You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Immunosuppressive myeloid cells under normal and neoplastic conditions Hamilton, Melisa June 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2012_spring_hamilton_melisa.pdf [ 3.71MB ]
Metadata
JSON: 24-1.0072452.json
JSON-LD: 24-1.0072452-ld.json
RDF/XML (Pretty): 24-1.0072452-rdf.xml
RDF/JSON: 24-1.0072452-rdf.json
Turtle: 24-1.0072452-turtle.txt
N-Triples: 24-1.0072452-rdf-ntriples.txt
Original Record: 24-1.0072452-source.json
Full Text
24-1.0072452-fulltext.txt
Citation
24-1.0072452.ris

Full Text

IMMUNOSUPPRESSIVE MYELOID CELLS UNDER NORMAL AND NEOPLASTIC CONDITIONS by MELISA JUNE HAMILTON B.Sc., The University of British Columbia, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Experimental Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2011  © Melisa June Hamilton, 2011  ABSTRACT Although the importance of immunomodulatory myeloid cells in both normal physiology and carcinogenesis is well established, many questions remain regarding the specific roles and regulation of these cells. In this thesis, we explore the immunosuppressive features of macrophages (Mφs) and elucidate the mechanisms by which they suppress T cell proliferation/activation, the factors that regulate their suppressive properties, the relative potency of Mφ suppression compared to other myeloid cells, such as myeloid‐derived suppressor cells (MDSCs), and the role these cells play in promoting tumor growth and metastasis. We demonstrate herein that in response to interferon (IFN)‐γ, which is secreted by activated T cells, resident Mφs from non‐tumor‐bearing mice acquire immunosuppressive properties that are mediated by nitric oxide (NO). Moreover, our data reveal a novel role for Toll‐like receptor (TLR)‐induced IFN‐β in regulating the immunosuppressive properties of Mφs. We also demonstrate for the first time that in vitro culture conditions profoundly affect the immunosuppressive functions of MDSCs. Specifically, we show that serum antagonizes the suppressive abilities of MDSCs from 4T1 tumor‐bearing mice and that the major serum protein albumin mediates these effects, in part by reducing reactive oxygen species (ROS) production from MDSCs. These findings have important implications, since the accurate detection and quantification of immunosuppression is critical for both the identification and functional analysis of tumor‐induced MDSCs. We also explore the phenotypic and functional heterogeneity of tumor‐induced myeloid cells and compare the immunosuppressive functions of different populations isolated from normal and tumor‐bearing mice. We show that tumors that induce the accumulation of myeloid cells also enhance the suppressive functions of these cells. In addition, we demonstrate that, in vitro, tumor‐induced Mφs are significantly more  ii  potent immune suppressors than tumor‐induced MDSCs on a per cell basis, and suppress T cell responses via distinct mechanisms. Finally, we present data showing that treating metastatic mammary tumor‐bearing mice with all‐trans‐retinoic acid (ATRA) decreases MDSCs, increases Mφs, and enhances metastatic growth. Taken together, these findings advance our understanding of the factors that regulate myeloid cell functions in normal and neoplastic tissues and may lead to improved immunotherapies to treat human disease.  iii  PREFACE All experiments were conducted by Melisa J. Hamilton, with the exception of those described below. I designed the studies presented in this dissertation, analyzed and interpreted all the data, and composed and edited the thesis. A version of Chapter 3 was previously published. Hamilton MJ, Antignano F, von Rossum A, Boucher JL, Bennewith KL, Krystal G. “TLR agonists that induce IFN‐β abrogate resident macrophage suppression of T cells”. The Journal of Immunology, 185(8):4545‐4553, 2010. Copyright 2010. The American Association of Immunologists, Inc. Several of the experiments in Chapter 3 were conducted with the technical assistance of co‐operative education students Anna von Rossum (Terry Fox Laboratory), who worked under my supervision, and Kim Snyder (Terry Fox Laboratory), who worked under the supervision of Vivian Lam (Terry Fox Laboratory). In addition, Dr. Frann Antignano (Terry Fox Laboratory) developed the IFN‐β ELISA in collaboration with Victor Ho (Terry Fox Laboratory), and participated in helpful discussions that led to hypothesis generation and experimental ideas. Dr. Scott Patterson (CFRI) assisted with the development of the Foxp3 staining protocol. A version of Chapter 4 is currently in press. Hamilton MJ, Banáth JP, Lam V, Lepard NE, Krystal G, Bennewith KL. “Serum inhibits the immunosuppressive function of myeloid‐derived suppressor cells isolated from 4T1 tumor‐bearing mice.” Cancer Immunology Immunotherapy. Studies in Chapters 4 and 5 involving tumor models were performed in collaboration with Dr. Kevin Bennewith’s laboratory, with the technical assistance of Dr. Judit Banáth, Nancy LePard, Momir Bosiljcic, Denise McDougal, Jessica Jia, and Dr. Kevin Bennewith, (Department of Integrative Oncology, BC Cancer Research Centre). Specifically, lab members assisted with tumor injections, tissue processing, immunofluorescence analyses, clonogenic assays, and some flow cytometry analyses. Victor Ho also assisted with tissue collection, though did not contribute directly to any specific figure. In addition, co‐operative education students Kim Snyder and Ann Hsu‐  iv  An Lin (Terry Fox Laboratory), under the supervision of me, Vivian Lam, and/or Victor Ho, assisted with Western blot analyses. In Chapter 4, the microscopy in Fig. 4.3 was performed by Dr. Judit Banáth and Dr. Kevin Bennewith, and Vivian Lam prepared the fetal calf serum that was used in Fig. 4.5 and Fig. 4.7. In Chapter 5, members of Dr. Kevin Bennewith’s laboratory collected the raw data presented in Fig. 5.1, Fig. 5.2, Fig. 5.3, Fig. 5.28, and Fig. 5.30A. Mice were housed in the Animal Resource Centre (ARC) with the assistance of ARC staff. Mouse studies were performed under UBC Animal Care protocols A07‐0221 (Dr. Gerald Krystal) and A09‐0251 (Dr. Kevin Bennewith). Drs. Gerald Krystal and Kevin Bennewith contributed to experimental design, data interpretation, and editing of the dissertation.  v  TABLE OF CONTENTS ABSTRACT ......................................................................................................................................... ii PREFACE............................................................................................................................................ iv TABLE OF CONTENTS ................................................................................................................... vi LIST OF TABLES ................................................................................................................................ x LIST OF FIGURES .............................................................................................................................xi LIST OF ABBREVIATIONS .......................................................................................................... xiv ACKNOWLEDGEMENTS .............................................................................................................. xix DEDICATION ....................................................................................................................................xx CHAPTER 1 : INTRODUCTION ......................................................................................................1 1.1  History of tumor immunology .................................................................................................................... 1  1.2  Cancer immunoediting versus tumor‐induced immune suppression ....................................... 2  1.3  Brief overview of the immune system .................................................................................................... 3  1.4  Tolerance versus autoimmunity................................................................................................................ 5  1.5  TLR signaling ..................................................................................................................................................... 6  1.6  Immune escape ............................................................................................................................................... 11  1.7  Pro‐ and anti‐tumor effects of the immune system ......................................................................... 12 1.7.1  Mφs ....................................................................................................................................................... 16 1.7.1.1 Mφ activation ................................................................................................................. 19 1.7.1.2 TAMs ................................................................................................................................. 23  1.7.2  Myeloid‐derived suppressor cells ........................................................................................... 26  1.8  Tumor metastasis and the pre‐metastatic niche .............................................................................. 29  1.9  4T1 and 67NR murine mammary tumor models ............................................................................. 31  1.10  Aims of study ................................................................................................................................................... 33  CHAPTER 2 : MATERIALS AND METHODS ............................................................................ 35 2.1  Mice ..................................................................................................................................................................... 35  2.2  Media................................................................................................................................................................... 35  2.3  Reagents and cytokines ............................................................................................................................... 36  2.4  Isolation of myeloid cells ............................................................................................................................ 37 2.4.1  Peritoneal macrophages ............................................................................................................. 37  2.4.2  Splenic, pulmonary, or tumor‐associated Mφs and MDSCs .......................................... 37  2.5  Flow cytometry ............................................................................................................................................... 38  2.6  T cell proliferation assay and cytokine assays................................................................................... 40 vi  2.7  In vitro Mφ skewing ....................................................................................................................................... 41  2.8  SDS‐PAGE and Western blot analysis .................................................................................................... 41  2.9  Viability and morphology assays............................................................................................................. 42  2.10  Nitric oxide assay ........................................................................................................................................... 43  2.11  In vitro Mφ pre‐treatment and stimulation ......................................................................................... 43  2.12  IFN‐β ELISA ...................................................................................................................................................... 44  2.13  Tumor models ................................................................................................................................................. 44  2.14  Serum treatment for MDSC assay studies............................................................................................ 45  2.15  ROS Detection .................................................................................................................................................. 45  2.16  Clonogenic assays .......................................................................................................................................... 46  2.17  Immunofluorescence .................................................................................................................................... 46  2.18  Statistical analysis ......................................................................................................................................... 47  CHAPTER 3 : TOLL­LIKE RECEPTOR AGONISTS THAT INDUCE IFN­β ABROGATE RESIDENT MACROPHAGE SUPPRESSION OF T CELLS ....................................................... 48 3.1  Introduction ..................................................................................................................................................... 48  3.2  Results ................................................................................................................................................................ 49 3.2.1  Characterization of the immunosuppressive properties of PC cells ......................... 49  3.2.2  Resident PMφs exhibit a naïve phenotype ........................................................................... 50  3.2.3  Resident Mφs suppress T cell proliferation and cytokine production ..................... 51  3.2.4  Mφs suppress T cell proliferation via a contact‐dependent mechanism................. 54  3.2.5  Mφs suppress T cell proliferation via IFN‐γ‐induced NO production ....................... 57  3.2.6  Pre‐treatment with LPS and dsRNA, but not CpG or PGN, decreases the ability of Mφs to produce NO and suppress T cells .......................................................... 59  3.2.7  IFN‐β contributes to the reduced suppressive abilities of LPS and dsRNA pre‐treated Mφs .............................................................................................................................. 62  3.2.8 3.3  Inhibition of Arg1 reduces the effect of LPS and dsRNA pre‐treatment of Mφs... 65  Discussion ......................................................................................................................................................... 66  CHAPTER 4 : SERUM INHIBITS THE IMMUNOSUPPRESSIVE FUNCTION OF MYELOID­DERIVED SUPPRESSOR CELLS ISOLATED FROM 4T1 TUMOR­ BEARING MICE ............................................................................................................................... 71 4.1  Introduction ..................................................................................................................................................... 71  4.2  Results ................................................................................................................................................................ 72 4.2.1  Validation of in vitro T cell proliferation assay systems ................................................ 72  vii  4.2.2  4T1‐induced MDSCs only suppress T cell proliferation under serum‐free conditions .......................................................................................................................................... 73  4.2.3  Serum does not increase the viability of 4T1‐induced MDSCs.................................... 75  4.2.4  Serum directly inhibits 4T1‐induced MDSC immunosuppression ............................ 76  4.2.5  FCS inhibits MDSC immunosuppression in a dose‐dependent manner .................. 76  4.2.6  Effect of different serum treatments on MDSC immune suppression...................... 78  4.2.7  BSA blunts 4T1‐induced MDSC immunosuppression ..................................................... 80  4.2.8  BSA antagonizes 4T1‐induced MDSC immunosuppression by inhibiting ROS production......................................................................................................................................... 82  4.3  Discussion ......................................................................................................................................................... 84  CHAPTER 5 : MACROPHAGES ARE MORE POTENT IMMUNE SUPPRESSORS THAN MYELOID­DERIVED SUPPRESSOR CELLS IN MURINE METASTATIC MAMMARY CARCINOMA ............................................................................................................. 89  5.1  Introduction ..................................................................................................................................................... 89  5.2  Results ................................................................................................................................................................ 91 5.2.1  Characterization of primary tumor growth and metastasis in 4T1 and 67NR tumor models................................................................................................................................... 91  5.2.2  4T1 but not 67NR tumors induce MDSCs and Mφs .......................................................... 93  5.2.3  Phenotypic characterization of myeloid cells from control versus tumor‐ bearing mice ..................................................................................................................................... 95 5.2.3.1 TAMs ................................................................................................................................. 95 5.2.3.2 PMφs .................................................................................................................................. 97 5.2.3.3 Splenic Mφs ..................................................................................................................... 97 5.2.3.4 Splenic Gr1+ cells ....................................................................................................... 100  5.2.4  Summary of the effect of 4T1 tumors on the phenotypes of myeloid cells ......... 100  5.2.5  Comparison of immunosuppressive abilities of myeloid cells from control versus tumor‐bearing mice ..................................................................................................... 102 5.2.5.1 PMφs ............................................................................................................................... 102 5.2.5.2 TAMs .............................................................................................................................. 103 5.2.5.3 SpMφs ............................................................................................................................. 103 5.2.5.4 Splenic Gr1+ cells ....................................................................................................... 107 5.2.5.5 Gr1+ cells from different tissues ......................................................................... 109  5.2.6  Summary of the relative immunosuppressive abilities of different myeloid cell populations from 4T1 tumor‐bearing mice ............................................................. 111 viii  5.2.7  4T1‐induced MDSCs and TAMs, but not PMφs, decrease T cell viability ............. 119  5.2.8  4T1‐induced MDSCs suppress T cells via a contact‐independent mechanism .. 121  5.2.9  PMφs from control and 67NR tumor‐bearing mice suppress via a NO‐ dependent mechanism.............................................................................................................. 123  5.2.10 The immunosuppressive properties of 4T1 PMφs are not abrogated by TLR stimulation ..................................................................................................................................... 125 5.2.11 4T1 PMφ suppression of T cells is reversed by N‐acetyl‐cysteine .......................... 126 5.2.12 The immunosuppressive properties of 4T1 pulmonary MDSCs are inhibited by catalase ...................................................................................................................................... 128 5.2.13 Mφs and MDSCs suppress T cell responses via different ROS‐mediated mechanisms ................................................................................................................................... 130 5.2.14 ATRA increases lung metastasis in 4T1 tumor‐bearing mice................................... 133 5.2.15 ATRA increases the proportion of Mφs, which are much more immunosuppressive than MDSCs in the lungs of 4T1 tumor‐bearing mice ....... 135 5.3  Discussion ...................................................................................................................................................... 139  CHAPTER 6 : SUMMARY AND FUTURE DIRECTIONS ...................................................... 149 REFERENCES ................................................................................................................................ 160  ix  LIST OF TABLES CHAPTER 1 Table 1.1  Summary of the properties of known TLRs .......................................................................... 8  Table 1.2  Diversity of Mφs in different tissues....................................................................................... 19  CHAPTER 2 Table 2.1  List of antibodies used in this thesis ...................................................................................... 39  x  LIST OF FIGURES CHAPTER 1 Figure 1.1  Immune cells contribute to both the promotion and inhibition of tumor growth ................................................................................................................................................ 13  Figure 1.2  Development and functions of the mononuclear phagocytic lineage ...................... 17  Figure 1.3  Polarization of Mφ phenotypes................................................................................................. 21  Figure 1.4  TAMs promote tumorigenesis via multiple mechanisms .............................................. 25  CHAPTER 3 Figure 3.1  Resident PMφs possess immunosuppressive properties .............................................. 50  Figure 3.2  Resident PMφs exhibit a naïve phenotype and can be skewed to either M1 or M2 with appropriate stimulation ............................................................................................ 51  Figure 3.3  Mφs suppress T cell proliferation and cytokine production in a dose‐ dependent manner ........................................................................................................................ 53  Figure 3.4  Mφs do not increase the proportion of dead T cells......................................................... 54  Figure 3.5  Mφs suppress T cell proliferation via a contact‐dependent mechanism................. 56  Figure 3.6  Mφs suppress T cell proliferation through NO production ........................................... 58  Figure 3.7  IFN‐γ is required for Mφ NO production ............................................................................... 59  Figure 3.8  Pre‐treatment with TLR agonists that signal through TRIF desensitizes Mφs to subsequent IFN‐γ stimulation, decreasing the ability of Mφs to produce NO and suppress T cells ...................................................................................................................... 61  Figure 3.9  Pre‐treatment with LPS, but not CpG, reduces Mφ sensitivity to subsequent stimulation ........................................................................................................................................ 62  Figure 3.10  LPS stimulation induces production of a secondary factor that inhibits IFN‐γ‐ induced NO production by Mφs ................................................................................................ 63  Figure 3.11  IFN‐β contributes to the reduced suppressive abilities of LPS and dsRNA pre‐treated Mφs .............................................................................................................................. 64  Figure 3.12  The Arg1 inhibitor BEC reduces the effects of LPS and dsRNA pre‐treatment .... 65  Figure 3.13  The ability of resident Mφs to suppress T cell proliferation is abrogated when Mφs are pre‐treated with LPS or dsRNA, but not CpG or PGN ........................ 67  xi  CHAPTER 4 Figure 4.1  Polyclonal‐ and Ag‐specific‐stimulated T cells proliferate in both serum‐free and serum‐containing conditions ........................................................................................... 73  Figure 4.2  MDSCs from 4T1 tumor‐bearing mice only suppress T cell proliferation under serum‐free conditions .................................................................................................... 74  Figure 4.3  Serum does not alter the proportion of different MDSC subtypes ............................ 75  Figure 4.4  Serum directly inhibits the immunosuppressive abilities of MDSCs........................ 77  Figure 4.5  The inhibitory effects of serum on 4T1‐induced MDSC immunosuppression are dose‐dependent ...................................................................................................................... 78  Figure 4.6  The effects of serum on 4T1‐induced MDSCs cannot be reversed by filtration, dialyzation, or heat inactivation of the serum ................................................................... 79  Figure 4.7  BSA reduces the immunosuppressive abilities of 4T1‐induced MDSCs .................. 81  Figure 4.8  Serum albumin restricts ROS production by 4T1‐induced MDSCs ........................... 83  CHAPTER 5 Figure 5.1  Characterization of 4T1 and 67NR tumor growth ........................................................... 92  Figure 5.2  Visualization of hypoxia and myeloid cell infiltration in 4T1 and 67NR tumors................................................................................................................................................. 93  Figure 5.3  Induction of myeloid cells by 4T1 and 67NR tumors...................................................... 94  Figure 5.4  Phenotypic characterization of TAMs ................................................................................... 96  Figure 5.5  Phenotypic characterization of PMφs .................................................................................... 98  Figure 5.6  Phenotypic characterization of SpMφs .................................................................................. 99  Figure 5.7  Phenotypic characterization of splenic Gr1+ cells ......................................................... 101  Figure 5.8  Analysis of protein expression in different myeloid cell populations................... 102  Figure 5.9  Tumors augment the immunosuppressive abilities of PMφs ................................... 104  Figure 5.10  67NR TAMs are more immunosuppressive than 4T1 TAMs .................................... 105  Figure 5.11  SpMφs are only immunosuppressive at high concentrations ................................... 106  Figure 5.12  4T1 tumors enhance the immunosuppressive abilities of splenic Gr1+ cells .... 108  Figure 5.13  4T1  Gr1+  cells  isolated  from  different  tissues  are  equally  immunosuppressive .................................................................................................................. 110 Figure 5.14  4T1 PMφs are more potent suppressors of T cell proliferation than 4T1 MDSCs on a per cell basis......................................................................................................... 112  xii  Figure 5.15  4T1 PMφs are more potent suppressors of T cell cytokine production than 4T1 MDSCs on a per cell basis ............................................................................................... 113  Figure 5.16  Visualization of the immunosuppressive effects of different myeloid cell populations .................................................................................................................................... 115  Figure 5.17  Summary of the ability of different myeloid cell populations to inhibit T cell division as determined by flow cytometric analysis of CFSE intensity ................ 117  Figure 5.18  Comparison of the ability of different myeloid cell populations to reduce the proportion of T cells................................................................................................................... 118  Figure 5.19  Effect of different myeloid cell types on cell viability .................................................. 