction allows proximity for ROS and/or substrate depletion to mediate their effects, or it initiates a specific signaling cascade remains to be determined. In addition to these cell intrinsic effects, MDSCs are also able induce T regs . T regs

are CD4 + T cells that express the fork-head transcription factor FOXP3 and inhibit the activity of CD8 T cells (101). In humans and mice, T regs localize to tumors and potentiate their growth (102). The mechanism by which MDSCs induce T regs varies by model. In an ovarian cancer model the tumor promoting effect of MDSCs is mediated by induction of T regs (54). The presence of CD80 on MDSC is critical for this conversion, as antibody blockade of CD80 or its ligand CD152 reduce the number of tumor infiltrating T regs and slows the growth of tumors. The ability of IL-10 and TGFβ to convert CD4 + T cells into

14

suppressive T regs is well documented (101, 103). In a colon cancer model, MDSCs produce IL-10 and TGFβ that are necessary for the induction of T regs (51). In the presence of MDSCs from mice with B cell lymphoma FOXP3 + T regs undergo an expansion that is not dependent on TGFβ, but is dependent on ARG1 activity in MDSCs. MDSCs harvested from other tumors, such as glioma, secrete TGFβ (53). Whether or not this exerts an immunosuppressive effect directly on T cells or through T reg induction is not clear. In addition to immunosuppression, MDSCs promote angiogenesis and vasculargenesis (104). During cancer growth, it is critical that the tumor maintains a supply oxygen and nutrients to the center of the growing tumor mass. The lack of oxygen at the site of tumor growth, a condition termed hypoxia, induces the expression of hypoxia-induced factor 1α (HIF1α), which activates transcription of genes that support angiogenesis and vasculogenesis. In a mouse model of glioblastoma, HIF1α recruits MDSCs to the site of tumor growth whereupon they promote tumor neovascularzation (105). In a colorectal cancer model, MDSCs directly intercalate into the tumor vascular network and MDSCs mediate the release of VEGF from the ECM through MMP9 activity (106). In addition to MMP9 mediated release of VEGF from the ECM, MDSCs can directly secrete VEGF (107). In a mammary tumor model, CD11b + cells from the bone marrow facilitate tumor growth through MMP9 dependent vasculogenesis (108). MDSCs induce the migration and vessel formation of endothelial cells in vitro under the control of STAT3 (107). Recent reports describe a novel pro-angiogenic molecule, Bv8, with similar properties to VEGF (109). Similar to VEGF, MDSCs secrete Bv8, and anti- Bv8 treatment of mice slows tumor growth (109).

15

MDSCs facilitate metastasis through their ability to re-organize the local tumor environment. MDSCs localize to the invasive front of mammary carcinomas (110). Indeed, an admixture of the 4T1 mammary tumor cell line and MDSCs increased the metastatic lesions in surrounding tissue (110). In murine ovarian cancer model, reduced myeloid cell recruitment to the peritoneum of tumor bearing mice decreases the spread of metastatic nodules throughout the abdominal wall (111).

1.6 Anti-Tumor Therapies Targeting MDSCs As the importance of immunosuppressive cells in the progression of cancer is clear, efforts have begun to develop novel strategies to target the activity of MDSCs. Current treatments include small molecule inhibitors to interfere with MDSCs function, cytotoxic drugs affecting MDSC expansion, and agents to cause the differentiation of MDSCs into mature myeloid cells. Moreover, current biologic therapies used to treat cancers are under investigation for their ability to modulate MDSCs biology. Targeting the ability of MDSCs to produce ROS, and subsequently reducing their immunosuppressive capacity, is an attractive therapeutic avenue. As outlined above, the administration of nor-NOHA can inhibit MDSC function in vitro as well as reduce tumor growth in vivo. This strategy has potential in humans, as the combination of the iNOS inhibitor N-monomethyl-L-arginine (L-NMMA) and nor-NOHA restored the activity of TILs from human prostate carcinomas in vitro (82). Nitroasprin, which contains an NO- donor moiety that inhibits iNOS activity and an aspirin group that blocks ARG1 function, is therapeutically efficacious in animal models (112). In the CT26 colon cancer model, nitroaspirin restores T cell proliferation through the reduction of ARG1 and iNOS

Full document contains 211 pages
Abstract: Immunosuppressive leukocytes are emerging as a critical factor in facilitating tumor progression. These leukocytes are converted by the tumor microenvironment to become tolerogenic, facilitate metastasis, and to aid in neo-vascularization. The predominant variety of suppressive leukocytes found in ovarian cancer are referred to as vascular leukocytes (VLCs), due to cell surface markers of both dendritic cells and endothelial cells. Despite the large infiltration of VLCs in ovarian ascites, their derivation, mechanisms by which they facilitate tumor progression, and effect on the global immune system are not well understood. Using the ID8 tumor cell line, and focusing on peritoneal tumor growth, we further define the phenotype and function of VLCs. We show that carrageenan-mediated depletion of peritoneal tumor-associated leukocytes inhibits ovarian tumor progression. Moreover, Scavenger Receptor-A (SR-A) is robustly and specifically expressed on VLCs within human and murine ovarian tumor ascites. Administration of anti-SR-A immunotoxin to mice challenged with peritoneal ID8 tumors substantially inhibits peritoneal tumor burden and ascites accumulation. Additionally, VLCs harvested from the peritoneum of tumor bearing mice are immunosuppressive cells. Similar to a variety of tumor infiltrating leukocytes, arginase 1 (ARG1) activity is necessary for VLCs immunosuppression. Moreover, ID8 cell growth induces in vivo immunosuppression in splenocytes, and CD11b+ cells mediate this suppression. The elimination of SR-A+ cells from the peritoneum of tumor bearing mice relives this suppression. Extending these results we demonstrate that the protective anti-tumor effect of VLC elimination does not occur when CD8+ T Cells are concomitantly depleted. Further more, using a genetic model, short-term depletion of VLCs blocks immunosuppression and allows for efficacious vaccination against model antigens in tumor bearing mice. This study further delineates the interplay between tumor growth and the immune system and has implications on the design of cancer immunotherapies. The immunosuppressive phenotype of VLCs places them among the growing number of suppressive tumor associated leukocytes. These data add to the body of literature describing the interactions between cancerous lesions and the immune system.