120  Figure 5.20  At high concentrations, both MDSCs and PMφs suppress T cells via contact‐ independent mechanisms........................................................................................................ 122  Figure 5.21  Control PMφs, but no other cell types tested, suppress T cell responses via NO production .............................................................................................................................. 124  Figure 5.22  The immunosuppressive properties of 4T1 PMφs are not reversed by pre‐ treatment with TLR ligands .................................................................................................... 125  Figure 5.23  NAC reduces the immunosuppressive abilities of 4T1 PMφs ................................... 127  Figure 5.24  The immunosuppressive properties of 4T1 MDSCs are not reversed by pre‐ treatment with TLR ligands .................................................................................................... 128  Figure 5.25  Catalase reduces the suppressive abilities of 4T1 MDSCs ......................................... 129  Figure 5.26  4T1 Mφs and MDSCs suppress T cell proliferation via different ROS‐mediated mechanisms ................................................................................................................................... 131  Figure 5.27  Myeloid cells suppress T cell proliferation via different mechanisms and to different degrees ......................................................................................................................... 132  Figure 5.28  ATRA treatment does not alter primary tumor growth but increases lung metastasis....................................................................................................................................... 134  Figure 5.29  ATRA does not change the immunosuppressive properties of MDSCs or Mφs . 136  Figure 5.30  ATRA increases the number of Mφs and reduces the number of MDSCs in 4T1 mice ......................................................................................................................................... 137  Figure 5.31  ATRA induces the differentiation of MDSCs into more immunosuppressive Mφs and promotes lung metastasis in the 4T1 and 4TO7 tumor models ............ 138  xiii  LIST OF ABBREVIATIONS Ab  antibody  Ag  antigen  AP‐1  activator protein 1  APC  antigen presenting cell  APCy  allophycocyanin  AML  acute myeloid leukemia  APL  acute promyelocytic leukemia  Arg1  arginase 1  ATRA  all‐trans retinoic acid  β‐ME  beta‐mercaptoethanol  Bcl  B cell lymphoma  BEC  [S]‐[2‐boronoethyl]‐L‐cysteine‐HCl  BM  bone marrow  BSA  bovine serum albumin  C/EBP  CCAAT/enhancer binding protein  CCL  CC chemokine ligand  CCR  CC chemokine receptor  CD  cluster of differentiation  CFSE  carboxyfluorescein succinimidyl ester  CMP  common myeloid progenitor  COX  cyclooxygenase  CpG  CpG‐dinucleotides  cpm  counts per minute  CSF‐1  colony‐stimulating factor 1  CTL  cytotoxic T lymphocyte  CTLA‐4  cytotoxic T‐lymphocyte antigen 4 (or CD152)  CXCL  CXC chemokine ligand  Cy  cyanine  xiv  DAMP  damage‐associated molecular pattern molecule  DAPI  4',6‐diamidino‐2‐phenylindole  DC  dendritic cell  dsRNA  double stranded ribonucleic acid  ECM  extracellular matrix  EDTA  ethylenediaminetetraacetic acid  ELISA  enzyme‐linked immunosorbent assay  FA  fatty acid  FCS  fetal calf serum  FITC  fluorescein isothiocyanate  Flt3L  Flt3 ligand  G‐MDSC  granulocytic‐myeloid‐derived suppressor cell  GAPDH  glyceraldehyde 3‐phosphate dehydrogenase  GM‐CFU  granulocyte macrophage colony‐forming unit  GM‐CSF  granulocyte macrophage colony‐stimulating factor  h  hour  H2O2  hydrogen peroxide  HBSS  Hank’s balanced salt solution  HEPES  4‐(2‐hydroxyethyl)‐1‐piperazineethanesulfonic acid  HFN  Hank’s balanced salt solution containing 2% FCS and 0.05% azide  HIF‐1α  hypoxia‐inducible factor‐1 alpha  HMGB1  high‐mobility group protein B1  HPC  hematopoietic progenitor cell  HPLC  high‐performance liquid chromatography  IDO  indoleamine 2,3‐dioxygenase  IFN  interferon  Ig  immunoglobulin  IκKi  inducible IκB kinase  IL  interleukin  IMC  immature myeloid cell  xv  IMDM  Iscove’s modified Dulbecco’s medium  iNOS  inducible nitric oxide synthase  IP  intraperitoneal  IRAK1  interleukin‐1 receptor‐associated kinase 1  IV  intravenous  JAK  Janus kinase  kDa  kilodalton  L‐Arg  L‐arginine  LFA‐1  lymphocyte function‐associated antigen 1  L‐NMMA  NG‐monomethyl‐L‐arginine  LAP  latency associated peptide  LPS  lipopolysaccharide  LRR  leucine‐rich repeat  Mφ  macrophage  M1 Mφ  classically activated macrophage  M2 Mφ  alternatively activated macrophage  mAb  monoclonal antibody  M‐CFU  macrophage colony‐forming unit  M‐CSF  macrophage colony‐stimulating factor  M‐CSFR  macrophage colony‐stimulating factor receptor  M‐MDSC  monocytic‐myeloid‐derived suppressor cell  MAL  MyD88 adaptor‐like  MCP‐1  monocyte chemotactic protein‐1  mDC  myeloid dendritic cell  MDSC  myeloid‐derived suppressor cell  MFI  mean fluorescence intensity  MHC  major histocompatibility complex  MHCII  major histocompatibility complex class II  min  minute  MMP  matrix metalloproteinase  xvi  MPS  mononuclear phagocyte system  MSF  migration stimulating factor  MTG  monothioglycerate  MyD88  myeloid differentiation factor 88  NAC  N‐acetyl cysteine  NF‐κB  nuclear factor kappa-light-chain-enhancer of activated B cells  NK  natural killer cell  NKT  natural killer T cell  NO  nitric oxide  NO2‐  nitrite  O2‐  superoxide  O/N  overnight  OCT  optimal cutting temperature  ONOO‐  peroxynitrite  OVA  ovalbumin  PAGE  polyacrylamide gel electrophoresis  PAMP  pathogen‐associated molecular pattern  PBS  phosphate buffered saline  PC  peritoneal cavity  PD‐L1  programmed death ligand 1  pDC  plasmacytoid dendritic cell  PE  phycoerythrin  PFA  paraformaldehyde  PG  prostaglandin  PGN  peptidoglycan  PI  propidium iodide  PMφ  peritoneal macrophage  PMN  polymorphonuclear neutrophil  Poly I:C  polyinosinic:polycytidylic  PRR  pattern recognition receptor  xvii  PTIO  carboxy‐2‐phenyl‐4,4,5,5‐tetramethylimidazoline‐1‐oxyl‐3‐oxide  RC  responder control  ROS  reactive oxygen species  rm  recombinant mouse  RNS  reactive nitrogen species  RPMI  Roswell Park Memorial Institute  SCCVII  squamous cell carcinoma VII  SDS  sodium dodecyl sulfate  SEM  standard error of the mean  SOD  superoxide dismutase  SpMφ  splenic macrophage  STAT  signal transducer and activator of transcription  TAA  tumor‐associated antigen  TAM  tumor‐associated macrophage  TBK1  TANK‐binding kinase 1  TCR  T cell receptor  TF  transcription factor  TGF‐β  transforming growth factor beta  Th  T helper cell  thy  thymidine  TIL  tumor‐infiltrating lymphocyte  TIR  Toll‐interleukin‐1 receptor  TIRAP  TIR‐domain containing adaptor protein  TLR  Toll‐like receptor  TNF  tumor necrosis factor  TRAM  TRIF‐related adaptor molecule  TRIF  TIR‐domain‐containing adapter‐inducing interferon‐β  Treg  regulatory T cell  VEGF  vascular endothelial growth factor  xviii  ACKNOWLEDGEMENTS I am extremely appreciative of the many people in my life that supported me in each step of this journey. I thank my family for their constant love, support, and encouragement. I also thank my husband for his never‐ending support, understanding, advice, and (sometimes tough) love. Thank you, Evan, for being my partner, in this and all aspects of life. I sincerely thank my supervisor, Dr. Gerald Krystal, for the opportunity to join his laboratory and conduct these studies under his guidance. Thank you, Gerry, for your support and encouragement these past six years. I appreciate your kind spirit, scientific curiosity, creativity, and passion for clinically‐relevant research. I am also grateful to Dr. Kevin Bennewith for his collaboration and mentorship. Kevin, thank you for all the time and effort you invested in these studies and your continual guidance and advice. I would also like to recognize the members of my supervisory committee, Drs. Fumio Takei, Megan Levings, and Kevin Bennewith, for their input, suggestions, and direction throughout my PhD studies. I thank all of my lab mates, past and present, especially Victor Ho, Vivian Lam, Frann Antignano, and Christina Thomas, for their help and friendship. I would also like to thank the members of the Terry Fox Laboratory for helpful discussions and collaborations, and express my gratitude to Drs. Vincent Duronio and Connie Eaves for their support of funding applications; the research described in this thesis was supported by personal fellowships awarded to me from the Canadian Institutes of Health Research, Michael Smith Foundation for Health Research, and Canadian Federation of University Women.  xix  DEDICATION I dedicate this dissertation to my family.  xx  CHAPTER 1 : INTRODUCTION 1.1 History of tumor immunology The relationship between the immune system and cancer has been the subject of much debate throughout the past century and remains one of the most challenging issues in the field of immunology. The concept that the immune system provides protection from cancer was first suggested by Paul Ehrlich in the early 1900s (Ehrlich, 1909), but it was not until after major advances in immune cell identification and function that experiments could be performed to test this hypothesis. Related to this, in the mid 1950s, Burnet and Thomas proposed the idea of cancer immunosurveillance, i.e., that the immune system can recognize and destroy emerging transformed cells (Burnet, 1957). However, initial studies produced mixed results and this idea was largely abandoned (Prehn and Main, 1957; Schreiber et al., 2011; Stutman, 1974; Stutman, 1975). Nevertheless, interest in cancer immunology was revived in 1982 when van Pel and Boon made the seminal discovery that vaccinating mice with mutagenized tumor cells provided protection from spontaneous tumors, indicating that spontaneous tumors possessed tumor antigens (Ags) (Van Pel and Boon, 1982). Advances in technology, including the development of improved mouse models, combined with a better understanding of the immune system has allowed major advances to be made in tumor immunology in the past two decades (Schreiber et al., 2011). Immune surveillance has now been clearly demonstrated, since mice with genetic immune deficiencies are more susceptible to chemically‐induced, as well as spontaneous, tumors (Laoui et al., 2011; Shankaran et al., 2001), and there is now a large body of evidence supporting the general hypothesis that the immune system can act to restrict cancer development and progression (Vesely et al., 2011). However, these advances have also revealed that the relationship between the immune system and cancer is extremely complex. In fact, we now appreciate that the immune system plays a dual role in cancer; on the one hand, it can inhibit tumor growth by recognizing and destroying cancer cells, but on the other, it can encourage tumor progression by  1  selecting for less immunogenic tumors and by providing resources to the tumor that support its growth and metastasis (Schreiber et al., 2011). 1.2 Cancer immunoediting versus tumor­induced immune suppression In 2001, experimental evidence from Schreiber and colleagues not only demonstrated the existence of immune surveillance, but also showed that the immune system can influence tumor immunogenicity (Dunn et al., 2002; Shankaran et al., 2001). Based on their findings, they proposed what they called cancer immunoediting (Dunn et al., 2002), which suggests three sequential steps: elimination, equilibrium, and escape (Schreiber et al., 2011). During the elimination phase, the immune system detects and destroys developing tumor cells (Dunn et al., 2002). Although the mechanisms by which the immune system becomes aware of these nascent tumor cells are not fully understood, there is evidence that a combination of danger signals may be involved, such as Type I interferons (IFNs), which are induced in the early stages of tumor development (Matzinger, 1994), damage‐associated molecular pattern molecules (DAMPs) (e.g. high mobility group box 1; HMGB1), which are released from dying tumor cells or injured tissues (Sims et al., 2010), or stress ligands (i.e. RAE‐1 or H60 in mice and MICA/B in humans), which are often expressed on tumor cells (Guerra et al., 2008). This elimination phase appears to require players from both the innate and adaptive immune system, and if it is fully successful, elimination is a possible endpoint of the cancer immunoediting process (Schreiber et al., 2011). However, if tumor variants survive, the equilibrium phase begins, whereby the adaptive immune system holds tumors in check by preventing the outgrowth of tumor cells, and tumor growth and immune surveillance enter a dynamic balance (Schreiber et al., 2011). Since the immune system is best able to recognize and control the most immunogenic tumor variants, it exerts a constant selective pressure on genetically unstable tumor cells and promotes the outgrowth of tumor variants that have acquired the most immunoevasive mutations (Schreiber et al., 2011). The final stage described by the immune surveillance hypothesis is escape, during which tumor cells that have evaded immune detection emerge and flourish to form clinically‐relevant tumors (Schreiber et al., 2011). 2  Although the reciprocal relationship between the immune system and tumors is widely accepted, there are some investigators who do not subscribe to the immunoediting hypothesis. Instead, these scientists propose that the most important determinants of successful tumorigenesis are the immunosuppressive mechanisms that are acquired by tumors to block anti‐tumor immune responses and induce a state of tumor‐specific tolerance. Although these two models have been proposed as conflicting hypotheses, they are likely not mutually exclusive (Whiteside, 2010). Rather, there is evidence to support both models, suggesting both the selection of less immunogenic tumors as well as tumor‐induced immune suppression contribute to carcinogenesis (Whiteside, 2010). In this thesis, the pro‐ and anti‐tumor effects of the immune system on tumorigenesis, as well as the mechanisms by which tumors suppress immune responses will be discussed. 1.3 Brief overview of the immune system The immune system is typically divided into two arms, innate and adaptive. Innate immune responses, which are carried out by monocytes, macrophages (Mφs), dendritic cells (DCs), polymorphonuclear leukocytes (PMNs, a.k.a. granulocytes), natural killer (NK) cells, platelets, mast cells, epithelial, and endothelial cells, provide rapid host protection against pathogens (Parkin and Cohen, 2001). Innate immune cells recognize foreign organisms, including nascent tumor cells, via sets of germ‐line encoded pattern recognition receptors (PRRs), such as Toll‐like receptors (TLRs) (Moresco et al., 2011). Due to the finite nature of receptor structure and expression, the innate immune response is limited in its breadth and is only activated by certain evolutionarily conserved pathogen‐associated molecular patterns (PAMPs) (Moresco et al., 2011). Receptor activation induces the maturation and activation of innate immune cells, which in turn triggers cytokine production, recruitment of immune cells, complement activation, and induction of the adaptive immune response via Ag presentation (Parkin and Cohen, 2001). Myeloid cell populations, including PMNs,  3  monocytes, Mφs, and DCs are critical mediators of immune responses (Mantovani et al., 2011; Soehnlein and Lindbom, 2010). Mφs and PMNs both eliminate pathogens and damaged cells via phagocytosis but also have distinct roles (Ueha et al., 2011); PMNs amplify inflammation by releasing cytotoxic granules, while Mφs resolve inflammation and restore tissue integrity after removal of inflammatory stimuli (Kennedy and DeLeo, 2009). The key players of the more recently evolved adaptive immune response are T and B lymphocytes. Unlike innate immune cells, T and B cells express a large repertoire of Ag receptors that are produced by site‐specific somatic recombination (Parkin and Cohen, 2001). This mechanism of receptor generation allows for a specific yet wide‐ ranging response, allowing the adaptive immune system to distinguish between vast numbers of different Ags (Parkin and Cohen, 2001). Although the adaptive response is not as fast‐acting as the innate immune system, it imparts long‐lasting specific protection to the host (Parkin and Cohen, 2001). There are two main types of effector T cells, cluster of differentiation (CD)4+ T helper (Th) cells and CD8+ cytotoxic T lymphocytes (CTLs) (Parkin and Cohen, 2001). Upon Ag recognition Th cells mature into Th1 or Th2 cells, which produce cytokines that activate players involved in the cell‐ mediated (i.e. Mφs, NK cells, CTLs) or antibody (Ab)‐mediated (i.e. B cell) response, respectively (Parkin and Cohen, 2001). CTLs, on the other hand, are directly cytotoxic to their target cells and thus have important functions in anti‐viral and anti‐tumor immunity (Parkin and Cohen, 2001). Natural killer T (NKT) cells are lymphocytes that share characteristics of both NK and T cells (Godfrey et al., 2010). NKT cells recognize glycolipid Ags in the context of CD1d, and express a semi‐invariant T cell receptor (TCR) and NK cell markers including NK1.1 (a.k.a. CD161) (Godfrey et al., 2010). NKT cells produce high levels of Th1 and Th2 cytokines and can therefore activate or suppress immune responses by modulating the functions of other immune cells (Godfrey et al., 2010). The final members of the adaptive immune system are B cells, which produce Abs upon activation. They also act as Ag‐presenting cells (APCs), presenting Abs in the context of major histocompatibility complex (MHC) to activate Th cells that help  4  amplify the response (Parkin and Cohen, 2001). Abs function via a number of different mechanisms to improve the effectiveness of both innate and adaptive‐mediated responses (Parkin and Cohen, 2001). There is overlap between the two arms of the immune system since innate APCs (i.e. Mφs and DCs) are critical modulators of the adaptive response. They reside in tissues where they continually survey the environment, taking up Ags and displaying them on their surface in the context of MHC molecules (Martinez et al., 2008). Immature APCs, which have not encountered a pathogen or harmful foreign entity, express low levels of MHC class II (MHCII) and lack expression of co‐stimulatory molecules (e.g. CD80/86, which bind to CD28 on T cells) (Martinez et al., 2008). However, in response to PRR activation or inflammatory cytokine production, Mφs and DC mature, increasing their expression of MHCII and co‐stimulatory molecules (Martinez et al., 2008). These APCs then migrate to specialized peripheral lymphoid organs, including the lymph nodes and spleen, where they interact with lymphocytes (Martinez et al., 2008). Depending on their activation state, APCs can either tolerize or activate naïve lymphocytes (Parkin and Cohen, 2001). When a naïve T cell recognizes a peptide displayed on an immature APC the T cell becomes anergic, and this is a mechanism of peripheral tolerance (Parkin and Cohen, 2001). On the other hand, binding of a naïve T cell to an Ag presented by a mature APC induces T cell activation and clonal expansion (Parkin and Cohen, 2001). 1.4 Tolerance versus autoimmunity Because of the process of somatic hypermutation by which they are generated, T and B cell receptors have the potential to recognize an almost unlimited number of Ags, including self‐Ags (Munn and Mellor, 2003). If left unchecked, these self‐reactive lymphocytes could cause tremendous damage to host tissues and lead to autoimmune diseases. Many self‐reactive T cells are deleted during development in the thymus via central tolerance (Munn and Mellor, 2003). However, since some self‐reactive cells are  5  able to mature and enter the periphery, peripheral tolerance strategies are also necessary to prevent autoimmunity (Munn and Mellor, 2003). These mechanisms include immune ignorance, such as in immune privileged sites, induction of anergy in lymphocytes that encounter their Ag in the absence of co‐stimulatory or in the presence of co‐inhibitory signals, and immune suppression by cells that actively suppress lymphocyte activation and proliferation (Nurieva et al., 2011). Although these tolerance mechanisms evolved to prevent autoimmune reactions, tumors have co‐opted them to hide from or disable the immune system. Therefore, maintaining the proper balance between immune tolerance and response is of the utmost importance; if the equilibrium shifts too strongly in favour of immune tolerance the organism forgoes its protection against harmful pathogens or tumors, yet if tolerance is diminished too much the organism is no longer protected from autoimmunity (Munn and Mellor, 2003). 1.5 TLR signaling One important factor that regulates the balance between immune tolerance and response is the ability of the immune system to distinguish self from non‐self. Innate immune cells possess PRRs that directly recognize pathogens and initiate an inflammatory response (Parkin and Cohen, 2001). These receptors are located either on the cell surface or within internal cell compartments of Mφs, PMNs, DCs, epithelial, and endothelial cells (Fitzner et al., 2008; Parkin and Cohen, 2001). These innate immune receptors are not clonally distributed like adaptive immune receptors; all cells of a particular subtype express the same set of PRRs (Parkin and Cohen, 2001). The TLR family is a class of PRRs that plays a key role in the innate immune response. To date, thirteen TLRs, named TLR1‐13, have been identified; 10 in humans and 12 in mice (Moresco et al., 2011). All family members share a common intracellular Toll‐ interleukin‐1 receptor (TIR) domain and have multiple extracellular leucine‐rich repeats (LRRs) (Moresco et al., 2011). TLRs, named for their homology to the protein coded by the Toll gene in Drosophila, are single membrane‐spanning non‐catalytic receptors that function as homo‐ or hetero‐dimers (Moresco et al., 2011). Each receptor  6  or receptor dimer is able to recognize a distinct set of PAMPs, which are molecular patterns that are broadly shared by a class of pathogens but are also distinguishable from normal host molecules (Moresco et al., 2011). Since TLRs are germ‐line encoded and their specificity cannot be easily altered by evolution, these receptors recognize molecules that are consistently and specifically associated with threats, including pathogens or cellular stress (Moresco et al., 2011). Stimulation of TLRs activates signaling cascades that lead to the induction of genes essential for immune responses, including those that encode chemokines, cytokines, and co‐stimulatory molecules (Akira and Takeda, 2004). Since different classes of microorganisms possess different components, and are thus recognized by different TLRs, the innate immune system is able to not only identify a pathogen as foreign, but also initiate an appropriately targeted response. For example, extracellular bacterial PAMPs induce increased phagocytosis, Ag processing and presentation, and secretion of pro‐inflammatory cytokines, which stimulate a Th2‐skewed adaptive immune response (Akira and Takeda, 2004). Viral PAMPs, on the other hand, trigger decreased protein synthesis, induction of apoptosis, and type 1 interferon (IFN) production (Akira and Takeda, 2004). After ligand binding, TLRs dimerize, undergo conformational changes, and recruit different TIR‐domain‐containing adaptor proteins including myeloid differentiation factor 88 (MyD88), TIR‐domain containing adaptor protein (TIRAP, a.k.a. MyD88 adaptor‐like, or MAL), TIR‐domain‐containing adaptor‐ inducing interferon‐β (TRIF), and TRIF‐related adaptor molecule (TRAM) (Moresco et al., 2011). These adaptors, in turn, activate other molecules within the cell, including specific protein kinases (i.e. interleukin‐1 receptor‐associated kinase 1 (IRAK1), TANK‐ binding kinase 1 (TBK1), and inducible IκB kinase (IKKi)) that amplify the signal and lead to the induction or inhibition of genes that regulate the inflammatory response (Moresco et al., 2011; Waltenbaugh et al., 2008). In general, TLR signaling pathways can be classified as MyD88‐dependent or MyD88‐independent, each of which activates distinct signaling pathways and ultimate end‐points (Akira and Takeda, 2004). Signaling through MyD88 activates nuclear factor kappa‐light‐chain‐enhancer of  7  activated B cells (NF‐κB) and activator protein 1 (AP‐1) transcription factors (TFs), resulting in the induction of pro‐inflammatory genes (Akira and Takeda, 2004). On the other hand, in addition to activating NF‐κB, signaling through the MyD88‐independent pathway also triggers the production of Type I IFNs (IFN‐α and IFN‐β), which have potent anti‐viral effects via their subsequent activation of IFN‐inducible genes (Akira and Takeda, 2004; Schroder et al., 2004). A summary of known mammalian TLRs, their PAMP ligands, signaling molecules, and expression patterns can be found in Table 1.1. Table 1.1  Summary of the properties of known TLRs.  TLRs recognize, bind, and become activated by different ligands, which are derived from different microbes including viruses, fungi, bacteria, and protozoa or from damaged host cells. TLRs signal through different adaptor proteins that directly bind to activated TLRs and recruit different downstream signaling components. While most receptors are displayed on the cell surface, TLRs that recognize nucleic acids (TLR3, 7, 8, 9) are located in intracellular endosomes. TLR10 is only present in humans and TLRs 11‐13 are only present in mice. Adapted from Moresco et al., 2011; Fitzner et al., 2008; and Waltenbaugh et al., 2008. Receptor  Ligand(s)  Adaptor(s)  Ligand Location Bacteria  MyD88/MAL  TLR1  Triacyl lipopeptides  TLR2  Glycolipids, lipopeptides, lipoproteins, lipoteichoic acid  Bacteria  MyD88/MAL  HSP70 Zymosan Double‐ stranded RNA (dsRNA), poly I:C  Host cells Fungi Viruses  TRIF  TLR3  Cell Types  Location  Monocytes, Mφs, DC subsets, B cells, endothelial cells Monocytes, Mφs, myeloid (m)DCs , mast cells, endothelial cells  Cell surface  DCs, B cells, endothelial cells  Cell compartment  Cell surface  8  Receptor TLR4  TLR5  TLR6  Ligand(s) Lipopolysacch‐ aride (LPS)  Heat shock proteins Fibrinogen, heparin sulfate fragments, hyaluronic acid fragments Flagellin Diacyl lipopeptides  Ligand Location Gram‐ negative bacteria  Bacteria and host cells Host Cells  Bacteria Mycoplasma  Adaptor(s)  Cell Types  Location  MyD88/MAL/ TRIF/TRAM  Monocytes, Mφs, mDCs, mast cells, intestinal epithelium, endothelial cells  Cell surface  MyD88  Monocytes, Mφs, DC subsets, intestinal epithelium, endothelial cells Monocytes, Mφs, mast cells, B cells, endothelial cells Monocytes, Mφs, plasma‐ cytoid (p)DCs, B cells, endothelial cells Monocytes, Mφs, DC subsets, B cells, endothelial cells Monocytes, Mφs, pDCs, B cells, endothelial cells  Cell surface  MyD88/MAL  TLR7  Single‐ stranded RNA and synthetic analogs (imidazoguino‐ line, loxoribine, bropirimine)  Synthetic compounds  MyD88  TLR8  Single‐ stranded RNA, small synthetic compounds  Synthetic compounds  MyD88  TLR9  Unmethylated CpG DNA  Bacteria  MyD88  Cell surface  Cell compartment  Cell compartment  Cell compartment  9  Receptor  Ligand(s)  Ligand Location Unknown  Adaptor(s)  TLR10  Unknown  Unknown  TLR11  Profilin  Toxoplasma gondii  MyD88  TLR12 TLR13  Unknown Unknown  Unknown Virus  Unknown MyD88, TAK‐1  Cell Types  Location  Monocytes, Mφs, B cells, endothelial cells Monocytes, Mφs, liver cells, kidney cells, bladder epithelium Neurons Unknown  Cell surface  Cell compartment  Unknown Unknown  10  1.6 Immune escape The fact that tumors are frequently able to develop and thrive in immune competent hosts suggests that the immune system is often unable to prevent cancer. There are a number of factors that contribute to this failure of anti‐tumor immunity including low immunogenicity of tumor cells, tolerance, and immune evasion (Dunn et al., 2002). In order to be successful, tumors must evolve mechanisms to escape from or suppress host immune responses, and this immune evasion is increasingly being recognized as an important component of cancer development (Hanahan and Weinberg, 2011). Immune evasion is mediated by several distinct molecular mechanisms that can be divided into two general categories; those that allow tumors to hide from the immune system and those that enable tumors to disable the immune system. Transformed cells can hide from the immune system by down‐regulating Ag presentation (either by decreasing Ag or co‐stimulatory molecule expression), by antigenic drift, or by acquiring defects in Ag processing or presentation machinery (Whiteside, 2006). Another way tumor cells hide is by decreasing T cell recruitment into tumor sites via reducing chemokine and adhesion molecule expression (Whiteside, 2006). Tumor cells can also surmount anti‐tumor immunity by increasing their resistance to immune effector mechanisms. For example, tumor cells can acquire mutations that make them less sensitive to apoptosis, including over‐expression of anti‐ apoptotic molecules (e.g. B cell lymphoma (Bcl)‐2, Bcl‐XL, survivin), up‐regulation of molecules that interfere with the perforin/granzyme pathway (e.g. serine protease inhibitors), expression of soluble death receptors that act as decoys, and shedding surface molecules that activate NK cells (Whiteside, 2006). Alternatively, tumor cells can escape from anti‐tumor immunity by disabling the immune system. One of the key mechanisms tumors use to do this is production of immunosuppressive factors that either disable the immune system directly or that promote recruitment and differentiation of immunosuppressive cells (Whiteside, 11  2006). Tumors secrete a plethora of molecules that induce an immunosuppressive state either locally, i.e., within the tumor microenvironment, or systemically. These can include immunosuppressive small molecules (i.e. prostaglandin (PG)E2, histamine, reactive oxygen species (ROS) and reactive nitrogen species (RNS)), cytokines (i.e. vascular endothelial growth factor (VEGF), granulocyte macrophage colony‐stimulating factor (GM‐CSF), interleukin (IL)‐4, IL‐10, transforming growth factor beta (TGF‐β)), and enzymes (i.e. indoleamine 2,3 dioxygenase (IDO), arginase 1(Arg1)) (Whiteside, 2010). Furthermore, tumors secrete growth factors and pro‐inflammatory cytokines that promote the recruitment and activation of regulatory cells (i.e. regulatory T cells (Tregs), tolerogenic DCs, mast cells, tumor‐associated macrophages (TAMs), and myeloid‐derived suppressor cells (MDSCs)) (Whiteside, 2006), whose functions will be discussed in detail in Section 1.7. 1.7 Pro­ and anti­tumor effects of the immune system It is well established that the immune system plays a role in both the promotion and inhibition of tumor growth and metastasis (Schreiber et al., 2011) (Fig. 1.1). In terms of anti‐tumor immunity, members from both the innate and adaptive arms of the immune system, including CD8+ CTLs, CD4+ Th cells, NK cells, DCs, Mφs, and B cells, work together to identify and kill foreign tumor cells using the same mechanisms used to eradicate microorganisms (Bhardwaj, 2007). As mentioned earlier, while there was initially some doubt about whether the immune system could successfully distinguish transformed cells from normal cells, there is now convincing evidence that cancer cells express specific Ags, termed tumor‐associated Ags (TAAs) that distinguish them from their non‐transformed counterparts (Whiteside, 2010). These Ags are distinctive due to a combination of mutagenic and epigenetic alterations (Laoui et al., 2011) and include products of viral, aberrantly expressed, or mutated genes (Whiteside, 2010). Tumor‐ specific T cells, i.e., T cells with TCRs that recognize TAAs, are theoretically able to eradicate even metastatic tumors (Vesely et al., 2011). CTLs are critical mediators of tumor cell killing and exert their effects via a number of different mechanisms,  12  Figure 1.1 Immune cells contribute to both the promotion and inhibition of tumor growth. Both innate and adaptive immune cells mediate anti‐tumor immunity, including CD8+ CTLs, CD4+ Th cells, NK cells, M1 Mφs, DCs, and B cells. Together these cell types can detect and induce apoptosis in tumor cells, which results in decreased tumor growth and metastasis. However, successful tumors develop strategies to evade immune detection, including the production of cytokines, growth factors, and chemokines that induce the development of immunosuppressive cells, including Tregs, mast cells, MDSCs, tolerogenic DCs, and M2 Mφs. These suppressive immune cells actively inhibit anti‐tumor immunity and produce factors that promote angiogenesis, as well as tumor cell growth and invasion, and thereby contribute to the progression of both primary and metastatic tumors. Adapted from references in section 1.7. 13  including indirect killing through release of IFN‐γ, tumor necrosis factor (TNF)‐α and soluble Fas ligand, and direct killing by engagement of membrane‐bound Fas ligand to Fas on tumor cells or by secretion of perforin and granzymes into the target cell (Vesely et al., 2011). In addition, the immune system is able to indirectly protect the host from tumors by both eliminating pathogens and limiting inflammation. Specifically, since the immune system evolved in part to protect the host from viral infection it is quite adept at preventing virus‐induced tumors (Schreiber et al., 2011). Similarly, by efficiently resolving infections, the immune system reduces the duration and extent of inflammation (Schreiber et al., 2011). Inflammation is a multifaceted process that acts to protect the host and maintain tissue homeostasis. Inflammation is initiated in response to stresses such as infection, physical injury, or local immune responses and functions to eliminate the threat to the organism and then trigger the healing process (Ostrand‐Rosenberg and Sinha, 2009). It involves the local accumulation of fluid, plasma proteins, and immune cells and is generally classified as either acute or chronic; acute inflammation is the initial response of the organism to a threat and is normally a short‐lived process, while chronic inflammation occurs when inflammation is prolonged (e.g. autoimmune disease, persistent foreign body, or cancer) (Sansone and Bromberg, 2011). Whereas acute inflammation is a necessary aspect of the anti‐tumor response, chronic inflammation has been shown to contribute to all stages of tumorigenesis, including cancer development, progression, and metastasis (Sansone and Bromberg, 2011). In fact, Mantovani and colleagues have recently suggested that inflammation should be considered the seventh hallmark of cancer (Mantovani, 2009; Mantovani and Sica, 2010) and in their recent review Hanahan and Weinberg highlighted tumor‐promoting inflammation as an enabling characteristic of cancer (Hanahan and Weinberg, 2011). Inflammation and cancer are linked in two major ways; on one hand, oncogene activation drives the expression of inflammatory mediators and leads to a pro‐inflammatory microenvironment and, on the other hand, inflammatory conditions promote cancer development (Mantovani and Sica, 2010). Therefore,  14  prompt resolution of inflammation by immune cells is critical to reducing cancer incidence and development. Although these protective mechanisms are effective in some cases, the fact that tumors are often able to develop and thrive in immune competent hosts indicates that the anti‐tumor functions of the immune system are frequently inadequate to protect the host from developing cancer. Related to this, there are numerous lines of evidence suggesting both early and late involvement of the immune system in promoting tumorigenesis (Whiteside, 2010). Early tumors and even premalignant foci are frequently infiltrated with hematopoietic cells, including lymphocytes, Mφs, and occasionally PMNs (Kornstein et al., 1983; von Kleist et al., 1987). Although the number of tumor‐infiltrating lymphocytes (TILs), especially CD8+ CTLs, has been associated with improved patient survival in some studies, they are largely unsuccessful in inhibiting tumor growth (Whiteside, 2010). Moreover, TILs isolated from advanced or metastatic tumors are often more functionally impaired than those isolated from early lesions (Whiteside, 2010). This is in large part due to bone marrow (BM) derived immunosuppressive cells that develop and are recruited to the tumor site by tumor‐ produced factors (Whiteside, 2008) including Tregs (Bettini and Vignali, 2009), alternatively activated (M2) Mφs (Gordon and Martinez, 2010), tolerogenic DCs (Whiteside, 2010), mast cells (Maltby et al., 2009), and MDSCs (Ostrand‐Rosenberg and Sinha, 2009; Youn and Gabrilovich, 2010). These immunosuppressive cells, which function physiologically to prevent auto‐immunity and resolve inflammation, directly oppose the anti‐tumor functions of the immune system by actively suppressing cytotoxic T cells and are critical mediators of tumor development and progression via production of factors that foster tumor growth and metastasis (Hanahan and Weinberg, 2011; Whiteside, 2006). Mφs and MDSCs are two of the most integral pro‐tumor immune cells and are the focus of this thesis. The roles MDSCs and Mφs play in suppressing anti‐tumor immunity and fostering tumorigenesis will now be discussed in greater detail.  15  1.7.1  Mφs  Mφs are terminally differentiated cells of the mononuclear phagocyte system (MPS), which is comprised of mononuclear cells that possess phagocytic abilities, including lineage committed BM precursors, MDSCs, monocytes, DCs, and Mφs (Pollard, 2009). Terminally‐differentiated cells of the MPS arise as part of normal myelopoiesis from pluripotent hematopoietic progenitor cells (HPCs) that give rise to a series of lineage‐restricted progenitor cells, or immature myeloid cells (IMCs) (Fig. 1.2). These IMCs, which lack immunosuppressive properties in healthy individuals (Delano et al., 2007), generate pro‐monocytes and monocytes that circulate in the bloodstream (Fig. 1.2) (Pollard, 2009). Murine monocytes express both CD11b and CD115 and can be further classified into functional subpopulations of resident (Gr1‐Ly6C‐) and inflammatory (Gr1+Ly6C+) monocytes (Geissmann et al., 2008; Geissmann et al., 2010). In response to inflammatory cytokines, chemokines, and other factors, monocytes migrate into tissues and undergo terminal differentiation into resident DCs or Mφs (Fig. 1.2) (Siveen and Kuttan, 2009). Recruitment of monocytes to these inflammatory sites is mediated to a large extent by CC chemokine ligand 2 (CCL2, a.k.a. monocyte chemotactic protein‐1, MCP‐1), which binds to CC chemokine receptor 2 (CCR2) on monocytes (Laoui et al., 2011). Overall, cells of the MPS are critical for a wide variety of physiological processes, including tissue development, homeostasis, and regulating the balance between pro‐ and anti‐inflammatory responses, and accordingly, exhibit enormous heterogeneity and versatility (Qian and Pollard, 2010).  16  Figure 1.2 Development and functions of the mononuclear phagocytic lineage. Mononuclear phagocyte development begins in the BM where pluripotent HPCs give rise to a series of progenitor cells, i.e., common myeloid progenitors (CMPs), granulocyte/macrophage colony‐forming units (GM‐CFUs), macrophage colony‐forming units (M‐CFUs), monoblasts, and pro‐monocytes. Terminally differentiated osteoclasts and Mφs arise in the BM, while all other mature cell types arise in the tissues from blood monocytes. M‐CSF plays an important role in promoting the development of mononuclear phagocytes at multiple stages. The presence of other factors in the tissues, including GM‐CSF, IL‐4, IL‐13, IFN‐γ, TNF‐α, promote the differentiation and determine the specific phenotype and physiological roles of mature DCs and Mφs. *’s indicate IMCs, which lack immunosuppressive properties in healthy individuals. However, in response to a variety of inflammatory factors, or disease states, these cells can acquire immunosuppressive functions and are termed MDSCs. Adapted from Pollard, 2009 and other references in section 1.7.1.  17  Mature Mφs are found in almost every tissue where they perform a wide variety of critical functions (Table 1.2). Under steady‐state conditions, terminally differentiated Mφs are very stable and can exist in tissues for months, or even years (Cuenca et al., 2011). Mφs are defined by specific phenotypic characteristics and by the expression of certain cell surface markers, none of which are specifically limited to Mφs (Gordon and Taylor, 2005). With the exception of alveolar Mφs, which express CD11c but not CD11b due to their unique lung environment (Guth et al., 2009), murine Mφs express CD11b, F4/80, and macrophage colony‐stimulating factor receptor (M‐CSFR; CD115) and lack expression of Gr1 (Qian and Pollard, 2010). In combination, these characteristics distinguish Mφs from other members of the myeloid lineage such as PMNs, monocytes, and MDSCs (Joyce and Pollard, 2009). The development of Mφs is regulated by a number of different growth factors, the most important of which is macrophage colony‐ stimulating factor (M‐CSF, a.k.a. colony‐stimulating factor‐1 or CSF‐1), as it is required at multiple stages of the differentiation process (Fig. 1.2) (Pollard, 2009). In fact, M‐CSF is not only essential for Mφ development, it is also necessary for their survival and proliferation in vitro (Chitu and Stanley, 2006). Mφs exhibit a variety of phenotypes and perform many diverse functions including phagocytosis of apoptotic cells, debris removal, destruction of pathogens, coordination of wound healing, tissue development and remodeling, Ag presentation, regulation of immune responses, and induction of both innate and adaptive immunity (Fig. 1.2) (Cesta, 2006; Gordon and Martinez, 2010; Martinez et al., 2008; Munn and Mellor, 2003; Pollard, 2009). These roles of Mφs in normal tissues can be exploited by the developing tumor and, as a result, Mφs can contribute to all phases of the cancer process (Laoui et al., 2011).  18  Table 1.2  Diversity of Mφs in different tissues.  Mφs are found in almost every tissue in mammals and perform a wide variety of immunogenic and non‐immunogenic roles. Adapted from Pollard, 2009; Guth et al., 2009; and Cesta, 2006. Tissue Bone  Specific Name Osteoclast  Function Bone remodelling; provide stem cell niches  Bone Marrow Mφ  Erythropoiesis (identify and degrade ejected nuclei) Neuronal survival and connectivity; repair Immune surveillance Vascular remodelling Immune surveillance Ductal development Clearance of debris from blood; tissue regeneration Regulation of immune responses to inhaled pathogens and allergens Branching morphogenesis; ductal development Steroid hormone production; ovulation Islet development Immune surveillance Clearance of debris and damaged erythrocytes Steroid hormone production Cervical ripening  Brain Epidermis Eye Intestine Kidney Liver  Microglial cell Langerhans cell N/A Crypt Mφ N/A Kupffer cell  Lung  Alveolar Mφ  Mammary gland Ovary Pancreas Peritoneal Cavity Spleen Testis Uterus  N/A N/A N/A Peritoneal Mφ Splenic Mφ N/A Uterine Mφ  N/A, not applicable 1.7.1.1  Mφ activation  One of the hallmark characteristics of Mφs is their functional plasticity (Mantovani and Sica, 2010). Mφs exhibit a wide range of phenotypes and perform many different functions depending on the environmental stimuli to which they are exposed (Martinez et al., 2008) (Fig. 1.3). This diversity of Mφs has led to a number of proposed classification systems based on activating stimuli (Gordon, 2003) or different homeostatic activities (i.e. host defense, wound healing, or immune regulation) (Mosser and Edwards, 2008). The M1/M2 classification scheme described below is the most widely used model and has been elaborated upon to include a number of different subclasses (i.e. M2a, M2b, M2c) (Mantovani et al., 2004). However, it is important to bear in mind that Mφ activation is likely a cyclic process that balances classical and 19  alternative activation to achieve proper immunologic function (McGrath and Kodelja, 1999). Thus, polarization of Mφs is a useful, simplified conceptual framework for describing what is most likely a continuum of diverse functional states (Mantovani et al., 2004). Classically activated, or M1 Mφs, are induced in response to type I, or pro‐ inflammatory cytokines, such as IFN‐γ and TNF‐α (Martinez et al., 2006). In turn, M1 Mφs produce high levels of pro‐inflammatory cytokines (i.e. IL‐1β, IL‐6, IL‐12, IL‐23, TNF‐α) and low levels of anti‐inflammatory cytokines (i.e. IL‐10) (Martinez et al., 2006). Upon activation by Th1 or NK cell produced IFN‐γ, M1 Mφs secrete ROS and RNS and thereby become potent killers of intracellular microorganisms and tumor cells (Parkin and Cohen, 2001). These cytotoxic functions, combined with the ability of these Mφs to process and present Ags, makes them important inducers and effectors of Th1 cell‐ mediated immune responses (Gordon and Martinez, 2010; Munn and Mellor, 2003). Alternatively activated, or M2 Mφs, on the other hand, have been reported to suppress both specific and non‐specific T cell activation. In general, they arise in response to stimulation by type II cytokines, such as IL‐4, IL‐10, IL‐13, TGF‐β, and PGE2 and produce anti‐inflammatory cytokines (i.e. IL‐10) (Martinez et al., 2008). Interestingly, M2 activation appears to be regulated in part by the p50 subunit of NF‐ κB, which inhibits NF‐κB induced M1‐polarizing IFN‐β production and consequently promotes M2 activation (Porta et al., 2009). M2 Mφs play important roles in killing extracellular microorganisms and parasites and in promoting wound healing by initiating tissue remodeling and angiogenesis (Martinez et al., 2008). A number of murine M2 Mφ markers have been described including scavenger, mannose, and galactose‐type receptors (Gordon, 2003; Stein et al., 1992), FIZZ1 and the mammalian chitinase Ym1 (Ho and Sly, 2009; Holcomb et al., 2000; Raes et al., 2002), and recently, migration‐stimulating factor (MSF) (Solinas et al., 2010). M2 Mφs have been reported to suppress T cell activation/proliferation via a variety of mechanisms including secretion  20  Figure 1.3 Polarization of Mφ phenotypes. Mφ activation is a cyclical process that balances classical (M1) and alternative (M2) activation to eradicate invading pathogens and promote subsequent healing. Blood monocytes that differentiate in the presence of LPS/IFN‐γ mature into M1 or “classically activated” Mφs. These M1 Mφs are effective APCs and thus promote immune responses. In addition, they produce high levels of pro‐inflammatory cytokines, ROS, and RNS, which enable them to kill pathogens and transformed cells. On the other hand, monocytes that differentiate in the presence of IL‐4, IL‐10, IL‐13, or corticosteroids mature into M2 or “alternatively activated” Mφs, which secrete IL‐10, TGF‐β, IL‐1ra, and the IL‐1R decoy protein. M2 Mφs scavenge dead cells and debris, remodel and repair damaged tissues, and promote angiogenesis. They also function to suppress M1 Mφ‐mediated immune responses as part of their role in resolving inflammation. Another key difference between M1 and M2 Mφs is how they metabolize the amino acid L‐Arg; M1 Mφs express iNOS and convert L‐Arg into NO while M2 Mφs express the enzyme Arg1, which converts L‐Arg into L‐ornithine, which is then converted into polyamines (i.e. putrescine, spermidine, spermine) that promote cell division and L‐proline, which supports collagen synthesis. Adapted from references in section 1.7.1.1.  21  of immunosuppressive cytokines (i.e. TGF‐β, IL‐10) (Munn and Mellor, 2003), Treg induction (Brem‐Exner et al., 2008; Hoves et al., 2006), production of PGs, ROS, and RNS (Albina et al., 1991; Allison, 1978; Metzger et al., 1980; Strickland et al., 1996), and by depletion of amino acids such as tryptophan (Munn et al., 1999) and arginine (Kung et al., 1977; Rodriguez et al., 2003) via expression of IDO and Arg1, respectively. Based on in depth studies exploring Mφ polarization, Mantovani et al. have proposed that M2 Mφs can be further subdivided based on the factors that induce their activation (Mantovani et al., 2004). Specifically, M2a are activated by IL‐4 or IL‐13, M2b by immune complexes in combination with IL‐1β or lipopolysaccharide (LPS), and M2c by IL‐10, TGF‐β, or glucocortocoids (Mantovani et al., 2004; Martinez et al., 2008). The phenotype and function of each subtype, including chemokine, cytokine, effector molecule, and receptor expression, has been thoroughly elucidated by genetic profiling, as well as functional assays (Mantovani et al., 2004; Martinez et al., 2006). Among their many intriguing findings, Martinez et al. report that M‐CSF driven human monocyte to Mφ differentiation leads to expression of M2 transcripts, suggesting that M2 activation is the default pathway under steady‐state conditions (Martinez et al., 2006). Furthermore, they found that treatment with M‐CSF upregulated a number of cell cycle‐ associated genes, revealing the previously unappreciated proliferation potential of monocytes (Martinez et al., 2006). These important findings regarding the phenotypic profiles of M1 versus M2 Mφs have been recently and comprehensively reviewed (Gordon and Martinez, 2010). Contributing to the distinct responses of M1 and M2 Mφs is the way they metabolize the amino acid L‐arginine (L‐Arg). M1 Mφs rapidly upregulate inducible nitric oxide synthase (iNOS, or NOS2) in response to danger signals (Rauh et al., 2005; Sinha et al., 2005a) to convert L‐Arg into nitric oxide (NO) to kill bacteria, viruses, and tumor cells (Gordon, 2003; Mantovani et al., 2002). M2 Mφs, on the other hand, which typically function in wound healing after an infectious agent has been destroyed, possess constitutively high levels of Arg1, which sequesters L‐Arg away from iNOS to  22  generate ornithine. Ornithine, in turn, stimulates host cell proliferation (via ornithine decarboxylase‐mediated conversion into polyamines) and collagen synthesis (via ornithine‐derived proline) (Gordon and Martinez, 2010). 1.7.1.2  TAMs  It is well established that Mφs play a critical role in tumor biology. Solid tumors, including both primary tumors and metastatic lesions, are infiltrated by significant numbers of TAMs (Siveen and Kuttan, 2009). In fact, there have been reports that TAMs can comprise up to 80% of the cell mass in breast tumors (Bingle et al., 2002). The presence of extensive TAM infiltration correlates with poor prognosis and metastasis in a variety of human cancers, including breast, cervix, and bladder (Leek et al., 1996; Siveen and Kuttan, 2009), a finding that underscores the vital role TAMs play in promoting carcinogenesis. Monocytes are recruited to the tumor site by tumor‐derived chemokines such as CCL2 and they mature and become activated in response to tumor‐ derived cytokines, growth factors (i.e. M‐CSF, VEGF, IL‐4, IL‐10, TGF‐β), and PGs (Siveen and Kuttan, 2009). Once at the tumor site, Mφs tend to cluster at the leading edge of tumors, i.e., at the tumor‐host tissue interface (Siveen and Kuttan, 2009). In general, TAMs are thought to display a molecular and functional phenotype similar to M2 Mφs (Mantovani et al., 2002). However, there are reports that some TAMs can express both iNOS and Arg1 and thus possess a combination of M1‐ and M2‐like features (Gordon and Martinez, 2010). Interestingly, although these M1/M2 TAMs produce pro‐ inflammatory cytokines and ROS, they do not seem to be cytotoxic to tumor cells (Gordon and Martinez, 2010). A recent study by Movahedi et al. highlighted the heterogeneity of TAMs, demonstrating the presence of multiple phenotypically and functionally distinct TAM subsets, which could be characterized as either M1‐ or M2‐ like (Movahedi et al., 2010). Consistent with the role Mφs play in normal wound healing, TAMs promote tumor growth and metastasis via promotion of angiogenesis and vascularization,  23  stroma formation and remodeling, enhancement of tumor cell growth and invasion, recruitment of additional inflammatory cells, and suppression of anti‐tumor immune responses (Mantovani et al., 2002; Siveen and Kuttan, 2009) (Fig. 1.4). In fact, it has been suggested that TAMs play a critical role in facilitating tumor cell migration out of the primary tumor, entry into blood or lymph vessels, and seeding in distant sites (Qian et al., 2009). TAMs carry out these functions by producing a plethora of growth factors, proteolytic enzymes, cytokines, and inflammatory mediators (Bingle et al., 2002; Siveen and Kuttan, 2009) and the specific mechanisms involved have been detailed in several reviews (Bingle et al., 2002; Condeelis and Pollard, 2006; Laoui et al., 2011; Martinez et al., 2008; Qian and Pollard, 2010).  24  Figure 1.4 TAMs promote tumorigenesis via multiple mechanisms. TAMs resemble M2 Mφs and play a critical role in tumor growth and metastasis. TAMs produce a plethora of cytokines, growth factors, chemokines, and other molecules that support tumor growth, metastasis, matrix remodeling, inflammation, angiogenesis, and immune suppression. Adapted from references in section 1.7.1.2.  25  1.7.2  Myeloid­derived suppressor cells  MDSCs are a heterogeneous population of relatively immature cells at different stages of differentiation including early myeloid progenitors and immature Mφs, monocytes, PMNs, and DCs (Pastula and Marcinkiewicz, 2011). Murine MDSCs express both the αM integrin CD11b and the myeloid lineage differentiation Ag Gr1 (Ly6G and Ly6C). Two subsets of murine MDSCs have been identified; CD11b+Ly6G‐Ly6Chigh mononuclear  cells,  which  are  termed  monocytic  MDSCs  (M‐MDSCs)  and  CD11b+Ly6G+Ly6Clow cells, which have multi‐lobed nuclei and are termed granulocytic‐ MDSCs (G‐MDSCs) (Movahedi et al., 2008; Youn et al., 2008). Human MDSCs are generally defined as cells that express CD11b and the common myeloid marker CD33, but lack expression of mature myeloid and lymphoid markers and the MHCII molecule HLA‐DR (Almand et al., 2001; Diaz‐Montero et al., 2009; Nagaraj and Gabrilovich, 2010; Zea et al., 2005). MDSCs are greatly expanded in practically all tested mouse tumor models (Gallina et al., 2006; Habibi et al., 2009; Rodriguez et al., 2005; Youn et al., 2008) and in cancer patients (Almand et al., 2001; Diaz‐Montero et al., 2009; Kusmartsev et al., 2008; Mandruzzato et al., 2009; Ochoa et al., 2007; Valenti et al., 2006; Young and Lathers, 1999; Zea et al., 2005) and are considered one of the main contributors to immune suppression in cancer (Nagaraj and Gabrilovich, 2010). In response to tumor‐ derived factors that promote myeloid cell recruitment and activation (i.e. M‐CSF, GM‐ CSF, VEGF, IL‐1β, IL‐3, IL‐6, IL‐10, Flt3 ligand (Flt3L)) MDSCs are recruited from the BM to the tumor site and secondary lymphoid organs (i.e. spleen, lymph nodes) via the bloodstream (Youn and Gabrilovich, 2010). In addition there is evidence that MDSCs can also home to and accumulate in the liver (Ilkovitch and Lopez, 2009), as well as other sites including the kidneys, bones, and lungs (Pulaski and Ostrand‐Rosenberg, 2001). A number of different combinations of markers have been suggested to more specifically identify murine and human MDSCs, but the heterogeneity of MDSCs makes this difficult (Gallina et al., 2006; Greenwald et al., 2005; Huang et al., 2006; Kryczek et al., 2006; Nagaraj and Gabrilovich, 2010). Moreover, although immature myeloid cells from tumor‐free mice lack the immunosuppressive activity of MDSCs they express  26  similar levels of these proposed markers (Youn et al., 2008). Consequently, both murine and human MDSCs must be identified functionally, which requires assays to accurately assess MDSC immunosuppression ex vivo. MDSCs are distinguished by their myeloid origin, immature state, and, most notably, by their ability to suppress T cell‐mediated immune responses (Youn and Gabrilovich, 2010). They have been reported to exert their suppressive effects by a number of different mechanisms, which are indicative of their heterogeneity. MDSCs can express both iNOS (induced by IFN‐γ) and Arg1 (induced by IL‐4, IL‐13, and PGE2), which contribute to MDSC‐mediated suppression both individually and synergistically (Bronte et al., 2003; Laoui et al., 2011). As mentioned previously, iNOS converts L‐Arg into NO, which inhibits IL‐2R signaling and T cell activation while Arg1 expression depletes L‐Arg from the microenvironment, which prevents re‐expression of the TCRζ chain and T cell signaling (Bronte et al., 2003). However, when both iNOS and Arg1 are expressed in the same cell these two enzymes work together in a synergistic fashion to produce ROS. In the absence of L‐Arg (due to Arg1 expression), iNOS can catalyze superoxide (O2‐) formation via its reductase domain, which combines with water to generate hydrogen peroxide (H2O2) or with NO to form peroxynitrite (ONOO‐), a powerful oxidant that potently induces T cell dysfunction and apoptosis (Bronte et al., 2003; Xia et al., 1998). Consequently, MDSCs have been reported to produce high levels of ROS such as H2O2 and ONOO‐ (Youn and Gabrilovich, 2010). MDSCs have also been reported to suppress T cells via the production of immunosuppressive factors including TGF‐β, IL‐10, and PGE2 (Ostrand‐Rosenberg and Sinha, 2009), by sequestering the amino acid L‐cysteine, which is required for T cell activation (Srivastava et al., 2010), by down‐regulating T cell expression of L‐selectin, which prevents T cell trafficking (Hanson et al., 2009), and by inducing the development of functional Tregs (Huang et al., 2006; Serafini et al., 2008). There are reports that M‐MDSCs and G‐MDSCs suppress T cell responses via different mechanisms (Movahedi et al., 2008; Youn and Gabrilovich, 2010). G‐MDSCs  27  suppress T cells via ROS production, which requires contact between G‐MDSCs and T cells for effective suppression. (Movahedi et al., 2008; Youn and Gabrilovich, 2010). M‐ MDSCs, on the other hand, have been reported to suppress T cells via a number of contact independent mechanisms including upregulation of iNOS and consequent NO production, upregulation of Arg1 and subsequent L‐Arg depletion, and production of immunosuppressive cytokines (Youn and Gabrilovich, 2010). In addition, some groups have reported that M‐MDSCs are more potent than G‐MDSCs on a per cell basis; however, in most tumor models the vast majority of MDSCs are comprised of the G‐ MDSC subtype (Youn and Gabrilovich, 2010). In addition to suppressing T cell responses, MDSCs can also impact anti‐tumor immunity via interactions with Mφs, DCs, B cells, NK cells, and NKT cells (Ostrand‐ Rosenberg, 2010). Briefly, MDSCs can inhibit Mφ production of IL‐12 and polarize Mφs towards an M2, tumor‐promoting phenotype (Sinha et al., 2005b). In addition, MDSCs increase Mφ expression of programmed death ligand 1 (PD‐L1), a negative T cell co‐ stimulatory molecule (Ilkovitch and Lopez, 2009). Moreover, studies have suggested that factors that induce MDSC activation (i.e. LPS and IFN‐γ stimulation) inhibit the development of DCs (Greifenberg et al., 2009) and there is some evidence that MDSCs can also restrict B cell development (Ostrand‐Rosenberg, 2010). However, the effect of MDSCs on NK and NKT cells is less clear; there are conflicting reports in the literature regarding whether MDSCs suppress or activate these cell types (Elkabets et al., 2010; Nausch et al., 2008; Ostrand‐Rosenberg, 2010; Pastula and Marcinkiewicz, 2011). It is possible that different MDSC subsets have distinct effects on NK and/or NKT cells and this may provide a rationale for the apparently dissimilar results. MDSCs have also been implicated in fostering cancer progression via promotion of angiogenesis, tumor cell invasion, and metastasis (Priceman et al., 2010; Sinha et al., 2005b; Yang et al., 2004; Yang et al., 2008; Youn and Gabrilovich, 2010). MDSCs are considered one of the main factors responsible for the failure of immunotherapy, both in cancer patients and tumor‐bearing mice, and are an attractive therapeutic target (Kusmartsev et al., 2008; Ostrand‐Rosenberg, 2010; Vieweg et al., 2007). Consequently, a better understanding of  28  MDSC immunosuppression in cancer will be vital to developing effective immune‐ mediated cancer therapies. It is worth noting that although MDSCs have been most commonly studied in the context of cancer, they are induced in a number of chronic and acute inflammatory conditions, including autoimmune disease, trauma, burns, parasitic infections, and sepsis (Cuenca et al., 2011; Van Ginderachter et al., 2010). Indeed, the factors that generate a normal immune response and mobilization of mature Mφs, PMNs, and immature populations from the BM and blood to inflammatory sites, also induce accelerated myelopoiesis and expansion of MDSCs (Ueda et al., 2009). Related to this, the expansion, accumulation (i.e. inhibition of terminal differentiation), and function of MDSCs are driven by pro‐inflammatory factors produced by the tumor, tumor stroma, and infiltrating T cells. These include cytokines (i.e. IL‐1β, IL‐6) (Bunt et al., 2007; Elkabets et al., 2010), chemokines, PGE2 (Eruslanov et al., 2010; Serafini, 2010), and the S100A8/A9 proteins (Ostrand‐Rosenberg and Sinha, 2009). In addition, NF‐κB signaling appears to be required for the complete activation of these cells (Delano et al., 2007). Exposure of MDSCs to these factors increases the pro‐inflammatory functions of MDSCs, including increased ROS, cytokine (i.e. IFN‐γ, IL‐10, TNF‐α), and chemokine production (i.e. CCL3, CCL4, CCL5, CXC chemokine ligand 12; CXCL12) (Delano et al., 2007). The mechanisms regulating these processes have been elucidated by a number of groups (Corzo et al., 2009; Nagaraj et al., 2007) and have been recently reviewed (Cuenca et al., 2011). 1.8 Tumor metastasis and the pre­metastatic niche Historically, the major focus of cancer research has been primary tumor development and progression. However, since the vast majority of cancer patients die from metastasis there has been a recent shift in focus towards investigating the factors that control tumor metastasis and this has led to the emergence of several novel hypotheses and models that shed light on this intricate process (Coghlin and Murray,  29  2010). In the past, metastasis was considered to be a stepwise accumulation of genetic events driven by clonal evolution (Fidler, 2003). However, new evidence from genetic studies suggests that the vast majority of tumor cells inherently possess the ability to metastasize (Singh et al., 2002; van 't Veer et al., 2002) and that tumor cells may separate from the primary tumor much earlier than previously appreciated (Husemann et al., 2008; Schmidt‐Kittler et al., 2003). Furthermore, the critical role that host‐derived factors play in driving tumor metastasis is now becoming clear. Our understanding of metastasis has been greatly influenced by the idea that the behaviour of individual tumor metastases is highly dependent on interactions between tumor cells (the ‘seeds’) and the host microenvironment (the ‘soil’) (Coghlin and Murray, 2010). This idea was first proposed by Stephen Paget in 1889 based on his analysis of metastatic spread in cancer patient autopsies (Paget, 1889). Today we acknowledge that multiple layers of cross‐talk occur between malignant cells and normal physiological cells, including stromal, endothelial, inflammatory, and/or BM‐ derived cells, which together comprise the metastatic microenvironment (Laoui et al., 2011; Psaila et al., 2006). There is evidence that although tumor cells may be continually released from a primary site, relatively few of them are able to efficiently form macrometastases (Luzzi et al., 1998). In fact, the most probable outcome for these cells is death, with extravasation and establishment of micrometastases serving as rate‐ limiting steps in this process (Joyce and Pollard, 2009). Perhaps one of the most important concepts that has transformed the way we conceptualize metastasis is that of the ‘pre‐metastatic niche’ (Coghlin and Murray, 2010; Kaplan et al., 2006), based on landmark findings by Kaplan et al. (Kaplan et al., 2005). These studies demonstrate that recruitment of VEGFR‐1 expressing HPCs represents an essential cellular event in the process of metastasis (Kaplan et al., 2005). Prior to the arrival of metastatic tumor cells, HPCs home to specific sites and form clusters, comprising the ‘pre‐metastatic niche’, that identify sites of future metastases (Kaplan et al., 2005). Their results suggest a functional role for VEGFR‐1, since anti‐VEGFR‐1 blocking Abs specifically prevented tumor metastases in their model (Kaplan et al., 2005). Furthermore, they report that soluble factors produced by primary tumor cells can induce VEGFR‐1 progenitor cells to 30  migrate from the BM to specific metastatic sites, where they direct the seeding and proliferation of metastatic lesions in a specific manner (Kaplan et al., 2005). However, there are conflicting reports regarding the cell types that comprise the niche as well as the manner in which the niche functions (Qian and Pollard, 2010). For example, some groups have challenged whether VEGFR‐1 is required for metastasis (Dawson et al., 2009), while others have proposed that the niche provides sites for tumor cells to adhere and grow (Psaila and Lyden, 2009) or acts as a reservoir of monocytes that differentiate into Mφs upon tumor cell arrival (Qian et al., 2009). There is also evidence that hypoxia mediates metastatic niche development. Specifically, hypoxia‐induced lysyl oxidase (LOX) has been reported to accumulate at pre‐metastatic sites, where it is required for the recruitment of CD11b+ myeloid cells, which in turn promote the invasion and recruitment of other BM‐derived cells, as well as metastatic tumor cells (Erler et al., 2009). Although it is now evident that BM‐derived progenitors and inflammatory cells contribute to metastasis (Psaila et al., 2006), the specific cell types and mechanisms involved are still being explored. Substantial evidence suggests both MDSCs and Mφs play roles in metastasis through production of soluble factors (including growth factors, proteolytic  enzymes,  cytokines,  chemokines,  and  inflammatory  mediators),  modification of the extracellular matrix (ECM), and interactions with tumor cells (Qian and Pollard, 2010; Siveen and Kuttan, 2009; Youn and Gabrilovich, 2010). Consequently, MDSCs and Mφs function to promote metastasis through several mechanisms including encouragement of tumor cell survival and proliferation, promotion of angiogenesis, and enhancement of tumor cell migration and invasion (Condeelis and Pollard, 2006; Youn and Gabrilovich, 2010). 1.9 4T1 and 67NR murine mammary tumor models As mentioned above, metastatic disease is responsible for the majority of deaths in breast cancer patients (Psaila et al., 2006) and a number of animal models have been  31  developed over the years to investigate key aspects of the metastatic process (Heppner et al., 2000). One of the most useful is the 4T1 murine mammary carcinoma model because it closely resembles the natural history and histopathology of human breast cancer and because it is one of a series of sister models that represent the spectrum of metastatic disease (Heppner et al., 2000). Specifically, a comparison of the metastatic 4T1 to its non‐metastatic sister 67NR facilitates studies aimed at elucidating the factors responsible for primary tumor growth versus metastasis. Both 4T1 and 67NR cell lines were derived from a single, spontaneous mouse mammary tumor (Aslakson and Miller, 1992; Miller et al., 1983). The thioguanine‐resistant 4T1 subline was derived from this parental population while the geneticin‐resistant 67NR subline was obtained by transfection of line 67 (Aslakson and Miller, 1992). While both lines are highly tumorigenic and form primary tumors in the mammary fat pad, the 4T1 cell line is highly metastatic while 67NR cells are non‐metastatic (Aslakson and Miller, 1992). Another key difference between these two cell lines is the vascularization of the primary tumor; primary 4T1 tumors are hypoxic while 67NR primary tumors are well vascularized and normoxic (Aslakson and Miller, 1992). A number of studies have used the 4T1 tumor model to investigate the role of myeloid cells in metastasis. When these cells are injected orthotopically into the mammary fat pad of syngeneic BALB/c mice, they share several important characteristics with human breast carcinomas (Heppner et al., 2000; Pulaski and Ostrand‐Rosenberg, 2001). In particular, 4T1 tumor cells spontaneously metastasize to the lung, liver, bone, and brain via the bloodstream (Aslakson and Miller, 1992; Heppner et al., 2000), consistent with late‐stage human breast cancer. As well, mice bearing 4T1 tumors exhibit significant expansion and accumulation of MDSCs in the lungs, spleen, blood, and primary tumor (Ostrand‐Rosenberg, 2010; Youn et al., 2008). These 4T1‐induced MDSCs have been shown to accumulate in response to pro‐ inflammatory mediators (i.e. IL‐1β, IL‐6, PGE2, S100A8/A9) (Ostrand‐Rosenberg, 2010), contribute to decreased immune surveillance (Sinha et al., 2005c), and block both innate and adaptive anti‐tumor immunity (Sinha et al., 2005b). They have been  32  reported to exert their immunosuppressive effects by a number of mechanisms including upregulation of Arg1 expression and the resulting depletion of L‐Arg, cysteine sequestration, and down‐regulation of L‐selectin expression on T cells (Ostrand‐ Rosenberg, 2010). Less is known about the role Mφs play in the 4T1 tumor model. However, it has been shown that 4T1 tumor cells skew the differentiation of naïve Mφs into M2 Mφs in an IL‐6‐dependent manner (Wang et al., 2010). Furthermore, 4T1‐induced MDSCs and Mφs have been reported to work together to suppress immune surveillance against metastasis; specifically, contact‐dependent cross‐talk between MDSCs and Mφs leads to increased IL‐10 production by MDSCs and decreased IL‐12 production by Mφs (Sinha et al., 2007a; Sinha et al., 2005b). On the other hand, relatively little is known about the role of Mφs or MDSCs in the 67NR tumor model. Additional studies will likely reveal whether myeloid cells are involved in tumor growth in this non‐metastatic model of mammary carcinoma. 1.10 Aims of study Although it is well established that immune suppression contributes to all stages of cancer, there are many aspects that remain unclear. First, multiple myeloid and lymphoid cell populations have been proposed to possess immunosuppressive properties. However, little is known about the specific roles each of these cell types play in different situations. For example, do the same cell types that contribute to peripheral tolerance under physiological circumstances also promote tumor growth? Moreover, the signaling pathways that regulate the immunosuppressive properties of different cells are not yet well defined. Given the dual roles these cells play in both the promotion and inhibition of immune responses, understanding the specific factors that regulate their function is of particular importance. Furthermore, very few studies have directly compared the immunosuppressive potencies of different cell types, or the different mechanisms they use to exert their suppressive effects. Finally, it is poorly understood  33  how the strength of immune suppression relates to the importance of these cells in tumor growth or metastasis. A better understanding of factors that control the development and function of immunosuppressive myeloid cells is critical for the design of therapies that could be used to treat diseases such as cancer and autoimmunity via the modulation of immune suppression. The work presented in this thesis addresses a number of these issues, especially the different roles immunosuppressive cells play in normal physiology and in cancer. We chose to focus on elucidating the immunosuppressive features of mononuclear phagocytes, and Mφs in particular, since the role of regulatory lymphoid cells (i.e. Tregs) was being carried out with a collaborator (Dr. Megan Levings, CFRI) and the suppressive properties of different DC subsets was being investigated by another graduate student in our laboratory (Antignano et al., 2011). Moreover, the regulatory functions of Mφs is a topic that has received very little attention to date (Brem‐Exner et al., 2008; Munn and Mellor, 2003). Our overall goal was to explore the immunosuppressive functions of Mφs and elucidate the factors that regulate their suppressive properties, the mechanisms by which Mφs suppress T cell responses, the relative potency of Mφ suppression compared to other myeloid cell types such as MDSCs, and their roles in promoting tumor development, growth, and metastasis.  34  CHAPTER 2 : MATERIALS AND METHODS 2.1 Mice BALB/c (Taconic; Germantown, NY) and C57BL/6 and OTII transgenic mice (The Jackson Laboratory; Bar Harbor, ME) were bred in house. BALB/c DO11.10 transgenic mice, which express a TCR specific for chicken ovalbumin (OVA) peptide (323‐339) restricted to I‐Ad (Murphy et al., 1990), were purchased from The Jackson Laboratory and spleens from OTI transgenic mice were kindly provided by Dr. Hung‐Sia Teh (The University of British Columbia, Vancouver, BC, Canada). Female mice, 6 to 12 weeks of age, were used for all experiments. Mice were maintained in the Animal Resource Centre at the BC Cancer Research Centre under specific‐pathogen free conditions. All animal experiments were performed in accordance with institutional and Canadian Council on Animal Care guidelines. 2.2 Media For the studies presented in Chapter 3, cells were cultured in Iscove’s modified Dulbecco’s medium (IMDM) + 10% fetal calf serum (FCS) medium (i.e. L‐Arg‐free IMDM (HyClone, Logan, UT) + 10% FCS, 150 µM monothioglycerate (MTG), 50 U/ml penicillin, 50 µg/ml streptomycin, and 100 µM L‐Arg). For the studies presented in Chapter 4, cells were cultured in HL1 medium (i.e. HL‐1 serum‐free medium (BioWhittaker; Basel, Switzerland) + 1% penicillin, 1% streptomycin, 1% Glutamax, 5 x 10‐5 M beta‐ mercaptoethanol (β‐ME)), or Roswell Park Memorial Institute (RPMI) 1640 + 10% FCS medium (i.e. RPMI 1640 (StemCell Technologies; Vancouver, BC, Canada) + 10% FCS, 150 µM MTG, 50 U/ml penicillin and 50 µg/ml streptomycin). For the studies presented in Chapter 5, cells were cultured in HL1 medium. For the MDSC serum studies, we tested four different sources of FCS purchased from HyClone, StemCell Technologies, Invitrogen (Burlington, ON, Canada), and PAA  35  Laboratories (Pasching, Austria) and two sources of bovine serum albumin (BSA) from Sigma‐Aldrich (St. Louis, MO) and Roche Diagnostics (Laval, QC, Canada). BALB/c mouse serum was purchased from Innovative Research (Novi, Michigan) or harvested by cardiac puncture from BALB/c mice. 2.3 Reagents and cytokines Neutralizing Abs to cytokines were purchased as follows: IL‐4 (eBioscience; San Diego, CA); IL‐10 and cytotoxic T‐lymphocyte antigen 4 (CTLA‐4) (BD Biosciences; Mississauga, ON, Canada); IL‐13, IFN‐γ, and TGF‐β (R&D Systems; Minneapolis, MN); and IFN‐β (Pestka Biomedical Laboratories; Piscataway, NJ). Reagents were purchased as follows: recombinant human latency‐associated peptide (LAP) (R&D Systems); N6‐ (1‐iminoethyl)‐L‐lysine (L‐NIL), a specific inhibitor of iNOS (Calbiochem; San Diego, CA); carboxy‐2‐phenyl‐4,4,5,5‐tetramethylimidazoline‐1‐oxyl‐3‐oxide (PTIO), an NO scavenger (Cayman Chemicals; Ann Arbor, MI); soluble anti‐CD3 and anti‐CD28 (eBioscience); OVA peptides (257‐264 and 323‐339) (GenScript; Piscataway, NJ); and Celebrex, a cyclooxygenase (COX)‐2 inhibitor (Pfizer; St. Louis, MO). [(S)‐(2‐ Boronoethyl)‐L‐cysteine] (BEC), a competitive inhibitor of Arg1 and 2 that does not inhibit iNOS, was from Dr. Jean‐Luc Boucher (Université Paris Descartes, Paris, France). TIB‐218, a rat IgG2aκ Ab selective for the β subunit of mouse lymphocyte function‐ associated antigen 1 (LFA‐1) and CD11b (CD18), was purified from hybridoma supernatants in house. E.coli LPS serotype O127:B8 and the double stranded ribonucleic acid (dsRNA) analog polyinosinic:polycytidylic (Poly I:C) acid were from Sigma‐Aldrich.  The  high‐performance  liquid  chromatography  (HPLC)‐purified  phosphorothioate‐modified CpG‐dinuclotide (CpG), 5'‐tccatgacgttcctgacgtt‐3' was from Invitrogen and peptidoglycan (PGN) from Staphylococcus aureus was from Fluka (Buchs, Switzerland). Recombinant mouse (rm) IFN‐γ and IL‐2 were from StemCell Technologies and IFN‐β was from Sigma‐Aldrich. Unless otherwise stated, all tissue culture reagents were from StemCell Technologies and all other reagents were from Sigma‐Aldrich.  36  2.4 Isolation of myeloid cells 2.4.1  Peritoneal macrophages  The peritoneal cavities (PCs) of mice were lavaged with 3 x 5 ml IMDM + 10% FCS medium or HL1 medium containing 1 mM ethylenediaminetetraacetic acid (EDTA) (USB Corporation, Cleveland, OH). The number of peritoneal macrophages (PMφs) was visually enumerated using a hemocytometer and cells were resuspended in fresh medium without EDTA. Cells were plated and allowed to adhere for at least 3 hours (h) before the non‐adherent cells were washed away. When required, adherent cells were harvested using cell dissociation buffer (Invitrogen/Gibco) and gentle scraping. Analysis of the adherent cells revealed that >95% co‐expressed F4/80 and CD11b. In experiments where the ratio of Mφs to splenocytes was varied, the number of Mφs plated per well was adjusted to give the desired ratio, while the number of splenocytes remained constant (2 x 105 cells/well [96‐well plate] or 1 x 106 cells/well [48‐well plate]). 2.4.2  Splenic, pulmonary, or tumor­associated Mφs and MDSCs  To isolate Mφs or MDSCs, single‐cell suspensions were prepared from the spleens, kidneys, livers, lungs, and/or tumors of naïve or tumor‐bearing female BALB/c mice. Spleens and livers were passed through a 70 µm filter to create a single‐cell suspension. Lungs, kidneys, and tumors were finely minced with crossed scalpels prior to agitation for 40 minutes (min) at 37°C with an enzyme suspension containing 0.5% trypsin (BD Biosciences) and 0.08% collagenase (Sigma‐Aldrich) in phosphate buffered saline (PBS) (lungs and kidneys) or 30 min at 37°C with 125 μg Liberase (Roche Diagnostics) in IMDM (tumors). After incubation, 0.06% DNase (Sigma‐Aldrich) was added, and the cell suspension was gently vortexed and filtered through 30 µm nylon mesh to remove clumps. Samples for flow cytometry were treated with ammonium chloride solution (0.8% with 0.1 mM EDTA; 7 min on ice) for erythrocyte lysis and  37  either fixed in 70% EtOH and stored at ‐20°C for subsequent flow cytometry analysis or analyzed immediately. Samples for magnetic separation were washed by centrifugation (1200 rpm, 5 min) and resuspended in PBS + 2% FCS + 1mM EDTA. Mφs were isolated using F4/80‐ PE positive selection (EasySep; StemCell Technologies) and MDSCs were isolated using Gr1‐PE positive selection (EasySep; StemCell Technologies), according to the manufacturer’s instructions (Table 2.1). Cell numbers were quantified using a hemocytometer and purity was determined to be >90% by flow cytometry. 2.5 Flow cytometry All flow cytometry samples were prepared in 96‐well V‐bottom plates. To label cells with surface marker Abs, cells were suspended in Hank’s balanced salt solution (HBSS) + 2% FCS + 0.05% NaN3 (HFN) and incubated with anti‐mouse CD16/CD32 (2.4G2) (BD Pharmigen, Mississauga, ON, Canada) for 10 min at 4°C to block Fc receptors before labeling. Flurochrome conjugated Abs (Table 2.1) were added at pre‐ determined optimal concentrations for 30 min at 4°C. Cells were then washed twice and resuspended in HFN for analysis. Data were acquired using a FACSCalibur flow cytometer (BD Biosciences) and analyzed using FlowJo software (Tree Star, Inc., Ashland, OR). Where reported, absolute numbers of cells were calculated by multiplying the total number of cells (enumerated using a hemocytometer) by the proportion of that specific cell type, as determined by flow cytometry. To stain cells for intracellular Foxp3 expression, cells were resuspended in 50 μl HFN + 2.4G2 Ab and 100 μl Fixation/Permeabilization working solution (eBioscience) was added to each sample. Following incubation for 30 min at 4°C, cells were washed twice with 100 μl 1x Permeabilization Buffer (eBioscience) and resuspended in 50 μl 1x Permeabilization Buffer containing 2% rat serum (Stem Cell Technologies). Abs were  38  added as indicated in Table 2.1 for 30 min at 4°C. Cells were then washed twice and resuspended in HFN and analyzed by flow cytometry.  Table 2.1  List of antibodies used in this thesis.  FC – Flow Cytometry, WB – Western Blotting, IF – Immunofluorescence, PS – Positive Selection Ms – mouse, Rb – rabbit, c – cells or cell equivalents pAb – polyclonal Ab, mAb – monoclonal Ab, Ig – immunoglobulin APCy – allophycocyanin, Cy – cyanine, FITC ‐ fluorescein isothiocyanate, PE ‐ phycoerythrin Antibody (Ab) Annexin V‐PE Arg 1 B220‐PE CD3‐APCy CD4‐APCy CD8‐PE CD8‐FITC CD11b CD11b‐APCy CD11b‐PE CD11b‐PE CD11c‐APCy CD40‐FITC CD86‐FITC FcεR1α‐FITC F4/80 F4/80‐PE F4/80‐PE GAPDH (glyceraldehyde 3‐ phosphate dehydrogenase) Gr1 Gr1‐PE  Type n/a mAb (Ms) Rat IgG2a Armenian Hamster IgG1 Rat IgG2b Rat IgG2a Rat IgG2a Rat IgG2b Rat IgG2b Rat IgG2b Rat IgG2b Armenian Hamster IgG Rat IgG2a Rat IgG2a Armenian Hamster IgG1 Rat IgG2a Rat IgG2a Rat IgG2a mAb (Ms)  Rat IgG2b Rat IgG2b  Concentration/ Dilution used 5 μl/106 c 1:2000 0.2 μg/106 c 0.2 μg/106 c  Procedure  Source  FC WB FC FC  BD Pharmingen BD Bioscience BD Pharmingen BD Pharmingen  0.125 μg/106 c 1 μg/106 c 0.5 μg/106 c 1:100 0.125 μg/106 c 1 μg/106 c 1 μg/ml 0.125 μg/106 c  FC FC FC IF FC FC PS FC  eBioscience StemCell Technologies eBioscience eBioscience eBioscience StemCell Technologies StemCell Technologies eBioscience  1 μg/106 c 1 μg/106 c 0.125 μg/106 c  FC FC FC  BD Pharmingen BD Pharmingen eBioscience  1:100 0.1 μg/106 c 1 μg/ml 1:40,000  IF FC PS WB  eBioscience Invitrogen Invitrogen Research Diagnostics Inc. (Flanders, NJ)  1:100 1 μg/106 c  IF FC  eBioscience eBioscience  39  Antibody (Ab)  Type  Procedure  Source  PS FC WB  eBioscience eBioscience Santa Cruz Biotechnology, (Santa Cruz, CA) BD Pharmingen Hydroxyprobe, Inc. (Burlington, MA) Cell Signalling (Beverly, MA) StemCell Technologies Invitrogen  Gr1‐PE Foxp3‐PE‐Cy7 iNOS  Rat IgG2b Rat IgG2a pAb (Rb)  Concentration/ Dilution used 1 μg/106 c 0.25 μg/106 c 1:750  MHCII‐FITC Pimonidazole‐FITC  Rat IgG2a Mouse IgG  1 μg/106 c 1:100  FC IF  pSTAT1 Y701  pAb (Rb)  1:1000  WB  Ym1 Goat anti‐rat Alexa 488 Goat anti‐rat Alexa 594  pAb (Rb) pAb (Goat)  1:2000 1:100  WB IF  pAb (Goat)  1:100  IF  Invitrogen  2.6 T cell proliferation assay and cytokine assays Erythrocyte‐depleted splenocytes were cultured in IMDM medium or RPMI medium + 10% FCS, or HL1 medium ± 10% FCS at 2 x 105 cells/well ± irradiated (2000 rads) test cells in a total volume of 150 μl in 96‐well, flat‐bottom tissue culture plates. These splenocytes were stimulated as follows: C57BL/6 or BALB/c splenocytes with 1 μg/ml anti‐CD3 + 5 μg/ml anti‐CD28 (polyclonal), OTII or DO11.10 splenocytes with 10 μg/ml OVA peptide (323‐339); and OTI splenocytes with 10 μg/ml OVA peptide (257‐ 264). Inhibitors were added as indicated. Cells were incubated at 37°C, 5% CO2 for 72 h, with 1 μCi/well 3H‐thymidine (thy) (2 Ci/mM; PerkinElmer, Woodbridge, ON, Canada) added for the last 18 h. For cell contact studies, test cells were added to the lower chamber of Transwell plates (0.4 μm pores, polycarbonate membrane; Corning Life Sciences, Corning, NY) and splenocytes were added to the upper chamber. Cells were harvested onto glass fiber filter mats using a 96‐well harvester (Molecular Devices, Sunnyvale, CA). Filter mats were sealed in plastic bags with 10 ml of scintillation fluid and 3H‐thy incorporation measured using a Betaplate liquid scintillation counter (Wallac, Waltham, MA). Data are expressed as counts per minute (cpm) (mean ± SEM) of triplicate cultures. Responder control (RC) indicates stimulated splenocytes alone.  40  The relative percentage of splenocyte proliferation was calculated as: (Proliferation of stimulated splenocytes with test cells) x 100% (Proliferation of stimulated splenocytes (RC) alone) To measure cytokine production, the above assay was performed with the following changes: 5 x 105 splenocytes/well were added to 48‐well plates, the assay was performed in a total volume of 375 μl, and no 3H‐thy was added. Cell‐free supernatants were collected after 72 h and IL‐2, IL‐4, IL‐10, and IFN‐γ production assayed using cytokine enzyme‐linked immunosorbent assay (ELISA) kits (BD Biosciences), according to the manufacturer’s instructions. To assay cell proliferation by flow cytometry, erythrocyte‐depleted splenocytes were resuspended in 37°C PBS and 5 μM carboxyfluorescein succinimidyl ester (CFSE) (Sigma‐Aldrich) was added to each sample. Cells were incubated for 10 min at 37°C and then washed twice in 4°C PBS prior to stimulation and co‐culture with test cells. After 72 h, non‐adherent cells were harvested, stained with fluorescent Abs against CD3, CD4, and/or CD8, and analyzed by flow cytometry. Histogram peaks represent the number of T cell divisions and the proportion of divided cells is indicated. 2.7 In vitro Mφ skewing Mφs were plated at 1.25 x 105 cells/well in 500 μl IMDM + 10% FCS and treated with this medium alone, with 100 ng/ml IFN‐γ + 100 ng/ml LPS, or with 10 ng/ml IL‐4 + 10 ng/ml IL‐13. After 72 h, Mφs were subjected to Western blot analysis. 2.8 SDS­PAGE and Western blot analysis Cells were washed with PBS and cell pellets lysed with 1x sodium dodecyl sulfate (SDS) sample buffer (i.e. 8.5% (v/v) glycerol, 0.5% (w/v) SDS and 0.71 M β‐ME). Samples were then boiled for 2 min, allowed to cool, and DNA sheared with a 26 gauge needle. Equal numbers of cells (total cell lysates) were then loaded onto 10% 41  polyacrylamide gels and subjected to SDS‐polyacrylamide gel electrophoresis (PAGE) and Western blot analysis as described previously (Damen et al., 1995). Abs used for Western blotting are indicated in Table 2.1. 2.9 Viability and morphology assays Following 72 h co‐culture of stimulated splenocytes with myeloid test cells, non‐ adherent cells were harvested and analyzed by flow cytometry for expression of CD3, CD4, CD8, Annexin V‐PE, and/or propidium iodide (PI) (Table 2.1). Annexin V staining was performed by washing cells (that had been previously stained with cell surface Abs) twice with cold PBS and then resuspending cells in 100μl 1x Binding Buffer (100 mM 4‐(2‐hydroxyethyl)‐1‐piperazineethanesulfonic acid (HEPES), pH 7.4; 140 mM NaCl; 2.5 mM CaCl2). 5 μl Annexin V‐PE and 100 ng PI were added to each sample, cells were gently mixed, and incubated for 15 min at 23°C in the dark. Cells were then diluted with 50 μl 1x Binding Buffer and analyzed by flow cytometry. Cells were classified as viable (Annexin V‐PE and PI negative), undergoing apoptosis (Annexin V‐PE positive, PI negative), or dead (Annexin V‐PE and PI positive). For some experiments cells were stained for CD3, CD4, and/or CD8, and PI prior to flow cytometric analysis to determine bulk (total cells), total T cell (CD4+ or CD8+), or CD4+ or CD8+ T cell viability. The increase in T cell viability was calculated as: (Percent viability of stimulated splenocytes with test cells) x 100% (Percent viability of stimulated splenocytes (RC) alone) Cell viability was also determined via trypan blue exclusion. Trypan blue was added to either adherent or non‐adherent cells and the proportion of cells that excluded trypan blue (viable cells) was determined using a hemocytometer. To assess MDSC morphology, MDSCs were plated at 1 x 106 cells/ml in HL1 medium with or without 10% FCS and incubated overnight (O/N) at 37°C. The next day, cells were harvested and cytospin preparations were stained with Giemsa‐Eosin. The  42  morphology of cells was assessed by light microscopy using an Axiovert S100 microscope (Carl Zeiss Canada Ltd., Toronto, ON, Canada) and images captured with a Retiga EXi camera (QImaging, Surrey, BC, Canada). 2.10 Nitric oxide assay NO production was determined indirectly by measuring the accumulation of the stable end product, nitrite (NO2‐), in cell‐free culture supernatants using the Griess assay (Kleinbongard et al., 2002). Briefly, 50 μl of supernatant was aliquoted into wells of a 96‐well plate at 23°C. Then 50 μl of solution A (1% sulfanilamide in 2.5% phosphoric acid) was added to each sample, followed by the addition of 50 μl of solution B (0.1% phenylnapthylenediamine dihydrochloride in 2.5% phosphoric acid). After 5 min, the absorbance of samples at 570 nm was determined and NO2‐ concentration calculated by comparison to a standard curve. 2.11 In vitro Mφ pre­treatment and stimulation Mφs were plated at 1.25 x 105 cells/well in 500 μl IMDM medium + 10% FCS or HL1 medium in 48‐well plates or 2.5 x 104 cells/well in 100 μl in 96‐well plates. Mφs were pre‐treated directly ± 100 ng/ml LPS, 0.3 μM CpG, 5 μg/ml dsRNA, 5 μg/ml PGN, or IFN‐β (concentrations as indicated), ± anti‐IFN‐β (500 or 1000 U/ml) or BEC (200 μM), or indirectly with polymyxin B‐treated (50 mg/ml) supernatants from Mφs stimulated for 24 h ± 100 ng/ml LPS. After 24 h, supernatants were collected and IFN‐β measured by ELISA. Mφs were then washed with fresh IMDM medium + 10% FCS and co‐cultured with activated splenocytes or stimulated with 100 ng/ml IFN‐γ + 100 ng/ml LPS for an additional 72 h. After 72 h, proliferation was measured, supernatants were collected for analysis of NO production, and/or cell lysates were prepared for Western blot analysis.  43  2.12 IFN­β ELISA To assay IFN‐β production we used an ELISA we adapted from two previous publications (Punturieri et al., 2004; Weinstein et al., 2000) and have since published (Sly et al., 2009). Briefly, Maxisorp ELISA 96‐well plates (Nalgene Nunc International, Rochester, NY) were coated O/N with 0.1 μg IFN‐β capture Ab (rat anti‐mouse IFN‐β mAb 7F‐D3; Seikagaku America, Falmouth, MA, USA) in 100 μl PBS. The next day, wells were washed five times with wash buffer (PBS + 0.05% Tween‐20), blocked for 2 h at 23°C with 200 μl assay diluent (PBS + 10% heat‐inactivated FCS), and then washed three times with wash buffer. 100 μl of each sample or standard (rmIFN‐β) was added to each well and incubated O/N at 4°C. The wells were washed 7x with wash buffer and then 100 μl of detection Ab (25 U/ml rabbit anti‐mouse IFN‐β polyclonal Ab in assay diluent; R&D Systems, Minneapolis, MN) was added to each well and incubated for 2 hrs at 23°C. The wells were washed 7x with wash buffer with 1 min soaks between each wash. ELISA substrate (100 μl) (BD OptEIA TMB Substrate Reagent Set; BD Biosciences, Mississauga, ON, Canada) was added to each well and the plate incubated for 20‐30 min at 23°C in the dark. The reaction was stopped with 50 μl 2N H2SO4, the absorbance of the samples at 450 nm determined, and the concentration of IFN‐β calculated from the standard curve. 2.13 Tumor models 4T1, 4TO7, and 67NR murine mammary carcinoma cells were a kind gift from Dr. Fred Miller (Karmanos Cancer Institutes, Detroit, MI). These cell lines were derived from a spontaneous mammary tumor in a BALB/cfC3H mouse (Dexter et al., 1978; Miller et al., 1983) and represent different levels of metastatic propensity (Aslakson and Miller, 1992). 4T1 tumor cells metastasize to the lung, liver, bone, and brain; 4TO7 cells metastasize to the lungs, but fail to grow into macroscopic metastases; 67NR cells do not metastasize. BALB/c mice were anesthetized with 2% isoflurane in O2 and were orthotopically inoculated with 105 4T1 cells, 106 4TO7 cells, or 2 x 105 67NR cells in 50  44  μl PBS in the fourth mammary fat pad. We have found that these cell concentrations produce consistent tumor growth rates with tumor volumes that approach ethical restrictions four weeks after implantation. Where indicated, mice were subcutaneously implanted with 5 mg or 10 mg slow‐release all‐trans retinoic acid (ATRA) pellets (Innovative Research of America, Sarasota, FL) at the base of neck. Mice were sacrificed at various time points and weights of spleens and primary tumors were measured. Unless otherwise indicated, mice were sacrificed three weeks post‐tumor implant or three weeks post‐ATRA pellet implant. 2.14 Serum treatment for MDSC assay studies FCS was untreated, filtered (0.22 µm), dialyzed (3.5 kDa (kilodalton) MW cut off), or heat inactivated (55°C, 45 min) and then added to HL1 medium to a final serum concentration of 10%. Albumin was removed from FCS using Affi‐Gel Blue Gel (Bio‐Rad; Mississauga, ON, Canada) affinity chromatography according to the manufacturer’s instructions. 2.15 ROS Detection Pulmonary and splenic MDSCs were plated O/N in 6‐well plates at 1 x 106 cells/ml in HL1 medium ± 10% FCS or 3% BSA. The next day, cells were harvested using 1 ml/well of cell dissociation buffer (Invitrogen/Gibco) and gentle scraping. Cells were  washed  twice  with  PBS,  and  stained  with  20  µM  DCFDA  dye  (Invitrogen/Molecular Probes) in PBS for 45 min at 37°C, with the cells gently mixed every 10 min during this incubation. MDSCs were then washed twice with 4°C PBS and then plated in 96‐well plates for cell surface staining, as previously described in section 2.5. MDSCs were stained for CD11b, Gr1, and PI, and then suspended for analysis by flow cytometry. The fluorescent intensity of DCFDA (measured in the FL‐1 channel) of live MDSCs (CD11b+Gr1+PI‐) was determined and mean fluorescence intensity (MFI) recorded.  45  2.16 Clonogenic assays Monodispersed lung cells were washed by centrifugation prior to treatment with ammonium chloride. Cells were washed in PBS, resuspended in medium, and aliquots of 3x103 to 106 cells were plated in clonogenic assays containing 60µM or 30µM 6‐ thioguanine (for 4T1 and 4TO7 respectively) or 250µg/ml geneticin (for 67NR). Cells were incubated for 9‐12 days (37°C, 5% CO2) and then colonies stained with malachite green and manually counted. The total number of tumor cells in the lungs was calculated as follows: (Proportion of colony forming tumor cells x Total number of lung cells) 2.17 Immunofluorescence Pimonidazole (100 mg/kg; Hydroxyprobe, Inc.; Burlington, MA) was injected intraperitoneally (IP) 90 min prior to tumor excision and Hoecsht 33342 (500 μg/mouse; Sigma‐Aldrich) was injected intravenously (IV) 20 min prior to tumor excision. Tumors were rinsed in 4°C PBS prior to freezing in Optimal Cutting Temperature (OCT) medium (Sakura Finetek; Torrance, CA). Lungs were inflated with 300 μl 50:50 OCT:PBS mixture and rinsed in 4°C PBS prior to freezing in OCT. Serial sections (8‐10 μm) were cut with a cryostat and stained with the appropriate Abs (Table 2.1) in PBS containing 4% calf serum with or without 0.1% Triton X, for anti‐ pimonidazole or surface marker Abs, respectively. Where indicated, samples were fixed with methanol and stained with 4',6‐diamidino‐2‐phenylindole (DAPI).  46  2.18 Statistical analysis Data are represented as the mean ± standard error of the mean (SEM) and all experiments were performed a minimum of three times. Two‐tailed Student t tests were performed using Microsoft Office Excel 2007. Values of p≥0.05 were considered not significant (NS) *,p<0.05; **,p<0.01; ***,p<0.001  47  CHAPTER 3 : TOLL­LIKE RECEPTOR AGONISTS THAT INDUCE IFN­β ABROGATE RESIDENT MACROPHAGE SUPPRESSION OF T CELLS 3.1 Introduction Immune cells that function to suppress T cell responses are critical for the maintenance of immune homeostasis and the prevention of auto‐immune disease. Negative regulation of T cells occurs, not only during development in the thymus, but also in peripheral tissues (Bettini and Vignali, 2009), such as the lung and PC. Tregs, myeloid DCs, and Mφs have all been shown to play a role in peripheral tolerance (Bettini and Vignali, 2009; Munn and Mellor, 2003); however, the role of Mφs as regulatory cells has received little attention (Brem‐Exner et al., 2008; Munn and Mellor, 2003). Although it is well established that Mφs can perform both immunogenic and immunosuppressive functions, the mechanisms that determine whether Mφs promote or inhibit a particular immune response are not well understood. Resting, or non‐ activated, tissue Mφs continually survey the microenvironment, ingesting large amounts of extracellular fluid, apoptotic cells, and cellular debris, which they display as Ags on their surface for recognition by T cells (Gordon and Martinez, 2010). Since these Ags can be derived from either host cells (i.e. self‐Ags) or pathogens, Mφs must be able to distinguish whether a particular Ag should provoke an immune response or be tolerated. Therefore, we investigated the factors that regulate the effect of Mφs on T cell activation; we determined the mechanism by which resident PMφs suppress in vitro T cell proliferation in the absence of pathogens, and then explored the effects of different pathogen‐derived molecules on Mφ immunosuppression. In this chapter we demonstrate that in response to IFN‐γ, which is secreted by TCR‐activated T cells, resident PMφs acquire immunosuppressive properties that are mediated by NO. Furthermore, we show that pre‐treatment with TLR agonists that activate TRIF (i.e. LPS and dsRNA), but not those that exclusively signal through MyD88 (i.e. CpG and PGN), eliminates the suppressive properties of Mφs, in part via the induction of IFN‐β. These 48  data further our understanding of how Mφs contribute to T cell tolerance and suggest a novel role for TLR signaling and IFN‐β in regulating the immunosuppressive functions of Mφs. This thesis chapter is based on previously published work (Hamilton et al., 2010) and is reprinted with the permission of The American Association of Immunologists, Inc. 3.2 Results 3.2.1  Characterization of the immunosuppressive properties of PC cells  Since it has been reported that negative regulation of T cells occurs in peripheral tissues (Bettini and Vignali, 2009) and T cells are known to reside in the PC (Shiku et al., 1975), we hypothesized that the PC may contain cells that possess immunosuppressive properties. To test this, we isolated the PC cells from mice and analyzed the cell types present and their relative proportions by flow cytometry. As shown in Fig. 3.1A, the most abundant cells were B cells and CD11b+ cells, followed by DCs, T cells, and mast cells. To see if any of these cells could suppress T cell proliferation, we separated PC cells based on their adherent properties and found that adherent cells significantly suppressed T cell proliferation while non‐adherent cells activated T cell proliferation (Fig. 3.1B). Analysis of the adherent PC cells revealed they were exclusively Mφs, co‐ expressing F4/80 and CD11b (Fig. 3.1C, left) while lacking Gr1 (Fig. 3.1C, right) and other non‐Mφ lineage markers (data not shown). To verify that these PMφs were indeed responsible for T cell suppression, we isolated F4/80+ or CD11b+ cells by positive selection and found that both cell populations effectively suppressed T cell proliferation to an equal extent (Fig. 3.1D).  49  Figure 3.1 Resident PMφs possess immunosuppressive properties. A, Bulk PC cells were stained with fluorescent Abs and analyzed by flow cytometry to assess the relative proportions of cells expressing the lineage markers B220 (B cells), CD11b (myeloid cells), CD11c (DCs), CD3 (T cells), and FcεR (mast cells). B, Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bar) adherent or non‐ adherent PC cells (1 PC cell: 8 splenocytes) and T cell proliferation measured. C, Adherent PC cells were stained with fluorescent Abs to CD11b and F4/80 (left) or CD11b and Gr1 (right) and analyzed by flow cytometry. The proportions of CD11b+F4/80+ and CD11b+Gr1‐ cells are indicated. D, Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bar) F4/80+ or CD11b+ PC cells (1 PC cell: 2 splenocytes) and T cell proliferation measured. Assays were performed in duplicate (A, C) or triplicate (B, D). Data are the mean ± SEM of two independent experiments (A) or representative of three independent experiments (B, C, D). *,p< 0.05; **,p<0.01; ***,p<0.001 relative to responder control (RC; stimulated splenocytes alone). Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  3.2.2  Resident PMφs exhibit a naïve phenotype  Mφs can exhibit a variety of phenotypes depending on their activation state, from classically activated, M1, to alternatively activated M2a, b, or c Mφ phenotypes (Martinez et al., 2008). We performed Western blot analysis on lysates from PMφs to determine their activation state and found that these resident Mφs lacked features of either M1, such as iNOS expression, or M2 activation, such as Arg1 or Ym1 expression  50  (Fig. 3.2, left). However, these Mφs could become either M1 or M2 polarized with appropriate stimulation (IFN‐γ + LPS or IL‐4 + IL‐13, respectively) (Fig. 3.2, right), confirming the naïve (non‐activated) state of resident PMφs.  Figure 3.2 Resident PMφs exhibit a naïve phenotype and can be skewed to either M1 or M2 with appropriate stimulation. PMφs isolated directly ex vivo (left) or treated with medium alone (‐), IFNγ+LPS, or IL‐4+IL‐ 13 for 72 h (right) were subjected to Western blot analysis. Data are representative of three independent experiments. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  3.2.3  Resident Mφs suppress T cell proliferation and cytokine production  To further evaluate the immunosuppressive properties of Mφs, we co‐cultured resident PMφs from C57/Bl6 mice with stimulated splenocytes and found that Mφs potently suppressed T cell proliferation in response to either polyclonal or Ag‐specific stimulation of CD4+ or CD8+ T cells in a dose‐dependent manner (Fig. 3.3A). We also conducted these studies using Mφs from BALB/c mice and obtained similar results (Fig. 3.3B). Next, we tested whether Mφs could suppress T cell cytokine production. In the presence of Mφs, activated T cells produced significantly less IFN‐γ and IL‐10 (Fig. 3.3C, left and middle), suggesting that Mφs can inhibit both Th1 and Th2 cytokine production.  51  We also assayed IL‐4 levels, but the concentration of IL‐4 was below the detection limit in cultures with and without Mφs. However, co‐culture with Mφs increased T cell IL‐2 levels (Fig. 3.3C, right), suggesting that the presence of Mφs does not prevent IL‐2 production, and may restrict IL‐2 uptake by T cells. This is consistent with previous studies suggesting that in the presence of Mφs, T cells secrete IL‐2, but are unable to utilize it and remain locked in the Go/G1 phase of the cell cycle (Bingisser et al., 1998; Strickland et al., 1996). To determine whether Mφs were inducing T cell death, we assayed the proportion of dead or apoptotic T cells following co‐culture of Mφs with polyclonal or Ag‐stimulated T cells. We found that Mφs did not increase T cell death. On the contrary, they reduced the proportion of dead CD4+ and CD8+ T cells (Fig. 3.4, left). This may be due, in part, to phagocytosis of dead cells by Mφs. Interestingly, analysis of the proportion of apoptotic T cells revealed that co‐culture with Mφs promoted apoptosis of CD8+, but not CD4+, T cells (Fig. 3.4, right).  52  Figure 3.3 Mφs suppress T cell proliferation and cytokine production in a dose­ dependent manner. A, anti‐CD3/anti‐CD28‐stimulated C57BL/6 (Polyclonal) or OVA‐peptide‐stimulated OTII (CD4+ T cell Ag‐specific) or OTI (CD8+ T cell Ag‐specific) splenocytes were cultured with (black bars) or without (white bars) Mφs at different ratios (Mφs:splenocytes; 1:2, 1:8, 1:32) and T cell proliferation measured. Significance is compared to RC. B, Responder splenocytes (polyclonal‐stimulated BALB/c (Polyclonal) or OVA‐peptide‐stimulated DO11.10 (CD4+ Ag‐ specific)) were cultured with (black bars) or without (white bars) Mφs at different ratios (Mφs:splenocytes; 1:2, 1:8, 1:32) and T cell proliferation measured. Significance is compared to RC. C, Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bars) Mφs (1 Mφ: 2 splenocytes). Supernatants were collected after 72 h and concentrations of IFN‐γ (left), IL‐10 (middle), and IL‐2 (right) measured. Data are the mean ± SEM of 2 independent experiments assayed in triplicate (A, B) or duplicate (C). *,p< 0.05; **,p<0.01; ***,p<0.001. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc. 53  Figure 3.4 Mφs do not increase the proportion of dead T cells. Polyclonal‐ (Poly) or Ag‐specific‐ (OTI or OTII) stimulated splenocytes were cultured with (black bars) or without (white bars) Mφs (1 Mφ: 8 splenocytes) and after 72 h the proportion of dead (PI+) (left) or apoptotic (Annexin V+ PI‐) (right) CD4+ and CD8+ T cells assessed by flow cytometry. Data shown are the mean ± SEM of 2 independent experiments assayed in duplicate. *,p< 0.05; **,p<0.01; ***,p<0.001; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  3.2.4  Mφs suppress T cell proliferation via a contact­dependent mechanism  To elucidate the mechanism(s) by which Mφs were suppressing T cells we utilized a panel of blocking Abs, inhibitors, and exogenous cytokines. Each of the reagents tested were shown to be biologically active in our hands in several different systems (Alcon et al., 2009; Hamilton et al., 2010; Kuroda et al., 2009; Mace et al., 2009; Rauh et al., 2005; Sly et al., 2009; Sly et al., 2008, and unpublished results). We first determined whether cytokines that have been reported to alter Mφ phenotype (i.e. IL‐4, IL‐13) (Rodriguez et al., 2003) or restrict T cell proliferation (i.e. IL‐10, TGF‐β) (Levings and Roncarolo, 2000) were involved. Adding neutralizing Abs to IL‐4, IL‐10, IL‐13, or TGF‐β did not reverse Mφ suppression of T cell proliferation, indicating that these cytokines were not involved (Fig. 3.5A). To determine whether Mφs were acting as a sink for IL‐2 (von Bergwelt‐Baildon et al., 2006) and thus restricting the availability of IL‐2 for T cell proliferation, we added exogenous IL‐2 to the cultures, but this did not reverse suppression (Fig. 3.5A), which was consistent with our finding that the presence of Mφs increased IL‐2 levels (Fig. 3.3C, right). Since PGs such as PGE2 have 54  been implicated in Mφ suppression of T cells (Allison, 1978; Metzger et al., 1980), we tested the effect of blocking PGE2 synthesis. However, adding the COX‐2 inhibitor Celebrex to co‐cultures did not alter suppression (Fig. 3.5A), which suggested PGE2 was not involved. Next, we tested whether Mφs were suppressing T cell proliferation by an arginine dependent mechanism. The amino acid L‐Arg is necessary for T cell functions and arginine depletion inhibits T cell proliferation through a number of different mechanisms, including preventing re‐expression of the CD3ζ‐chain (Rodriguez et al., 2002; Rodriguez et al., 2003). However, adding exogenous L‐Arg did not abrogate Mφ suppression of T cells (Fig. 3.5A). Consistent with this result, inhibiting Arg1, which converts L‐Arg to urea and L‐ornithine, with the inhibitor BEC did not reverse suppression; in fact, BEC increased the suppressive effects of Mφs (Fig. 3.5A). In addition, we tested whether Mφs were suppressing T cell responses indirectly via Treg induction. However, Mφ co‐culture with activated T cells did not significantly increase the proportion of Tregs (Fig. 3.5B). Since none of the soluble‐mediated mechanisms we tested appeared to be involved in Mφ‐mediated immunosuppression, we asked whether the mechanism might be contact‐dependent. To test this, we assessed proliferation of activated splenocytes using a Transwell system. As shown in Fig. 3.5C, when Mφs and responder splenocytes were separated by a semi‐permeable membrane, T cell suppression was abrogated, suggesting that Mφs suppress T cell proliferation via a contact‐dependent mechanism.  55  Figure 3.5 Mφs suppress T cell proliferation via a contact­dependent mechanism. A, Mφs were co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes) ± 10 μg/ml rat IgG, 10 μg/ml anti‐IL‐4, 2 μg/ml anti‐IL‐10, 10 μg/ml anti‐IL‐13, 10 μg/ml anti‐ TGF‐β, 100 U/well mIL‐2, 20 μM Celebrex, 2 mM L‐arg, or 200 μM BEC. T cell proliferation was measured and percent of RC proliferation calculated. Significance is compared to control (‐) co‐cultures, where no inhibitors were added. B, Polyclonal‐stimulated splenocytes were cultured with (black bar) or without (white bar) Mφs (1 Mφ: 8 splenocytes). After 72 h, the proportion of Tregs (CD4+Foxp3+ cells) was analyzed by flow cytometry. C, Mφs were co‐ cultured with polyclonal‐stimulated splenocytes (1 Mφ: 2 splenocytes) in control (Mφs and splenocytes in contact) or Transwell (Mφs and splenocytes separated by a semi‐permeable membrane) 96‐well plates. T cell proliferation was measured and percent of RC proliferation calculated. Data is representative of three independent experiments performed in triplicate (A, C) or the mean ± SEM of two independent experiments performed in duplicate. *,p< 0.05; ***,p<0.001; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  56  3.2.5  Mφs suppress T cell proliferation via IFN­γ­induced NO production  Since direct contact between Mφs and T cells was required for effective suppression, we tested whether Mφs suppressed T cells either by activating CTLA‐4, a negative co‐stimulatory molecule found on T cells, by producing membrane‐bound TGF‐β, or via integrin‐mediated cell‐cell interactions. However, suppression was not reversed by adding a neutralizing Ab to CTLA‐4, LAP, to bind and neutralize bioactive TGF‐β at the cell membrane (Gandhi et al., 2007), or TIB218, to block CD11b/LFA‐1 (Fig. 3.6A, left). Next, we investigated whether ROS or RNS were involved. Although ROS and RNS are soluble mediators, they are extremely short‐lived and require cells to be in close proximity for their effects to be transmitted. Addition of the ROS inhibitors catalase, N‐acetyl‐L‐cysteine (NAC), and superoxide dismutase (SOD) did not reverse suppression; in fact, catalase and NAC increased the suppressive effects of Mφs (Fig. 3.6A, right). To elucidate whether RNS were playing a role, we first assayed for production of NO, a free radical gas that has cytostatic/cytotoxic effects (Bogdan, 2001). We found that Mφs produced NO in a dose‐dependent manner in the presence of activated T cells (Fig. 3.6B and C). The addition of iNOS inhibitors NG‐monomethyl‐L‐ arginine (L‐NMMA) and L‐NIL, or the NO scavenger PTIO, to co‐cultures effectively reduced Mφ NO production (Fig. 3.6D, left) and completely abrogated the inhibitory effect of Mφs on T cells (Fig. 3.6D, right), indicating that Mφs suppress T cell activation in large part via NO production.  57  Figure 3.6 Mφs suppress T cell proliferation through NO production. A, Mφs were co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes) ± 10 μg/ml anti‐CTLA4, 250 ng/ml LAP, or 5 μg/ml TIB218 (left), or ± 1 mg/ml catalase, 10 mM NAC, or 200 U/ml SOD (right). T cell proliferation was measured and percent of RC proliferation calculated. Significance is compared to control (‐) co‐cultures, where no inhibitors were added. B, Mφs were cultured alone or with polyclonal‐stimulated splenocytes at different ratios (Mφs:splenocytes; 1:4, 1:8, 1:16) for 72 h and NO production measured. C, Responder splenocytes (polyclonal‐stimulated C57BL/6 or OVA‐peptide‐ stimulated OTI or OTII) were cultured with (black bars) or without (white bars) Mφs at different ratios (Mφs:splenocytes; 1:4, 1:8, 1:16, 1:32). After 72 h, NO production was measured. D, Mφs were co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes) ± 0.5 mM L‐NMMA, 1 mM L‐NIL, or 25 μg/ml PTIO. NO production (left) and T cell proliferation were measured and percent of RC proliferation calculated (right). Significance is compared to control (‐) co‐cultures, where no inhibitors were added. Data are representative of three to five independent experiments performed in triplicate. *,p< 0.05; **,p<0.01; ***,p<0.001; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  58  Since Mφs did not express iNOS (Fig. 3.2) or produce NO (Fig. 3.6B) in the absence of activated T cells, we hypothesized that a factor produced by activated T cells was required to induce Mφ NO production. Since IFN‐γ is known to induce iNOS in Mφ via Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signaling (Schroder et al., 2004), we added an IFN‐γ‐neutralizing Ab during Mφ co‐culture with stimulated T cells and found this eliminated Mφ NO production (Fig. 3.7, left) and restored T cell proliferative ability (Fig. 3.7, right). Taken together, these data indicate that IFN‐γ, produced by activated T cells, induces naïve, resident Mφs to express iNOS and produce NO, which negatively regulates T cell responses.  Figure 3.7 IFN­γ is required for Mφ NO production. Mφs were co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes) ± 10 μg/ml anti‐IFN‐γ and NO production (left) and T cell proliferation measured and percent of RC proliferation calculated (right). Data are representative of three to five independent experiments performed in triplicate. ***,p<0.001. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  3.2.6  Pre­treatment with LPS and dsRNA, but not CpG or PGN, decreases the ability of Mφs to produce NO and suppress T cells  Our results suggested that in the absence of a pathogen signal, resident Mφs become potent suppressors of T cell proliferation upon stimulation by T cell‐derived IFN‐γ. Consequently, we wanted to determine whether Mφs matured with TLR ligands  59  would exhibit the same suppressive phenotype. Related to this, we found that although LPS, CpG, dsRNA, and PGN were equally potent at stimulating the maturation of Mφs, i.e., increasing Mφ expression of MHCII, CD40, and CD86 to a similar extent (Fig. 3.8A; LPS and CpG, and data not shown), they had different effects on Mφ immunosuppression. Pre‐treatment with LPS or dsRNA, which signal through the adaptor TRIF (Akira and Takeda, 2004), reduced the ability of Mφs co‐cultured with activated T cells to upregulate iNOS expression (Fig. 3.8B), produce NO (Fig. 3.8C), or suppress T cell proliferation (Fig. 3.8D). In contrast, Mφs pre‐treated with CpG or PGN, which signal through the adaptor MyD88 (Akira and Takeda, 2004), behaved like control Mφs, expressing iNOS (Fig. 3.8B), producing NO (Fig. 3.8C), and significantly inhibiting T cell proliferation (Fig. 3.8D). It is worth noting that although LPS and dsRNA pre‐treatment reduced Mφ iNOS expression upon exposure to activated T cells, the level of STAT1 phosphorylation did not change (Fig. 3.8B), indicating that JAK/STAT signaling was not reduced. Since previous reports suggested that pre‐treatment of Mφs with sub‐lethal doses of LPS can cause a transient state of hypo‐responsiveness to subsequent exposure to LPS or other pro‐inflammatory mediators (Lorsbach and Russell, 1992), we hypothesized that pre‐treatment with LPS might render Mφs less responsive to subsequent IFN‐γ stimulation. To test this, we pre‐treated Mφs with or without LPS or CpG and then stimulated the cells with or without IFN‐γ + LPS. As shown in Fig. 3.9, Mφs that had been pre‐treated with LPS produced significantly less NO than CpG or non‐pre‐treated Mφs, indicating that pre‐treatment with LPS, but not CpG, reduces the ability of Mφs to respond to IFN‐γ.  60  Figure 3.8 Pre­treatment with TLR agonists that signal through TRIF desensitizes Mφs to subsequent IFN­γ stimulation, decreasing the ability of Mφs to produce NO and suppress T cells. A, Mφs were treated with medium alone (grey fill), LPS (black line), or CpG (grey line). After 24 h, cells were harvested and analyzed by flow cytometry for surface expression of MHC class II, CD40, and CD86. Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bars) Mφs that had been pre‐treated ± LPS, CpG, dsRNA, or PGN for 24 h and washed prior to co‐culture (1 Mφ: 8 splenocytes). After 72 h, Mφs were subjected to B, Western blot analysis, and C, NO production and D, T cell proliferation measured. Data shown are representative of 4 independent experiments, performed in duplicate (A) or the mean ± SEM of 3 independent experiments, performed in duplicate (B, C, D). *,p< 0.05; **,p<0.01; ***,p<0.001; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  61  Figure 3.9 Pre­treatment with LPS, but not CpG, reduces Mφ sensitivity to subsequent stimulation. Mφs were pre‐treated ± LPS or CpG for 24 h, washed, and then stimulated ± IFN‐γ+LPS. After 72 h, NO production was measured. Data are the mean ± SEM of 3 independent experiments, performed in duplicate. ***,p<0.001; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  3.2.7  IFN­β contributes to the reduced suppressive abilities of LPS and dsRNA pre­treated Mφs  To gain insight into how LPS or dsRNA pre‐treatment reduced the ability of Mφs to respond to IFN‐γ and suppress T cell responses, we tested whether the effect of LPS was dependent on a secondary, secreted factor. To perform these studies we utilized polymyxin B, a compound that binds to and inactivates LPS (Morrison and Jacobs, 1976) and effectively blocks the effect of LPS on Mφ NO production (Fig. 3.10A). We treated Mφs with or without LPS for 24 h, and then added polymyxin B to the supernatants before transferring them onto unstimulated Mφs. After 24 h, the Mφs were washed and co‐cultured with activated T cells. As shown in Fig. 3.10B, Mφs exposed to supernatants from LPS‐treated cells had decreased NO production when compared to Mφs treated with unstimulated supernatants. These data suggest that LPS stimulation induces Mφs to secrete a secondary factor that acts in an autocrine manner to reduce Mφ NO production in response to IFN‐γ from activated T cells. 62  Figure 3.10 LPS stimulation induces production of a secondary factor that inhibits IFN­γ­induced NO production by Mφs. A, Mφs were pre‐treated with medium alone (‐), LPS, or LPS + polymyxin B (PB) for 24 h, washed, and co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes). After 72 h, NO production was measured. Significance is compared to non‐pre‐treated Mφs (‐). B, Mφs were stimulated ± LPS for 24 h and the supernatants were supplemented with polymyxin B (PB), and then added to unstimulated Mφs. After 24 h, Mφs were washed and co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes) for 72 h and NO production measured. Data are representative of three independent experiments. *,p< 0.05; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  We then asked if this secondary factor could be IFN‐β since activation of the TRIF pathway has been shown to produce this cytokine (Akira and Takeda, 2004). Indeed, we found that stimulation with LPS and dsRNA, but not CpG or PGN, induced Mφs to produce IFN‐β (Fig. 3.11A). Therefore, we tested whether IFN‐β was involved in the abrogation of T cell suppression by LPS and dsRNA and found that IFN‐β pre‐ treatment of Mφs dose‐dependently reduced iNOS expression (Fig. 3.11B, left), NO production (Fig. 3.11B, middle) and Mφ immunosuppression (Fig. 3.11B, right). Adding a neutralizing Ab to IFN‐β significantly increased NO production (Fig. 3.11C, left) and suppression of T cell proliferation (Fig. 3.11C, right) by LPS‐ and dsRNA‐treated Mφs. Moreover, neutralizing IFN‐β also increased iNOS expression in Mφs exposed to supernatants from LPS‐stimulated Mφs treated with polymyxin B (Fig. 3.11D). Taken  63  together, these results suggest that the induction of autocrine‐acting IFN‐β plays an important role in abrogating the immunosuppressive abilities of Mφs.  Figure 3.11 IFN­β contributes to the reduced suppressive abilities of LPS and dsRNA pre­treated Mφs. A, Mφs were stimulated ± LPS, CpG, dsRNA, or PGN for 24 h. Supernatants were collected and IFN‐β levels measured. Significance is compared to non‐pre‐treated Mφs (‐). B, Polyclonal‐ stimulated splenocytes were cultured with (black bars) or without (white bars) Mφs that had been pre‐treated with IFN‐β (0, 100, 1000, 5000 U/ml), LPS, or dsRNA for 24 h and washed prior to co‐culture (1 Mφ: 8 splenocytes). After 72 h, Mφs were subjected to Western blot analysis (left), and NO production (middle) and T cell proliferation (right) measured. Significance is compared to untreated Mφs. C, Mφs were pre‐treated with medium alone (‐), LPS, or dsRNA in the presence (black bars) or absence (white bars) of neutralizing anti‐IFN‐β (1000 U/ml) for 24 h, washed, and co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes). After 72 h, NO production (left) and T cell proliferation (right) were measured. D, Mφs were stimulated with medium alone (‐), LPS, or LPS + anti‐IFNβ (500 U/ml) for 24 h. Supernatants were collected, treated with polymyxin B, and then added to unstimulated Mφs. After 24 h, Mφs were washed and co‐cultured with polyclonal‐stimulated splenocytes (1 Mφ: 8 splenocytes) for 72 h and then subjected to Western blot analysis. Data are representative of three to five experiments performed in duplicate (A, D) or triplicate (B, C). *,p< 0.05; **,p<0.01; ***,p<0.001; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  64  3.2.8  Inhibition of Arg1 reduces the effect of LPS and dsRNA pre­treatment of Mφs  We observed that, in addition to inhibiting iNOS expression (Fig. 3.11B), LPS and, to a lesser extent, dsRNA increased the expression of Arg1 in Mφs (Fig. 3.11D and data not shown). Thus, we investigated whether Arg1 was playing a role in LPS‐ and dsRNA‐induced abrogation of Mφ immunosuppression. As shown in Fig. 3.12, inhibiting Arg1 with the Arg‐specific inhibitor BEC increased NO production (left) and suppression of T cells (right) by LPS‐ and dsRNA‐treated Mφs, but, as expected, did not alter iNOS expression (data not shown). This is consistent with the observation that increasing Arg1 expression or activity can decrease NO production, given that iNOS and Arg1 compete for the common substrate, L‐Arg (Gordon, 2003). Therefore, LPS and dsRNA decrease the ability of Mφs to produce NO in response to T cell‐derived IFN‐γ by both decreasing iNOS and increasing Arg1 expression.  Figure 3.12 The Arg1 inhibitor BEC reduces the effects of LPS and dsRNA pre­ treatment. Mφs were pre‐treated with medium alone (‐), LPS, or dsRNA in the presence (black bars) or absence (white bars) of 200 μM BEC for 24 h, washed, and co‐cultured with polyclonal‐ stimulated splenocytes (1 Mφ: 8 splenocytes). After 72 h, NO production (left) and T cell proliferation (right) were measured. Data are representative of 3 independent experiments performed in triplicate. *,p< 0.05; NS, not significant. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  65  3.3 Discussion One of the most important features of the immune system is its ability to discriminate between Ags that should provoke an immune response and those that should be tolerated. It is well known that APCs, such as Mφs and DCs, regulate the ability of the adaptive immune system to respond to Ags. However, the specific mechanisms by which this occurs have not been fully elucidated (Munn and Mellor, 2003). In this chapter, we investigated the process by which resident PMφs suppress T cell proliferation and determined how the presence of pathogen‐derived molecules influences Mφ immunosuppression. Our results more clearly elucidate the mechanisms by which Mφs exert their immunosuppressive effects and suggest a key role for TLR signaling and IFN‐β in regulating Mφ phenotype. Resident PMφs, but not other cells within the PC, potently suppress T cell proliferation via an IFN‐γ and NO‐dependent mechanism. As shown in the model in Fig. 3.13, IFN‐γ produced by TCR‐activated T cells induces naïve resident Mφs to express iNOS and produce NO, which subsequently suppresses T cell activation. However, this negative feedback loop is eliminated by TLR agonists that signal through the TRIF cascade (i.e. LPS or dsRNA), but not by those that signal exclusively through MyD88 (i.e. CpG or PGN). Mφs stimulated with LPS or dsRNA exhibit low responsiveness to T cell‐ produced IFN‐γ and, consequently, do not upregulate iNOS, produce NO, or exhibit immunosuppressive properties. The reduced ability of LPS or dsRNA pre‐treated Mφs to respond to IFN‐γ is due, at least in part, to the production of autocrine‐acting IFN‐β, which occurs following TRIF recruitment and IRF‐3 activation (Akira and Takeda, 2004). This, together with induction of Arg1, reduces NO production by Mφs, allowing T cells to proliferate and mount an immune response. Although the inhibitory effects of LPS and dsRNA on IFN‐γ responsiveness and iNOS expression are mediated by IFN‐β, the ability of LPS and dsRNA to induce Arg1 does not appear to be IFN‐β‐dependent, since blocking IFN‐β did not reduce Arg1 expression (Fig. 3.11D). Related to this, there is some evidence that Arg1 may be induced via the NF‐κB pathway (Hagemann et al., 66  2008), which is activated by all TLR signaling pathways, although more rapidly via the MyD88‐dependent pathway (Akira and Takeda, 2004). However, our data suggest that increasing Arg1 alone is not sufficient to reduce NO production and Mφ immunosuppression, but that increased Arg1 expression is only capable of decreasing NO production in combination with decreased iNOS expression.  Figure 3.13 The ability of resident Mφs to suppress T cell proliferation is abrogated when Mφs are pre­treated with LPS or dsRNA, but not CpG or PGN. Resident peritoneal Mφs, or Mφs pre‐treated with CpG or PGN, suppress T cell proliferation via IFN‐γ‐induced NO production. IFN‐γ produced by T cells following TCR stimulation binds to its receptor on resident Mφs and recruits two JAKs (JAK‐1 and JAK‐2), which trigger the recruitment, phosphorylation, and activation of the transcription factor STAT1. The phosphorylated STAT1 homodimer translocates into the nucleus and binds to IFN‐γ‐ activation sites (GAS) on the iNOS promoter, inducing iNOS expression and NO production, which in turn suppresses T cell activation. However, Mφs treated with LPS or dsRNA lack immunosuppressive properties. Exposure of resident peritoneal Mφs to LPS or dsRNA triggers TRIF recruitment, IRF‐3 activation, and induction of IFN‐β, and later, IFN‐inducible genes. When these pre‐treated Mφs subsequently detect IFN‐γ produced by activated T cells they fail to upregulate iNOS or produce NO, and consequently exhibit significantly reduced immunosuppressive abilities as compared to naïve, or CpG or PGN pre‐treated, Mφs. This failure of LPS or dsRNA pre‐treated Mφs to respond to IFN‐γ is, at least in part, due to autocrine‐acting IFN‐β. This, together with induction of Arg1, reduces NO production by Mφs and eliminates Mφ immunosuppression. Reprinted from Hamilton et al., 2010 with permission of The American Association of Immunologists, Inc.  67  There is some controversy in the literature regarding the effect of IFN‐β on IFN‐γ signaling (Gough et al., 2010; Inaba et al., 1986; Ling et al., 1985; Schroder et al., 2004; Yoshida et al., 1988). The IFN‐γ and IFN‐α/β signal pathways are partially overlapping, which allows for cross‐talk at multiple levels (Schroder et al., 2004). IFN‐γ binds to its receptor, which triggers receptor clustering and activates a JAK‐STAT signaling pathway that culminates in the binding of a phosphorylated STAT1 homodimer to promoter elements. This, in turn, activates or suppresses transcription of IFN‐γ‐ regulated genes (Schroder et al., 2004). There is evidence that type I IFNs can prime cells and increase responsiveness to IFN‐γ by inducing IFN‐γ signaling molecules, such as STAT1 (Gough et al., 2010; Schroder et al., 2004). On the other hand, our finding that IFN‐β antagonizes the response of Mφs to subsequent treatment with IFN‐γ is consistent with several reports (Inaba et al., 1986; Ling et al., 1985; Yoshida et al., 1988) including a recent study suggesting the mechanism may be down‐regulation of the IFN‐γR in response to IFN‐ α/β (Rayamajhi et al., 2010). However, since we did not see a decrease in STAT1 phosphorylation following IFN‐β treatment (data not shown), down‐ regulation of the IFN‐γR may not be consistent with our findings. IFN‐γ is a critical regulator of Mφ phenotype and function. It is well known that IFN‐γ stimulation triggers Mφs to acquire anti‐microbial and pro‐inflammatory properties, including the ability to promote T cell proliferation (Schroder et al., 2004). However, as our data demonstrate, IFN‐γ can also trigger Mφs to become immunosuppressive. Thus, IFN‐γ‐induced Mφs have the potential to both stimulate and suppress T cell responses and the dominating effect likely depends on the ratio of Mφs to T cells and the resulting local concentration of NO. When Mφs are present in low numbers, little NO is produced and the immunostimulatory effects of Mφs predominate (Fig. 3.3A, left). However, when the ratio of Mφs to T cells is high, IFN‐γ‐induced Mφs exhibit anti‐inflammatory properties, since the resulting NO is sufficient to suppress local T cell proliferation. This situation is physiologically relevant both in sites of acute infection (e.g. wounds) and in non‐infected tissues, where Mφs are more abundant than lymphocytes (Gordon, 2003; Martin and Muir, 1990). This is supported by a recent 68  report (Composto et al., 2011) demonstrating that although resident PC T cells exhibit a phenotype indicative of activation, they are suppressed by resident PMφs, which outnumber T cells and suppress via an IFN‐γ and iNOS‐dependent mechanism. The co‐existence of Mφs and activated T cells in tissues such as the PC (Composto et al., 2011) suggests Mφs play an important role in maintaining immune homeostasis. Our results, which indicate that TLR signaling can regulate the immunosuppressive properties of Mφs, add to our understanding of how this homeostasis is maintained and are consistent with the role of TLRs in pathogen detection. In the absence of PAMPs, it is likely that Mφs that detect T cell‐produced IFN‐γ sense the response as inappropriate and exert immunosuppressive effects. On the other hand, if a pathogen is present, predisposition of Mφs to allow T cells to mount an immune response would be advantageous. However, our data suggest that not all PAMPs negate Mφ suppressive abilities. Specifically, while TLR4 (LPS) and TLR3 (dsRNA) agonists abrogated Mφ immunosuppression, TLR9 (CpG) and TLR2 (PGN) agonists did not (Fig. 3.8). Although the physiological rationale for these differences is not immediately apparent, we hypothesize that the abrogation of Mφ immunosuppression may be specific to viral stimuli, given that only PAMPs that trigger the induction of IFN‐β (Fig. 3.11A), a critical mediator of the anti‐viral immune response (Schroder et al., 2004), eliminated Mφ immunosuppression (Fig. 3.8). Bacterial stimuli, on the other hand, did not reduce NO production by Mφs, which may be advantageous given the ability of NO to kill extracellular pathogens, such as certain classes of bacteria (Bogdan, 2001). As well, there is indirect in vitro (Taylor‐Robinson et al., 1994) and in vivo (Wei et al., 1995) evidence that Th1 cells, which target intracellular pathogens (i.e. viruses) may be more susceptible to the inhibitory effects of NO than Th2 cells, which promote attacks on extracellular pathogens (e.g. extracellular bacteria), and this may suggest a potential rationale for our findings. Further studies to address the physiological effects of different pathogens will provide additional insights into the regulation of Mφ phenotype by TLR agonists.  69  The central role that Mφs play in regulating T cell activation makes them attractive therapeutic targets. Strategies that promote Mφ immune suppression could reduce autoimmune disease and enhance transplantation tolerance. Alternatively, therapies that reduce the suppressive effects of Mφs, such as treatment with iNOS inhibitors, IFN‐β, LPS, or dsRNA could boost immune responses and lead to more effective treatments for infection and cancer. Furthermore, given the current debate concerning the most effective vaccine adjuvants (McKee et al., 2007) and the contrasting effects of different TLR agonists on the immunosuppressive functions of Mφs, and potentially DCs, agonists that induce IFN‐β, such as LPS or dsRNA, may be more effective than CpG at stimulating immune responses. In summary, the ability of Mφs to contribute to both immune response and tolerance highlights their considerable therapeutic potential, as well as the importance of increasing our understanding of the factors that regulate Mφ function.  70  CHAPTER 4 : SERUM INHIBITS THE IMMUNOSUPPRESSIVE FUNCTION OF MYELOID­DERIVED SUPPRESSOR CELLS ISOLATED FROM 4T1 TUMOR­BEARING MICE 4.1 Introduction After investigating the role that Mφs play in regulating peripheral tolerance under physiological conditions, we next wanted to compare the roles of Mφs and other myeloid cell types, especially MDSCs, under neoplastic conditions. MDSCs are a heterogeneous population of myeloid cells that expand and become activated in response to tumor‐derived factors in both cancer patients and tumor‐bearing mice (Nagaraj and Gabrilovich, 2010). MDSCs can contribute to cancer progression by a number of different mechanisms including suppression of T cell‐mediated immune responses (Priceman et al., 2010; Sinha et al., 2005b; Yang et al., 2004; Yang et al., 2008; Youn and Gabrilovich, 2010). The cell surface phenotype of MDSCs is not clearly elucidated and the heterogeneity of MDSCs has made both their identification and isolation difficult (Nagaraj and Gabrilovich, 2010). Consequently, both human and murine MDSCs must be identified functionally, which requires assays to accurately assess MDSC immunosuppression ex vivo. A number of different culture systems have been used to investigate the properties of cancer‐induced MDSCs. However, the influence of different culture conditions on the ability of MDSCs to suppress T cell responses ex vivo remains unknown. Given that MDSCs are functionally defined, accurate assessment of MDSC immunosuppression is extremely important for studying the role of MDSCs in tumor progression. One murine cancer model often employed to study MDSC‐induced suppression involves orthotopic injection of 4T1 mammary carcinoma cells that metastasize to the lungs and other tissues (Aslakson and Miller, 1992; Heppner et al., 2000). One of the primary differences in the culture conditions used by different groups studying 4T1‐induced MDSCs is the presence (Ko et al., 2010) or absence (Srivastava et 71  al., 2010) of serum in ex vivo immunosuppression assays. In this chapter we demonstrate that MDSCs from 4T1 tumor‐bearing mice effectively suppress T cell proliferation under serum‐free conditions, but fail to do so in the presence of FCS. Furthermore, we show that the major serum protein albumin mediates the effect of FCS on 4T1‐induced MDSCs by inhibiting their ability to produce ROS. Our data indicate that different culture conditions can profoundly alter MDSC phenotype. These findings have important implications for investigating the role of MDSCs in cancer, since accurate assessment of MDSC immunosuppression is essential for identifying tumor‐induced MDSCs and for studying the functional roles of MDSCs in primary and metastatic tumor growth.  This thesis chapter is based on work presented in a manuscript currently in press (Hamilton et al., 2011). 4.2 Results 4.2.1  Validation of in vitro T cell proliferation assay systems  In order to test our hypothesis that the immunosuppressive abilities of MDSCs are influenced by different culture conditions, we designed assay systems to assess the ability of MDSCs to suppress polyclonal‐ and Ag‐specific‐stimulated T cell proliferation. As expected, splenic T cells isolated from BALB/c mice proliferated only in response to polyclonal‐stimulation, while splenic T cells from BALB/c DO11.10 mice proliferated in response to both polyclonal‐ and Ag‐specific‐stimulation (Fig. 4.1A). In the latter system, polyclonal‐stimulation induces the proliferation of all T cells and is therefore a more potent stimulus than Ag‐driven‐stimulation, which specifically activates CD4+ T cells. Initial tests also revealed that polyclonal‐stimulated T cells could proliferate to an equal extent whether or not 10% FCS was present, while Ag‐specific‐stimulated T cells proliferated more in the presence of 10% FCS (Fig. 4.1B).  72  Figure 4.1 Polyclonal­ and Ag­specific­stimulated T cells proliferate in both serum­ free and serum­containing conditions. A, BALB/c (left) or DO11.10 (right) splenocytes were stimulated with anti‐CD3/anti‐CD28 (polyclonal) or OVA peptide (Ag‐specific) in serum‐free HL1 medium and T cell proliferation measured. No stim = unstimulated splenocytes. B, Splenocytes were polyclonal‐ (left) or Ag‐ specific‐ (right) stimulated in HL1 medium ± 10% FCS and T cell proliferation measured. Data are the mean ± SEM of three (a) or five (b) independent experiments. ***,p<0.001, NS, not significant  4.2.2  4T1­induced MDSCs only suppress T cell proliferation under serum­ free conditions  We isolated MDSCs from the spleens of 4T1 mice based on Gr1 positive selection and established that the recovered cells were >95% CD11b+Gr1+ and contained few F4/80+ Mφs (Fig. 4.2A). We then investigated the ability of 4T1‐induced MDSCs to suppress T cell proliferation and found that MDSCs significantly suppressed both 73  polyclonal‐ and Ag‐specific‐stimulated T cell proliferation in serum‐free medium, but were no longer immunosuppressive when 10% FCS was present (Fig. 4.2B). To ensure that these results were not due to a specific source of FCS we tested four different sources of FCS and found that they all completely prevented MDSC‐induced suppression of T cells, and did so to an equal extent (data not shown). We also tested the effect of mouse serum in our assay, but found that polyclonal‐stimulated T cells were unable to proliferate in medium containing 10% mouse serum (data not shown).  Figure 4.2 MDSCs from 4T1 tumor­bearing mice only suppress T cell proliferation under serum­free conditions. A, MDSCs were isolated from the spleens of 4T1 tumor‐bearing mice by Gr1 positive selection and surface expression of Gr1, CD11b, and F4/80 assayed. B, Polyclonal‐ (left) or Ag‐specific‐ (right) stimulated splenocytes were cultured with (black bars) or without (white bars) 4T1‐induced MDSCs (1 MDSC: 1 splenocyte) in HL1 medium ± 10% FCS and T cell proliferation measured. Data are representative of five independent experiments. ***,p<0.001  74  4.2.3  Serum does not increase the viability of 4T1­induced MDSCs  We next addressed whether the inability of MDSCs to suppress T cell proliferation in the presence of serum was due to an effect of serum on the MDSCs or on the T cells. First, we pre‐incubated MDSCs O/N with or without 10% FCS, washed the cells, and performed morphological analysis to determine the proportion of M‐ versus G‐MDSC subsets (Movahedi et al., 2008). We found that although M‐MDSCs were present in both serum‐free and serum‐containing conditions, the majority of cells were G‐MDSCs under both conditions, and that the presence or absence of serum did not alter this ratio (Fig. 4.3A). We also assayed cell viability and found that culturing MDSCs in medium containing 10% FCS slightly reduced cell viability and increased the proportion of non‐adherent cells (Fig. 4.3B).  Figure 4.3 Serum does not alter the proportion of different MDSC subtypes. MDSCs were cultured overnight (O/N) in HL1 medium ± 10% FCS and A, cell morphology (20x magnification), and B, viability and adherence were determined. Data are representative of three independent experiments.  75  4.2.4  Serum directly inhibits 4T1­induced MDSC immunosuppression  Next, we tested the ability of these pre‐incubated MDSCs to suppress T cell proliferation under serum‐free conditions. MDSCs pre‐incubated with 10% FCS displayed significantly reduced immunosuppressive abilities under both polyclonal and Ag‐specific conditions (Fig. 4.4A). Importantly, serum did not compromise the ability of T cells to be suppressed since TAMs (Fig. 4.4B) and PMφs (Fig. 4.4C) from 4T1 tumor‐ bearing mice potently suppressed T cell proliferation both in the presence and absence of serum. Taken together, these results indicate that FCS directly ameliorates the immunosuppressive properties of 4T1‐induced MDSCs. 4.2.5  FCS inhibits MDSC immunosuppression in a dose­dependent manner  To understand how serum influences the immunosuppressive phenotype of MDSCs, we first assayed the ability of 4T1‐induced MDSCs to suppress T cell proliferation in the presence of different concentrations of FCS. As shown in Fig. 4.5, serum inhibited MDSC immunosuppression in a dose‐dependent manner. While pulmonary and splenic 4T1‐induced MDSCs strongly suppressed T cell proliferation in serum‐free or 1% FCS‐containing medium, they lost all immunosuppressive function when assayed in 3% or 10% FCS‐containing medium (Fig. 4.5).  76  Figure 4.4 Serum directly inhibits the immunosuppressive abilities of MDSCs. A, Splenic MDSCs were cultured O/N in serum‐free (white bars) or 10% FCS‐containing (black bars) HL1 medium, washed, and co‐cultured at different ratios with polyclonal‐ (left) or Ag‐specific‐ (right) stimulated splenocytes and T cell proliferation measured. B, TAMs or C, PMφs were co‐cultured (1 Mφ: 2 splenocytes) with polyclonal‐ (left) or Ag‐specific‐ (right) stimulated splenocytes in HL1 medium ± 10% FCS and T cell proliferation measured. Data are representative of three independent experiments. *,p<0.05; **,p<0.01; ***,p<0.001  77  Figure 4.5 The inhibitory effects of serum on 4T1­induced MDSC immunosuppression are dose­dependent. Polyclonal‐stimulated splenocytes were co‐cultured with (black bars) or without (white bars) pulmonary (left) or splenic (right) MDSCs (1 MDSC: 1 splenocyte) in HL1 medium ± 1, 3, or 10% FCS and T cell proliferation measured. Data are the mean ± SEM of two independent experiments. ***,p<0.001; NS, not significant  4.2.6  Effect of different serum treatments on MDSC immune suppression  Next, we tested the ability of MDSCs to suppress T cells in the presence of FCS that had been filtered to remove components >0.22 µm or dialyzed to remove components <3.5 kDa. Neither treatment restored the immunosuppressive abilities of pulmonary or splenic 4T1‐induced MDSCs, although dialysis may have slightly increased the ability of pulmonary MDSCs to suppress T cell proliferation (Fig. 4.6A). Similarly, heat‐inactivating the serum to destroy complement activity did not restore MDSC immunosuppression (Fig. 4.6B). Our experiments were performed using HL1 medium and clearly demonstrate the inhibitory effect of serum on MDSC immune suppression. There are some reports in the literature that splenic 4T1‐induced MDSCs are immunosuppressive when assayed in RPMI 1640 medium containing 10% FCS (Ko et al., 2010; Kodumudi et al., 2010; Le et al., 2009). However, in our hands, both splenic and pulmonary MDSCs were unable to suppress T cell proliferation in RPMI 1640 medium containing 10% FCS and, in fact, pulmonary 4T1‐induced MDSCs appeared to activate T cell proliferation under these conditions (Fig. 4.6C). Similarly, 4T1‐induced  78  MDSCs failed to suppress T cell proliferation when cultured in IMDM medium containing 10% FCS (data not shown).  Figure 4.6 The effects of serum on 4T1­induced MDSCs cannot be reversed by filtration, dialyzation, or heat inactivation of the serum. Polyclonal‐stimulated splenocytes were co‐cultured with (black bars) or without (white bars) pulmonary (left) or splenic (right) MDSCs (1 MDSC: 1 splenocyte) in A, HL1 medium ± 10% untreated, filtered, or dialyzed FCS, or B, HL1 medium ± 10% heat‐inactivated FCS, or C, RPMI 1640 medium ± 10% untreated FCS, and T cell proliferation measured. Data are the mean ± SEM of two independent experiments. *,p<0.05; ***,p<0.001; NS, not significant 79  4.2.7  BSA blunts 4T1­induced MDSC immunosuppression  Since the antagonism by serum of MDSC‐induced immune suppression could not be ameliorated by filtration, dialysis, or heat‐inactivation of the serum, we considered whether any of the key protein components of serum could be playing a role. BSA is the most abundant protein in FCS (Seifert and Resman‐Targoff, 2006) and is commonly added to serum‐deprived cultures to support cell growth (Monette and Sigounas, 1990). We tested the effect of BSA on 4T1‐induced MDSC immunosuppression and found that BSA decreased the ability of both pulmonary and splenic MDSCs to suppress T cell proliferation in a dose‐dependent manner (Fig. 4.7A). In the absence of BSA, pulmonary and splenic MDSCs suppressed 96.9% and 98.3% of T cell proliferation, respectively. This suppression was reduced to 51.4% and 43.6% with 1% BSA, and to 29.8% and 21.6% when 3% BSA was added to the assay. Soluble BSA is 400 x 1400 nm in size with a molecular mass of 66.4 kDa and, therefore, would not have been removed by either filtration (>0.22 µm) or dialysis (<3.5 kDa) of FCS, which is consistent with our results (Fig.4.6A). We also tested the effect of human albumin and a second source of BSA and found that both significantly reduced the suppressive functions of 4T1‐induced MDSCs (data not shown). To verify that the effect of serum on MDSC immunosuppression was due to BSA, we removed BSA from FCS using Affi‐Gel Blue bead affinity chromatography and compared the ability of MDSCs to suppress T cell proliferation in serum with or without BSA. At all concentrations of serum tested, removing BSA restored the ability of 4T1‐induced MDSCs to suppress T cells (Fig. 4.7B), indicating that BSA inhibits MDSC immunosuppression. Since BSA contains both high and low affinity binding sites for fatty acids (FAs), and thus is an efficient source of lipids (Fasano et al., 2005; Roche et al.,  2008),  we  tested  whether  BSA‐associated  FAs  were  affecting  MDSC  immunosuppression. Interestingly, we found that MDSC suppression was restored when the assay was performed in medium containing FA‐free BSA (Fig. 4.7C), suggesting that BSA‐associated FAs may be involved in inhibiting MDSC immune suppression.  80  Figure 4.7 BSA reduces the immunosuppressive abilities of 4T1­induced MDSCs. A, Polyclonal‐stimulated splenocytes were co‐cultured with (black bars) or without (white bars) pulmonary (left) or splenic (right) MDSCs (1 MDSC: 1 splenocyte) in HL1 medium ± 0.3, 1, or 3% BSA and T cell proliferation measured. B, Polyclonal‐stimulated splenocytes were co‐cultured with (black bars) or without (white bars) pulmonary (left) or splenic (right) MDSCs (1 MDSC:1 splenocyte) in HL1 medium ± 1, 3, or 10% FCS that was untreated or that had been depleted of BSA and T cell proliferation measured. C, Polyclonal‐stimulated splenocytes were co‐cultured with (black bars) or without (white bars) pulmonary (left) or splenic (right) MDSCs (1 MDSC:1 splenocyte) in HL1 medium containing 3% BSA ± fatty‐ acids (FAs) and T cell proliferation measured. Data are the mean ± SEM of two independent experiments. *,p<0.05; **,p<0.01; ***,p<0.001; NS, not significant  81  4.2.8  BSA antagonizes 4T1­induced MDSC immunosuppression by inhibiting ROS production  Next, we embarked on a series of studies to investigate the mechanism(s) by which BSA inhibits MDSC‐induced immune suppression. MDSCs have been reported to suppress T cell proliferation via a number of different mechanisms including expression of iNOS and/or Arg1, production of inhibitory cytokines (i.e. TGF‐β, IL‐10), and generation of ROS species (Youn and Gabrilovich, 2010). Therefore, we investigated whether BSA was interfering with any of these processes. We cultured MDSCs O/N in serum‐free or 10% FCS or 3% BSA containing medium (as our previous studies had demonstrated this was sufficient to reduce the suppressive function of MDSCs (Fig. 4.4A)). Western blot analysis of cell lysates revealed that both pulmonary and splenic 4T1‐induced MDSCs failed to express either iNOS or Arg1 under any of these culture conditions (data not shown). Furthermore, the presence of FCS or BSA did not influence the production of TGF‐β, which was below the detection limit in all samples, or IL‐10, which was produced at low, but similar, levels in all samples (data not shown). Next, we tested the effect of FCS or BSA on the ability of 4T1‐induced MDSCs to generate ROS and found that culturing MDSCs O/N in FCS or BSA reduced ROS production to a large extent (Figure 4.8). Specifically, pulmonary MDSCs cultured in medium containing FCS or BSA produced 3.6 fold and 6.5 fold less ROS, respectively. Similarly, splenic MDSCs produced 5.2 and 6.5 fold less ROS when cultured in the presence of FCS or BSA, respectively. Taken together, our results indicate that the immunosuppressive properties of MDSCs are greatly influenced by different culture conditions and that BSA contained in FCS blunts the immunosuppressive properties of 4T1 tumor‐derived MDSCs, at least in part, by restricting ROS production. These findings are vital for the appropriate design and interpretation of immunosuppression assays and may shed light on the physiological roles of MDSCs in tumor biology.  82  Figure 4.8 Serum albumin restricts ROS production by 4T1­induced MDSCs. A, Pulmonary (left) or splenic (right) MDSCs were cultured O/N in HL1 medium containing no serum (grey fill), 10% FCS (black line) or 3% BSA (grey line) and ROS production measured using DCFDA dye. B, Mean fluorescent intensity (MFI) of MDSCs cultured O/N in HL1 medium containing no serum (grey fill), 10% FCS (black line) or 3% BSA (grey line) and labeled with DCFDA dye. Data are (A) representative or (B) the mean ± SEM of two independent experiments.  83  4.3 Discussion Although MDSCs are defined by their immunosuppressive properties, our work demonstrates that the ability of MDSCs to suppress T cell responses can be influenced by different culture conditions, such as the presence or absence of serum. Specifically, while 4T1‐induced MDSCs potently suppressed T cell proliferation under serum‐free conditions, they were not suppressive when 10% FCS was included in the assay. Our data identify BSA as the principal factor responsible for the effect of serum on 4T1‐ induced MDSCs and are consistent with a model in which serum albumin acts directly on MDSCs to decrease their immunosuppressive properties by decreasing ROS production. Moreover, we have some evidence that BSA‐associated FAs may play a role in regulating MDSC suppression. These findings have important implications for researchers investigating the role of MDSCs in cancer, as the accurate detection and quantification of immunosuppression is critical for both the identification and functional analysis of MDSCs. Since our studies revealed that serum interferes with the immunosuppressive functions of 4T1‐induced MDSCs, we performed a series of experiments to elucidate the mechanism(s) by which this occurs. Although an understanding of the individual components of serum would be most helpful to address this, serum represents a complex mixture of proteins and other factors that are not well defined (Barnes and Sato, 1980; Dainiak, 1985). Nevertheless, our data indicate that albumin, a soluble protein that comprises approximately 50 to 60% of total protein in blood plasma (Seifert and Resman‐Targoff, 2006), strongly inhibits MDSC immunosuppression (Fig. 4.7). Albumin, which is synthesized in the liver, plays a number of important roles in vivo: it controls oncotic pressure in the plasma, transports amino acids synthesized in the liver to other tissues, and acts as a carrier protein for ligands that are relatively insoluble in plasma, such as fatty acids, metals, cholesterol, bile pigments, hormones, and drugs (Fasano et al., 2005; Seifert and Resman‐Targoff, 2006). In addition, albumin serves as an important antioxidant, both directly via its thiol group and indirectly via its ligand‐binding capabilities (Fasano et al., 2005; Roche et al., 2008). Albumin likely 84  carries out similar functions in vitro. In addition to serving as a carrier for factors that support cell growth, including FAs, trace minerals, hormones, and growth factors, there are also reports that albumin can protect cells from mechanical shear damage and act as a detoxifying agent for H2O2 (Monette and Sigounas, 1990; Roche et al., 2008). Interestingly, we found that albumin stripped of FAs blocked MDSC‐induced immunosuppression to a lesser extent than control albumin (Fig. 4.7c), suggesting that FAs bound to albumin may play a role in inhibiting MDSC‐induced immunosuppression. In addition, it is possible that albumin may act as a carrier for other molecules that restrict the immunosuppressive function of MDSCs. Our data indicate that albumin restricts the suppressive function of MDSCs, at least in part, by decreasing the formation of ROS. MDSCs isolated from a variety of murine and human tumors have been reported to suppress T cell responses via production of ROS, including H2O2, O2‐, and peroxynitrites (Nagaraj and Gabrilovich, 2010; Youn and Gabrilovich, 2010). Although our data clearly demonstrate that exposing 4T1‐induced MDSCs to serum or albumin decreases ROS production (Fig. 4.8), the flow cytometric approach we used to assay ROS detects generalized oxidative stress rather than any specific ROS (Eruslanov and Kusmartsev, 2010); therefore, the specific ROS involved remain unknown and their identification will be the subject of future studies. Although we have demonstrated that serum albumin inhibits the production of ROS, it is possible that other components of serum and/or other mechanisms could also be involved. Related to this, a recent report has suggested that 4T1‐induced MDSCs can suppress Ag‐stimulated T cell proliferation by restricting the availability of cysteine, which is an essential amino acid for T cell proliferation (Srivastava et al., 2010). It is therefore possible that serum contains additional cysteine/cystine and thus prevents 4T1‐induced MDSCs from suppressing T cell proliferation. However, this does not appear to be the case since we found that MDSCs were unable to potently suppress T cell proliferation in the presence of dialyzed serum (Fig. 4.6A), which lacks molecules smaller than 3.5 kDa, including amino acids such as cysteine (121.15 Da). Previous 85  studies have also suggested that 4T1‐induced MDSCs suppress T cells via an Arg1‐ dependent mechanism (Sinha et al., 2005c). Conceivably, proteins contained in serum may inhibit MDSC immunosuppression by interfering with Arg1 expression or L‐Arg metabolism, although our finding that neither pulmonary nor splenic 4T1‐induced MDSCs express Arg1 directly ex vivo (data not shown) does not support this hypothesis. Related to this, it is intriguing that serum inhibits the ability of MDSCs from 4T1 tumor‐ bearing mice to potently suppress T cell proliferation, but does not affect TAM‐ or PMφ‐ mediated immunosuppression (Fig. 4.4). This could be due to differences in the immunosuppressive mechanism(s) used by these cells or to differences in susceptibility to serum factors. While there are some reports of 4T1‐induced MDSCs suppressing T cell responses under serum‐containing culture conditions, i.e., in RPMI 1640 supplemented with 10% FCS (Ko et al., 2010; Kodumudi et al., 2010; Le et al., 2009), the magnitude of suppression is difficult to determine since the ratios of MDSCs to splenocytes have not always been reported. However, in our hands, 4T1‐induced MDSCs could not suppress T cell proliferation in RPMI 1640 or IMDM media containing 10% FCS, even at very high ratios (1 MDSC: 1 splenocyte). In fact, in the presence of FCS, MDSCs stimulated T cell proliferation in most experiments (Fig. 4.6C). It is possible that the type of T cell stimulation may affect MDSC immunosuppression in the presence of serum, given that one group reported 4T1‐induced MDSC suppression of T cells stimulated by a combination of bryostatin, ionomycin, IL‐2, and cultured in the presence of IL‐5, IL‐7, and 10% FCS (Le et al., 2009). In light of these studies, we propose that MDSCs may be capable of suppressing T cell responses in serum‐containing RPMI 1640 medium under certain conditions, but the use of serum in immunosuppression assays can mask the full suppressive effect of MDSCs. Taken together, our results demonstrate that different culture conditions have a considerable influence on the properties of MDSCs isolated from 4T1 tumor‐bearing mice. Since MDSCs are defined functionally via in vitro assays, these findings are of  86  critical importance to accurately assess the role of MDSC in promoting primary tumor growth and tumor metastasis. The presence of serum, and specifically albumin, conceals the immunosuppressive properties of MDSCs, which can lead to erroneous conclusions regarding the importance of 4T1‐induced MDSCs in immunosuppression. Furthermore, these findings may provide insight into the in vivo functions of MDSCs. It is possible that high levels of albumin in the peripheral blood may restrict the immunosuppressive abilities of MDSCs until they reach the tissues. This hypothesis is supported by a study by Sinha et al., in which they demonstrate that, at higher ratios, blood MDSCs were less suppressive than MDSCs isolated from the spleen, BM, and lung of 4T1 tumor‐bearing mice (Sinha et al., 2008). Related to this, we found that stimulated T cells were unable to proliferate in medium containing 10% mouse serum (data not shown). Although the basis for this is unclear, it is possible that mouse serum lacks some of the factors required to support T cell survival and/or proliferation or that it contains inhibitory factors that induce T cell anergy or death. This effect may have physiological importance, since proliferation of T cells in the peripheral circulation could have deleterious effects. In fact, this may be an example of a more general phenomenon, whereby serum constrains the activation of certain cell types, including MDSCs and T cells, and thereby maintains appropriate immune cell function. A recent report by Haverkamp et al. also indicates that MDSCs from different tissues possess different capacities to regulate T cell responses (Haverkamp et al., 2011). Using the RM‐ 1 prostate tumor model, they found that while MDSCs in peripheral tissues (i.e. spleen and liver) could acquire immunosuppressive abilities upon exposure to T cell‐produced IFN‐γ in vitro, only MDSCs isolated from the tumor site revealed immediate suppressive abilities (Haverkamp et al., 2011). Our findings, which are consistent with those of Haverkamp et al. and others (Nausch et al., 2008), support the idea that MDSCs may exhibit diverse properties in different contexts or in response to certain treatments (i.e. docetaxel) (Kodumudi et al., 2010), a feature that could be exploited therapeutically. Overall, our data clearly highlight the importance of testing different in vitro culture conditions on MDSC phenotype to ensure that the presence of serum is not masking the full immunosuppressive properties of MDSCs. These results will enable more accurate  87  identification of MDSCs based on their immunosuppressive function and will consequently advance our understanding of the roles that MDSCs perform in promoting primary and metastatic tumor growth.  88  CHAPTER 5 : MACROPHAGES ARE MORE POTENT IMMUNE SUPPRESSORS THAN MYELOID­DERIVED SUPPRESSOR CELLS IN MURINE METASTATIC MAMMARY CARCINOMA 5.1 Introduction Having established appropriate culture conditions to accurately assay the immunosuppressive abilities of myeloid cells, we embarked on a series of experiments to investigate the roles of different myeloid cells in tumorigenesis. Immunosuppressive myeloid cells, including MDSCs and mature Mφs, play pivotal roles in the promotion of primary tumor growth and are emerging as possible players in metastasis (Qian and Pollard, 2010; Youn and Gabrilovich, 2010). However, little is known about the relative immunosuppressive potencies of different myeloid cell types, the different mechanisms by which they exert their suppressive effects, or how the strength of immune suppression relates to the functions of these cells in tumor growth or metastasis. Characterization of the specific phenotypes of different myeloid cell types and the roles each play in vivo is difficult because of the extreme heterogeneity exhibited by tumor‐induced myeloid cells. Different subtypes of MDSCs and Mφs have been classified based on characteristics such as morphology (i.e. M‐MDSCs versus G‐MDSCs), activation state (i.e. M1 versus M2a, M2b, or M2c Mφs), or tissue‐specificity (Corzo et al., 2010; Haverkamp et al., 2011). However, these classifications only begin to describe the true heterogeneity of myeloid cells, since the phenotype of an individual cell can be influenced by a number of key factors including cell lineage and molecules present in the microenvironment both during and subsequent to differentiation (Laoui et al., 2011). Related to this, different populations of MDSCs and Mφs are thought to be able to exert their suppressive effects via different mechanisms, including, but not limited to, production of inhibitory cytokines, expression of inhibitory receptors, and expression of Arg1 and iNOS, which lead to L‐Arg depletion and formation of ROS and RNS (Cuenca et al., 2011). 89  Although there is evidence that MDSCs and TAMs contribute to metastasis, many questions still remain. For example, it is not well established if Mφ subtypes other than TAMs play a role in metastasis. Specifically, the contribution of Mφs from extra‐tumoral (i.e. peripheral tissue) sites in tumor‐bearing hosts remains unclear (Torroella‐Kouri et al., 2009). Also, given that MDSCs and TAMs have been reported to promote metastasis via similar mechanisms (Siveen and Kuttan, 2009; Youn and Gabrilovich, 2010) it is not known whether MDSCs and Mφs play distinct roles. Finally, little is known about the relationship between immune suppression and metastasis. The concept of the pre‐ metastatic niche proposes that BM‐derived cells migrate to future sites of metastasis ahead of tumor cells and promote the homing and growth of metastatic tumor cells (Kaplan  et  al.,  2005).  However,  whether  these  BM‐derived  cells  possess  immunosuppressive properties or whether immune suppression contributes to tumor metastasis is not well established. To investigate some of these questions, including whether MDSCs and Mφs play distinct roles in vivo, we implanted 4T1 tumor‐bearing mice with ATRA pellets to promote the differentiation of MDSCs to Mφs and DCs. ATRA, a member of the retinoid family and structurally related to vitamin A, exerts strong effects on myeloid cell proliferation, differentiation, and apoptosis in both normal and cancer cells (Bastien and Rochette‐Egly, 2004). ATRA induces immature cells to terminally differentiate and is used clinically to treat acute promyelocytic leukemia (APL) (a form of acute myeloid leukemia (AML)), in combination with chemotherapy (Sanz and Lo‐Coco, 2011). Based on its ability to reduce MDSC levels and abrogate MDSC‐mediated immune suppression, the use of ATRA as a possible cancer therapy has been proposed in the literature (Kusmartsev et al., 2003; Kusmartsev et al., 2008; Mirza et al., 2006). In this chapter we report that, as expected, mice treated with ATRA had decreased MDSCs and increased Mφs in their lungs. However, mice treated with ATRA displayed significantly more lung metastases than untreated or placebo‐treated mice. Furthermore, these data are consistent with our in vitro experiments, which reveal that lung Mφs are more potent  90  immune suppressors than lung MDSCs on a per cell basis. These findings contribute to our understanding of the processes and cell types involved in metastasis, which may lead to novel and more effective treatment strategies for cancer patients. 5.2 Results 5.2.1  Characterization of primary tumor growth and metastasis in 4T1 and 67NR tumor models  To address the roles of Mφs and MDSCs in tumor growth and metastasis we used two murine mammary carcinoma models. After three weeks, mice orthotopically implanted with metastatic 4T1 tumor cells (i.e. 4T1 mice) exhibited large primary tumors (Fig. 5.1A), considerable splenomegaly (Fig. 5.1B) due to extramedullary myelopoiesis (Ueha et al., 2011), and high numbers of metastatic tumor cells in their lungs (Fig. 5.1C). In contrast, although non‐metastatic 67NR primary tumors reached the same size as 4T1 tumors, mice bearing 67NR tumors (i.e. 67NR mice) did not exhibit splenomegaly (Fig. 5.1D), and, as expected, did not exhibit any lung metastases (data not shown). Next, we performed immunofluorescence on 4T1 and 67NR primary tumor sections to examine differences in perfusion and myeloid cell infiltration. 4T1 tumors were poorly vascularised and contained a high proportion of hypoxic tumor cells while 67NR tumors were well vascularised and hypoxic cells were not detectable (Fig. 5.2A). Moreover, the number of CD11b+ cells was much higher in 4T1 tumors (Fig. 5.2A), indicating greater myeloid cell infiltration. We also examined the lungs of 4T1 mice and found that the metastatic tumor nodules and the lung tissue were highly infiltrated with myeloid cells (Fig. 5.2B). Interestingly, we saw a marked difference in the location of different myeloid cell types; Gr1+ cells were located around the periphery of the tumor nodules (Fig. 5.2B, left), while F4/80+ cells had invaded deep into the tumor interior (Fig. 5.2B, right).  91  Figure 5.1 Characterization of 4T1 and 67NR tumor growth. Mice were implanted with 4T1 tumor cells and A, primary tumor weight, B, spleen weight, and C, number of metastatic tumor cells in the lungs were measured over time. Data are mean ± SEM with at least 15 mice per point. D, Mice were implanted with 4T1 (•) or 67NR (□) tumor cells, harvested at different times, and primary tumor weights and spleen weights measured. Each point represents an individual mouse.  92  Figure 5.2 Visualization of hypoxia and myeloid cell infiltration in 4T1 and 67NR tumors. A, 4T1 (left) or 67NR (right) primary tumor sections analyzed for anti‐CD11b (red), pimonidazole (hypoxia; green), and Hoechst 33342 (perfusion; blue). B, Lung sections from 4T1 mice analyzed for DAPI (nuclei; blue) and anti‐Gr1 (red; left) or anti‐F4/80 (red; right). Areas in dotted lines are metastatic tumor nodules. All images are 10x magnification. Data are representative of at least two experiments performed in duplicate.  5.2.2  4T1 but not 67NR tumors induce MDSCs and Mφs  To further characterize the involvement of myeloid cells in the 4T1 and 67NR tumor models, we investigated the ability of each tumor type to induce the accumulation of MDSCs (CD11b+Gr1+ cells) and Mφs (F4/80+CD11b+ cells) using flow cytometry. Mice bearing 4T1, but not 67NR tumors, exhibited significantly higher proportions of MDSCs and Mφs in both their lungs and spleens, compared to control (i.e. 93  non‐tumor‐bearing) mice (Fig. 5.3A). Similarly, the absolute number of both myeloid cell types was considerably higher in 4T1, but not 67NR, mice, compared to control mice (Fig. 5.3B). These contrasting effects were underscored by our observation that, in some cases, 67NR mice possessed significantly lower proportions and numbers of myeloid cells than control animals, while 4T1 mice displayed up to 230 or 600 fold more Mφs or MDSCs, respectively (Fig. 5.3A,B). These data are consistent with our ex vivo imaging results (Fig. 5.2) and reveal clear and intriguing differences in the induction of myeloid cells by 4T1 and 67NR tumors.  Figure 5.3 Induction of myeloid cells by 4T1 and 67NR tumors. A, Proportion and B, total number of MDSCs (CD11b+Gr1+) and Mφs (F4/80+CD11b+) in the lungs and spleens of control (i.e. non‐tumor bearing) (white bars), 4T1 (black bars), or 67NR (grey bars) mice three weeks post‐implant. Data are the mean ± SEM of at least two independent experiments performed in duplicate. *,p< 0.05; **,p<0.01; ***,p<0.001; NS, not significant relative to control mice; nd, not determined  94  5.2.3  Phenotypic characterization of myeloid cells from control versus tumor­bearing mice  5.2.3.1  TAMs  Since our data suggested that 4T1, but not 67NR, tumors induce the development of myeloid cells, we commenced a series of studies to investigate the roles of different myeloid cell subtypes in promoting tumor growth and metastasis. We began by comparing the cell surface phenotype and protein expression of 4T1 and 67NR TAMs. Flow cytometric analysis revealed that TAMs from both tumor models represented a heterogeneous population. Although almost all cells lacked expression of Gr1, subsets of TAMs exhibited different degrees of F4/80 and CD11b expression (Fig. 5.4A). In general, a larger proportion of 4T1 TAMs expressed high levels of both F4/80 and CD11b compared to 67NR TAMs (Fig. 5.4A). We also analyzed cell lysates to determine if TAMs were M1 or M2 skewed. As Figure 5.4B shows, 4T1 TAMs expressed Arg1 and 67NR TAMs expressed iNOS, indicating that 4T1 TAMs are M2 skewed while 67NR TAMs exhibit features of M1 activation.  95  Figure 5.4 Phenotypic characterization of TAMs. TAMs were isolated from 4T1 or 67NR mice and A, surface expression of F4/80, CD11b, and Gr1 was analyzed by flow cytometry and B, expression of Ym1, Arg1, and GAPDH was analyzed by Western blot. Data are representative of at least three experiments performed in duplicate.  96  5.2.3.2  PMφs  Next, we investigated the phenotype of PMφs from control and tumor‐bearing mice. PMφs from both control and 4T1 mice expressed CD11b and F4/80, but not Gr1 (Fig. 5.5A). Interestingly, control PMφs were a single, relatively uniform population, while 4T1 PMφs appeared to be composed of two distinct populations, one that expressed moderate levels of both CD11b and F4/80 and one that expressed high levels of these two markers (Fig. 5.5A). Next, we investigated whether these cells expressed features of M1 or M2 activation. Western blot analysis revealed that 4T1 PMφs expressed high levels of both Arg1 and Ym1, 67NR PMφs expressed very low Arg1 and no Ym1, and control PMφs lacked expression of both proteins (Fig. 5.5B). Furthermore, none of the cells expressed iNOS (data not shown). Overall, these data suggest that 4T1 PMφs are M2 skewed, while 67NR and control PMφs exhibit a naïve phenotype. 5.2.3.3  Splenic Mφs  Analysis of the surface phenotype of splenic Mφs (SpMφs) revealed that although, as expected, a large proportion expressed F4/80, few expressed CD11b (Fig. 5.6A). Moreover, SpMφ samples isolated from 4T1 mice appeared to contain some Gr1+ cells (Fig. 5.6B). This is perhaps not surprising given the large proportion of CD11b+/Gr1+ MDSCs present in 4T1 spleens (Fig. 5.3A). However, since there are reports that some MDSCs express F4/80, CD11b, and Gr1 (Ueha et al., 2011), it is difficult to determine whether these cells should be classified as SpMφs, MDSCs, or a mixed population. Interestingly very few F4/80+ cells from the spleens of control (Fig. 5.6A) or 67NR (Fig. 5.6C) mice co‐expressed Gr1+ .  97  Figure 5.5 Phenotypic characterization of PMφs. PMφs were isolated from control, 4T1, or 67NR mice and A, surface expression of F4/80, CD11b, and Gr1 was analyzed by flow cytometry and B, expression of Ym1, Arg1, and GAPDH was analyzed by Western blot. Data are representative of at least three experiments performed in duplicate.  98  Figure 5.6 Phenotypic characterization of SpMφs. SpMφs were isolated from A, control, B, 4T1, or C, 67NR mice and surface expression of F4/80, CD11b, and Gr1 was analyzed by flow cytometry. Data are representative of at least three experiments performed in duplicate.  99  5.2.3.4  Splenic Gr1+ cells  Lastly, we considered the phenotypic properties of Gr1+ cells isolated from the spleens of control, 4T1, or 67NR mice. Intriguingly, the cell surface phenotypes of these three populations varied significantly. Control and 67NR splenic Gr1+ cells appeared to be comprised of three distinct populations that differentially expressed Gr1: Gr1loCD11b‐, Gr1midCD11b+, and Gr1brightCD11b+ (Fig. 5.7). In contrast, cells from 4T1 mice were a uniform Gr1highCD11b+ population (Fig. 5.7), consistent with the general phenotype of MDSCs. In addition, we found that none of the Gr1+ cell populations isolated from any group of mice expressed F4/80 (Fig. 5.7). 5.2.4  Summary of the effect of 4T1 tumors on the phenotypes of myeloid cells  Taken together, our data demonstrate that, unlike non‐metastatic 67NR tumors, metastatic 4T1 tumors have a profound effect on the phenotypes of different myeloid cell populations. Whereas the phenotype of cells isolated from 67NR mice closely resembled that of cells from control mice, 4T1 tumors modified both cell surface and intracellular protein expression. In general, 4T1 tumors increased expression of proteins associated with M2 activation (i.e. inducing Arg1, but not iNOS) in PMφs and TAMs (Fig. 5.8A). Furthermore, it is noteworthy that neither Gr1+ nor F4/80+ splenic myeloid cells from control or 4T1 mice exhibited Arg1 expression, but Gr1+ cells isolated from the tumors of 4T1 mice did express both Arg1 and Ym1 (Fig. 5.8B).  100  Figure 5.7 Phenotypic characterization of splenic Gr1+ cells. Gr1+ cells were isolated from the spleens of control, 4T1, or 67NR mice and surface expression of Gr1, F4/80, and CD11b was analyzed by flow cytometry. Data are representative of at least three experiments performed in duplicate.  101  Figure 5.8 Analysis of protein expression in different myeloid cell populations. A, PMφs, splenic Gr1+ cells, SpMφs, or TAMs were isolated from control or 4T1 mice and expression of iNOS, Arg1, and GAPDH analyzed by Western blot. B, Gr1+ cells were isolated from the tumors of 4T1 mice and expression of Ym1, Arg1, and GAPDH analyzed by Western blot. Data are representative of at least three experiments performed in duplicate.  5.2.5  Comparison of immunosuppressive abilities of myeloid cells from control versus tumor­bearing mice  5.2.5.1  PMφs  Next, we investigated the effect of tumors on the immunosuppressive properties of different myeloid cell populations. Our previous results revealed that control PMφs could suppress T cell responses and we were interested to discover whether PMφs from tumor‐bearing mice also possessed immunosuppressive properties. We found that PMφs from control, 4T1, and 67NR mice all potently suppressed T cell proliferation (Fig. 5.9A). However, consistent with our phenotypic data, different tumor models had distinct effects on immunosuppressive function. Specifically, 4T1 PMφs, but not 67NR PMφs were significantly more immunosuppressive on a per cell basis than control PMφs (Fig. 5.9A) and this trend held true over multiple PMφ:splenocyte ratios (Fig. 5.9B). We 102  also performed these studies using C3H/HeN mice bearing metastatic squamous cell carcinoma (SCCVII) tumors (Suit et al., 1985) and found that PMφs from tumor‐bearing mice were significantly more suppressive than control PMφs at higher concentrations (i.e. lower ratios) (Fig. 5.9C). Taken together, these data suggest that certain tumors, such as the metastatic 4T1 and SCCVII cell lines, enhance the immunosuppressive functions of PMφs. 5.2.5.2  TAMs  We isolated TAMs from 4T1 or 67NR mice and compared their abilities to suppress T cell proliferation in vitro. At each ratio tested, 67NR TAMs were more potent immune suppressors, on a per cell basis (Fig. 5.10). This was somewhat surprising, since 4T1 TAMs, but not 67NR TAMs, expressed Arg1 (Fig. 5.4B), which is often associated with immune suppression. However, this finding could contribute to our understanding of the roles of TAMs in normoxic (i.e. 67NR) versus hypoxic (i.e. 4T1) primary tumors. 5.2.5.3  SpMφs  Studies quantifying the immunosuppressive properties of SpMφs revealed that they were only able to suppress T cell proliferation at high concentrations (Fig. 5.11). As the ratio of SpMφs: splenocytes was decreased, SpMφs were no longer immunosuppressive (Fig. 5.11A). Interestingly, SpMφs isolated from 4T1 mice exhibited very similar suppressive properties to control SpMφs (Fig. 5.11A). In fact, neither 4T1 or 67NR tumors altered the immunosuppressive abilities of SpMφs in a consistent manner (Fig. 5.11B).  103  Figure 5.9 Tumors augment the immunosuppressive abilities of PMφs. A, PMφs were isolated from control, 4T1, or 67NR mice and co‐cultured with polyclonal‐ stimulated splenocytes (1 PMφ: 8 splenocyte) and T cell proliferation measured. Data are representative of three experiments performed in triplicate. B, PMφs were isolated from control (o), 4T1 (■), or 67NR (∆) mice and co‐cultured with polyclonal‐stimulated splenocytes at different ratios (PMφs: splenocytes; 1:2, 1:4, 1:8, 1:16, 1:32). Proliferation was measured and percent of RC proliferation calculated. Data are the mean ± SEM of two independent experiments performed in triplicate. C, Polyclonal‐stimulated splenocytes were cultured alone (white bar) or with control (grey bar) or SCCVII (black bar) PMφs at different ratios (PMφs: splenocytes; 1:4, 1:8, 1:16, 1:32) and proliferation measured. Data are representative of two experiments performed in triplicate. *,p< 0.05; **,p<0.01; NS, not significant relative to control PMφs  104  Figure 5.10 67NR TAMs are more immunosuppressive than 4T1 TAMs. TAMs were isolated from 4T1 (open square) or 67NR (closed square) mice using F4/80+ selection and co‐cultured with polyclonal‐stimulated splenocytes at different concentrations (TAMs: splenocytes; 1:1, 1:2, 1:4, 1:8). T cell proliferation was measured and percent of RC proliferation calculated. Data are representative of three experiments performed in duplicate.  105  Figure 5.11 SpMφs are only immunosuppressive at high concentrations. A, SpMφs were isolated from control (ο) or 4T1 mice (■) and co‐cultured with polyclonal‐ stimulated splenocytes at different concentrations (SpMφs: splenocytes; 2:1, 1:1, 1:2, 1:4, 1:8, 1:16, 1:32, 1:64). T cell proliferation was measured and percent of RC proliferation calculated. Data are the mean ± SEM of three experiments performed in duplicate. B, SpMφs were isolated from control (white bars), 4T1 (black bars), or 67NR (grey bars) mice and co‐ cultured with polyclonal‐stimulated splenocytes at different concentrations (SPMφs: splenocytes; 1:1, 1:2). T cell proliferation was measured and percent of RC proliferation calculated. Data are representative of three experiments performed in triplicate.  106  5.2.5.4  Splenic Gr1+ cells  Next, we investigated the immunosuppressive properties of splenic Gr1+ cells. Although our phenotypic data revealed that most of these cells, particularly those isolated from 4T1 mice, express both CD11b and Gr1 (Fig. 5.7), before we could define these cells as MDSCs we first had to determine their immunosuppressive abilities. Interestingly, we observed an immense difference in the ability of splenic Gr1+ cells to suppress T cell responses depending on whether they were isolated from control, 4T1, or 67NR mice. As shown in Fig. 5.12A, while 4T1 Gr1+ cells potently suppressed T cell proliferation, control or 67NR cells were much less suppressive at all cell ratios tested. We also assayed the effect of splenic Gr1+ cells on T cell cytokine production. Consistent with the proliferation data, 4T1 Gr1+ cells but not control or 67NR Gr1+ cells, inhibited T cell‐produced IFN‐γ and increased levels of IL‐2 in the culture medium (Fig. 5.12B), which is consistent with a model in which T cells produce IL‐2 but are unable to utilize it for cell division (Strickland et al., 1996). We also found that all three types of Gr1+ cells suppressed T cell IL‐10 production to an equal, albeit minor, extent (Fig. 5.12B). Finally, we assessed the levels of NO and found that none of the splenic Gr1+ subsets produced considerable levels of NO upon co‐culture with activated splenocytes (Fig. 5.12B). Taken together, these data suggest that 4T1 splenic Gr1+ cells can be characterized as MDSCs, since they are capable of potently suppressing T cell responses, whereas Gr1+ cells from 67NR or control mice lack immunosuppressive properties and are therefore not MDSCs. These findings are consistent with the idea that many tumors induce both the differentiation and immunosuppressive function of MDSCs, yet also suggest that not all tumors induce splenic Gr1+ cells to acquire immunosuppressive abilities.  107  Figure 5.12 4T1 tumors enhance the immunosuppressive abilities of splenic Gr1+ cells. A, Splenic Gr1+ cells were isolated from control (o), 4T1 (■), or 67NR (∆) mice and co‐ cultured with polyclonal‐stimulated splenocytes at different ratios (Gr1+ cells: splenocytes; 2:1,1:1, 1:2, 1:4, 1:8, 1:16, 1:32). T cell proliferation was measured and percent of RC proliferation calculated. Data are representative of three experiments, performed in triplicate. B, Splenic Gr1+ cells were isolated from control, 4T1, or 67NR mice and co‐ cultured with polyclonal‐stimulated splenocytes (1 Gr1+ cell: 2 splenocytes). Supernatants were collected and concentrations of IFN‐γ, IL‐10, IL‐2, or NO were determined. Data are the mean ± SEM of two independent experiments, performed in duplicate.  108  5.2.5.5  Gr1+ cells from different tissues  Since MDSCs are known to accumulate in multiple organs and tissues (Ilkovitch and Lopez, 2009; Youn and Gabrilovich, 2010), we investigated the immunosuppressive abilities of Gr1+ cells isolated from various sites. Gr1+ cells isolated from the spleens or lungs of 4T1 mice suppressed T cell proliferation significantly more than their counterparts from control mice, and did so to an equal extent (Fig. 5.13A). Similarly, splenic and pulmonary 4T1 Gr1+ cells both suppressed T cell IFN‐γ and IL‐10 production and did so to a similar degree (Fig. 5.13B). Finally, we compared the immunosuppressive abilities of Gr1+ cells harvested from the spleens, lungs, tumors, kidneys, and livers of 4T1 mice and found that they all potently suppressed T cell proliferation to the same degree (Fig. 5.13C). Consequently, from this point forward we will refer to Gr1+ cells from 4T1 mice as MDSCs.  109  Figure 5.13 4T1 Gr1+ cells isolated from different tissues are equally immunosuppressive. A, Polyclonal‐stimulated splenocytes were co‐cultured with or without (white bar) Gr1+ cells isolated from the spleens or lungs of control (grey bars) or 4T1 mice (black bars) (1 Gr1+ cell: 2 splenocytes) and T cell proliferation measured. Data are representative of three independent experiments performed in triplicate. B, Gr1+ cells were isolated from the spleens or lungs of 4T1 mice and co‐cultured with polyclonal‐stimulated splenocytes (1 Gr1+ cell: 2 splenocytes). Supernatants were collected, and IFN‐γ and IL‐10 were measured. Data are representative of three independent experiments performed in duplicate. C, Gr1+ cells were isolated from the spleens, lungs, tumors, kidneys, or livers of 4T1 mice and co‐cultured with polyclonal‐stimulated splenocytes (1 Gr1+ cell: 2 splenocytes) and proliferation measured. Data are representative of two experiments, performed in triplicate. **,p<0.01; ***,p< 0.001 relative to control Gr1+ cells. 110  5.2.6  Summary of the relative immunosuppressive abilities of different myeloid cell populations from 4T1 tumor­bearing mice  As  previously  mentioned,  few  studies  have  directly  compared  the  immunosuppressive potencies of different myeloid cell types. To address this we plated equal numbers of splenic MDSCs and PMφs from control or 4T1 mice and analyzed their abilities to suppress polyclonal and Ag‐specific stimulated T cell proliferation. Our results clearly demonstrated that PMφs were much more immunosuppressive than splenic MDSCs, on a per cell basis (Fig. 5.14A). Specifically, these data suggest the following: 1) control PMφs, but not control Gr1+ cells possess immunosuppressive properties, 2) control PMφs are more immunosuppressive than 4T1 MDSCs, and 3) the presence of 4T1 tumors increases the immunosuppressive abilities of both Gr1+ cells and PMφs (Fig. 5.14A). As Figure 5.14B illustrates, PMφs were more than 32‐fold (polyclonal) or 10‐fold (Ag‐specific) more immunosuppressive than MDSCs isolated from the same 4T1 mouse. We also performed a direct comparison of all the myeloid cell types we isolated and found the following immunosuppressive hierarchy (listed from most suppressive to least suppressive): 4T1 PMφs > 4T1 TAMs ≈ control PMφs > 4T1 SpMφs > control SpMφs > 4T1 Gr1+ cells > control Gr1+ cells, and these trends were comparable whether splenocytes were activated in a polyclonal or Ag‐specific manner (Fig. 5.14C).  111  Figure 5.14 4T1 PMφs are more potent suppressors of T cell proliferation than 4T1 MDSCs on a per cell basis. A, Splenic Gr1+ cells and PMφs were isolated from control or 4T1 mice and co‐cultured with polyclonal‐ (left) or Ag‐specific‐ (right) stimulated splenocytes (1 myeloid cell: 2 splenocytes). T cell proliferation was measured and percent of RC proliferation calculated. B, Splenic Gr1+ cells (black bars) or PMφs (white bars) were isolated from 4T1 mice and co‐ cultured with polyclonal‐ or Ag‐specific‐ stimulated splenocytes (1 myeloid cell: 2 splenocytes). T cell proliferation was measured and percent of RC proliferation calculated. C, Splenic Gr1+ cells, PMφs, TAMs, or SpMφs were isolated from control or 4T1 mice and co‐ cultured with polyclonal‐ (left) or Ag‐specific‐ (right) splenocytes. T cell proliferation was measured and percent of RC proliferation calculated. Data are representative of three independent experiments performed in triplicate. **,p<0.01; ***,p< 0.001  112  These results are also consistent with data comparing the abilities of different myeloid cell types from 4T1 mice to suppress T cell cytokine production. As Figure 5.15A shows, PMφs and TAMs inhibited T cell‐produced IFN‐γ and IL‐10 to a greater extent than MDSCs or SpMφs. Interestingly, we also measured levels of NO in culture supernatants and found that only control PMφs, which we had previously determined suppress via a NO‐mediated mechanism (Hamilton et al., 2010), produced NO upon co‐ culture with activated splenocytes (Fig. 5.15B). These data suggest that none of the 4T1 myeloid cells we tested exerted their suppressive effects via NO production and are consistent with a previous report that PMφs from mammary tumor‐bearing mice produce less NO upon LPS stimulation than control PMφs (Di Napoli et al., 2005).  Figure 5.15 4T1 PMφs are more potent suppressors of T cell cytokine production than 4T1 MDSCs on a per cell basis. A, PMφs, splenic Gr1+ cells, TAMs, and SpMφs were isolated from 4T1 mice and co‐cultured with polyclonal‐stimulated splenocytes (1 myeloid cell: 2 splenocytes) and IFN‐γ and IL‐10 were measured in the culture supernatants. B, PMφs were isolated from control mice and PMφs, splenic Gr1+ cells, TAMs, and SpMφs were isolated from 4T1 mice. Myeloid cells were co‐cultured with polyclonal‐stimulated splenocytes (1 myeloid cell: 2 splenocytes). Supernatants were collected and NO measured. Data are the mean ± SEM of two independent experiments, performed in duplicate.  113  Finally, we corroborated these data using a flow cytometric approach. We labelled responder splenocytes with CFSE, a fluorescent dye that becomes stably trapped in the cytoplasm of individual cells. Using this method, we were able to visualize T cell proliferation, since each time a cell divides its fluorescence intensity is halved, resulting in a series of peaks that each represent one cell division (Lyons, 2000). We co‐cultured different myeloid cell types with stimulated CFSE‐labeled splenocytes and after 72 h analyzed the proportion of CD4+ and CD8+ T cells that had undergone cell division, as well as the number of divisions (Fig. 5.16). However, due to the absence of viable cells we were unable to obtain data for CD8+ T cells co‐cultured with 4T1 Gr1+ cells, and this is discussed further in Section 5.3. In general, the results were consistent with our 3H‐thy incorporation data; 4T1 PMφs, 4T1 Gr1+ cells, and 4T1 TAMs exhibited strong immunosuppressive effects, reducing the maximal number of T cell divisions (Fig. 5.16) and proportion of divided T cells (Fig. 5.17), while control Gr1+ cells, control SpMφs, and 4T1 SpMφs did not. In both proliferation assays, 4T1 PMφs were the most immunosuppressive, completely preventing all T cell division (Fig. 5.14 and Fig 5.17). Finally, we compared the ability of different myeloid cell types to suppress CD4+ versus CD8+ T cell proliferation and found they did so to a similar extent (Fig. 5.18).  114  Figure 5.16 Visualization of the immunosuppressive effects of different myeloid cell populations. Responder splenocytes were labelled with fluorescent CFSE dye and cultured with or without PMφs, splenic Gr1+ cells, TAMs, and SpMφs isolated from control or 4T1 mice ± polyclonal stimulation. After 72 h cells were harvested and analyzed by flow cytometry for expression of CD4, CD8, and CFSE. The proportion of CD4+ and CD8+ T cells is shown, as well as the proportion of each that had undergone cell division. Data are representative of three experiments performed in duplicate. N/A; assay could not be performed due to insufficient live cells.  115  116  Figure 5.17 Summary of the ability of different myeloid cell populations to inhibit T cell division as determined by flow cytometric analysis of CFSE intensity. Proportion of CFSE‐labelled polyclonal‐stimulated CD4+ (black bars) and CD8+ (white bars) T cells that divided at least once during 72 h co‐culture with or without different myeloid cell populations. Data are representative of three experiments performed in duplicate. n/a; assay could not be performed due to insufficient live cells.  117  Figure 5.18 Comparison of the ability of different myeloid cell populations to reduce the proportion of T cells. Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bars) different populations of myeloid cells and the proportion of viable CD4+ (upper) or CD8+ (lower) T cells after 72 h was determined by flow cytometry. Data are the mean ± SEM of two independent experiments performed in duplicate. **,p<0.01; ***,p< 0.001; NS, not significant relative to RC  118  5.2.7  4T1­induced MDSCs and TAMs, but not PMφs, decrease T cell viability  We subsequently performed a series of experiments to elucidate the mechanisms of immunosuppression utilized by different myeloid cell types. We first investigated the effect of different myeloid cells isolated from control or 4T1 mice on T cell viability. We found that co‐culture with 4T1 MDSCs or TAMs significantly decreased the proportion of viable splenocytes, but co‐culture with control or 4T1 PMφs, or with control splenic Gr1+ cells did not (Fig. 5.19A). However, when we specifically assayed the effect of myeloid cells on T cell viability, the results were less clear. The data suggest that control or 4T1 PMφs greatly increased the proportion of viable CD4+ and CD8+ T cells, while control or 4T1 splenic Gr1+ cells both decreased T cell viability, and 4T1 TAMs decreased CD4+ and increased CD8+ T cell viability (Fig. 5.19B). These results indicate that PMφs and splenic Gr1+ have contrasting effects on T cell survival and suggest that PMφs, MDSCs, and TAMs suppress T cells via different mechanisms.  119  Figure 5.19 Effect of different myeloid cell types on cell viability. Polyclonal‐stimulated splenocytes were co‐cultured for 72 h with or without different myeloid cell populations and A, Bulk cell viability (total splenocytes) and B, CD4+ (black bars) and CD8+ (white bars) T cell viability was determined by exclusion of PI by flow cytometry. For B, data are reported as the fold increase in T cell viability compared to stimulated splenocytes alone (responder control). Data are the mean ± SEM of two independent experiments performed in duplicate. ***,p< 0.001 relative to RC  120  5.2.8  4T1­induced MDSCs suppress T cells via a contact­independent mechanism  To further explore the different suppressive mechanisms used by Mφs and MDSCs we used a Transwell assay system to determine whether direct cell‐cell contact was required. We found that MDSCs isolated from either the spleen or lungs of 4T1 mice could suppress T cell proliferation whether or not they were separated from T cells by a semi‐permeable membrane (Fig. 5.20A), suggesting contact is not required for MDSC immunosuppression. In contrast, 4T1 PMφs were only immunosuppressive when in direct contact with T cells and promoted T cell proliferation when contact was removed (Fig. 5.20A). However, when we tested the ability of 4T1 PMφs to suppress T cell proliferation at different ratios, we found that at high concentrations PMφs could inhibit T cell proliferation even when contact with T cells was prevented, but this was not true at lower concentrations of PMφs (Fig. 5.20B). One hypothesis that is consistent with these data is that 4T1 PMφs produce one or more secreted factors that function along a concentration gradient, and as the number of PMφs increases, so does the level of the factor(s), which enables PMφs to suppress T cells from a greater distance. However, when the concentration of PMφs, and thus the secreted factor, are low, PMφs must be in close contact with T cells to inhibit their activation. Finally, we tested whether there were differences in contact requirement for control, 4T1, and 67NR PMφ immunosuppression and found that even at high concentrations, control PMφs required contact to effectively suppress T cell proliferation, while 4T1 and 67NR PMφs could suppress in both control and Transwell systems (Fig. 5.20C). However, all myeloid cell types we tested suppressed T cell responses to a greater degree when contact was allowed between myeloid cells and test cells (Fig. 5.20).  121  Figure 5.20 At high concentrations, both MDSCs and PMφs suppress T cells via contact­independent mechanisms. A, MDSCs (splenic and pulmonary) and PMφs were harvested from 4T1 mice and co‐cultured with polyclonal‐stimulated splenocytes (1 MDSC: 2 splenocytes or 1 PMφ: 8 splenocytes) in control (black bars) or Transwell (white bars) wells. T cell proliferation was measured and percent of RC proliferation calculated. B, 4T1 PMφs were co‐cultured with polyclonal‐ stimulated splenocytes at different concentrations (PMφ: splenocytes; 1:2, 1:8) in control (black bars) or Transwell (white bars) wells. T cell proliferation was measured and percent of RC proliferation calculated. C, PMφs were isolated from control, 4T1, or 67NR mice and co‐ cultured with polyclonal‐stimulated splenocytes (1 PMφ: 2 splenocytes) in control (black bars) or Transwell (white bars) wells. T cell proliferation was measured and percent of RC proliferation calculated. Data are representative of three independent experiments performed in triplicate.  122  5.2.9  PMφs from control and 67NR tumor­bearing mice suppress via a NO­ dependent mechanism  Since our previous work, presented in Chapter 3, had revealed that PMφs from control mice suppressed T cells via NO production (Hamilton et al., 2010), we tested whether any other myeloid cell types suppressed by this mechanism. We found that control and 67NR, but not 4T1 PMφs produced NO in response to co‐culture with T cells (Fig. 5.21A). Consistent with this, inhibiting NO production via iNOS inhibitors (i.e. L‐ NMMA, L‐NIL) or iNOS induction by IFN‐γ (i.e. anti‐IFN‐γ) reversed the suppressive effects of control but not 4T1 PMφs (Fig. 5.21B). Conversely, although 67NR PMφs produced moderate amounts of NO upon co‐culture with T cells, these inhibitors did not eradicate T cell suppression by 67NR PMφs (Fig. 5.21B), which may suggest that 67NR PMφs may suppress via different or additional mechanisms.  123  Figure 5.21 Control PMφs, but no other cell types tested, suppress T cell responses via NO production. A, PMφs were isolated from control mice; PMφs, splenic Gr1+ cells, TAMs, and SpMφs were isolated from 4T1 mice; and PMφs and TAMs were isolated from 67NR mice. Each cell type was cultured with polyclonal‐stimulated splenocytes (1 myeloid cell: 2 splenocytes) and supernatant NO production measured. B, PMφs were cultured with polyclonal‐stimulated splenocytes (1 PMφ: 8 splenocytes) ± L‐NMMA, L‐NIL, or αIFN‐γ. T cell proliferation was measured and percent of RC proliferation calculated.  124  5.2.10 The immunosuppressive properties of 4T1 PMφs are not abrogated by TLR stimulation Since we had previously elucidated the mechanism by which control PMφs suppress T cell responses, we were particularly interested in investigating the suppressive mechanism(s) used by 4T1 PMφs, to gain insight into the effect of tumors on the suppressive phenotypes of Mφs. Furthermore, our results had revealed 4T1 PMφs to be the most potent immunosuppressive cells of those we tested, which made them particularly intriguing. Since we had found that the suppressive properties of control PMφs could be reversed by pre‐treatment with LPS or dsRNA we first determined the effect of TLR agonists on 4T1 PMφ immunosuppression. Unlike control PMφs, pre‐treating 4T1 PMφs did not inhibit their immunosuppressive abilities; in fact, pre‐treatment with LPS and dsRNA increased the 4T1 PMφ‐induced suppression of T cells (Fig. 5.22).  Figure 5.22 The immunosuppressive properties of 4T1 PMφs are not reversed by pre­treatment with TLR ligands. Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bars) 4T1 PMφs (1 PMφ: 8 splenocytes) that had been pre‐treated ± LPS, CpG, dsRNA, or PGN for 24 h and washed, and T cell proliferation measured. Data are representative of three independent experiments performed in triplicate. *,p< 0.05; ***,p<0.001; NS, not significant relative to untreated PMφs 125  5.2.11 4T1 PMφ suppression of T cells is reversed by N­acetyl­cysteine Next, we utilized the same panel of inhibitors described in Chapter 3 to elucidate the mechanism of 4T1 PMφ‐induced immune suppression. Consistent with our previous results (Fig. 5.21B), inhibitors of the NO pathway did not reverse suppression (Fig. 5.23A). Similarly, blocking Abs to immunosuppressive cytokines also did not abrogate suppression (Fig. 5.23B), suggesting that 4T1 PMφs do not suppress T cell proliferation by production of IL‐4, IL‐10, IL‐13, or TGF‐β. Next we tested the effect of IL‐2 or L‐Arg supplementation, inhibiting Arg1, or blocking the membrane‐bound inhibitory molecules CTLA4 or TGF‐β, but none of these treatments restored T cell proliferation (Fig. 5.23C). Finally, we tested the effects of different ROS inhibitors on 4T1 PMφ immunosuppression. Catalase catalyzes the conversion of H2O2 to H2O and O2, and SOD catalyzes the dismutation of O2‐ into O2 and H2O2, while NAC, a derivative of cysteine, has both indirect and direct antioxidant functions; NAC acts as a scavenger of free radicals and also serves as a precursor in the formation of the antioxidant glutathione (GSH) (Zhang et al., 2011). While catalase and SOD had no effect in our assay, the presence of NAC completely eliminated 4T1 PMφ immunosuppression (Fig. 5.23). In fact, 4T1 PMφs stimulated T cell proliferation in the presence of NAC. Since NAC can also act as a source of cysteine we supplemented cultures with different concentrations of L‐cysteine to see if this mimicked the effect of NAC, but L‐cysteine did not reverse 4T1 PMφ‐induced immune suppression (data not shown). Taken together these data suggest that 4T1 PMφs suppress T cell responses by production of ROS other than H2O2 or O2‐.  126  Figure 5.23 NAC reduces the immunosuppressive abilities of 4T1 PMφs. Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bars) 4T1 PMφs (1 PMφ: 8 splenocytes) ± A, 0.5 mM L‐NMMA, 1 mM L‐NIL, or 10 μg/ml anti‐IFN‐γ; B, 10 μg/ml rat IgG, 10 μg/ml anti‐IL‐4, 2 μg/ml anti‐IL‐10, 10 μg/ml anti‐IL‐13, or 10 μg/ml anti‐TGF‐β; C, 100 U/well mIL‐2, 2 mM L‐Arg, 200 μM BEC, ± 10 μg/ml anti‐CTLA4, or 250 ng/ml LAP; or D, 1 mg/ml catalase, 200 U/ml SOD, or 10 mM NAC and T cell proliferation measured. Data are representative of five independent experiments performed in triplicate. 127  5.2.12 The immunosuppressive properties of 4T1 pulmonary MDSCs are inhibited by catalase Next, we performed a similar series of experiments to investigate the mechanism by which MDSCs isolated from the lungs of 4T1 mice suppress T cell responses. We tested the effect of pre‐treating MDSCs with different TLR ligands but none decreased MDSC‐induced immunosuppression (Fig. 5.24). We then used our panel of inhibitors to try to elucidate the mechanism of suppression. The only inhibitor that had an effect on 4T1 MDSC‐induced suppression was catalase (Fig. 5.25). Although treatment with catalase did not completely eliminate 4T1 MDSC suppressive function, it significantly reduced suppression and increased T cell proliferation (Fig. 5.25C), indicating that 4T1 pulmonary MDSCs exert their immunosuppressive effects, at least in part, by H2O2 production.  Figure 5.24 The immunosuppressive properties of 4T1 MDSCs are not reversed by pre­treatment with TLR ligands. Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bars) 4T1 splenic MDSCs (i.e. Gr1+ cells) (1 MDSC: 2 splenocytes) that had been pre‐treated ± LPS, CpG, dsRNA, or PGN for 24 h and washed, and T cell proliferation measured. Data are representative of three independent experiments performed in triplicate.  128  Figure 5.25 Catalase reduces the suppressive abilities of 4T1 MDSCs. Polyclonal‐stimulated splenocytes were cultured with (black bars) or without (white bars) 4T1 pulmonary MDSCs (1 PMφ: 2 splenocytes) ± A, 10 μg/ml rat IgG, 10 μg/ml anti‐IL‐4, 100 U/well mIL‐2, 10 μg/ml anti‐CTLA4, or 250 ng/ml LAP + 10 μg/ml anti‐TGF‐β; B, 0.5 mM L‐ NMMA, 1 mM L‐NIL, 10 μg/ml anti‐IFN‐γ, 2 mM L‐Arg, or 200 μM BEC; or C, 1 mg/ml catalase, , 10 mM NAC, or 200 U/ml SOD and T cell proliferation measured. Data are representative of three independent experiments performed in triplicate. ***,p< 0.001 relative to untreated 4T1 MDSCs 129  5.2.13 Mφs and MDSCs suppress T cell responses via different ROS­mediated mechanisms To follow up on these experiments, we determined whether our results regarding the different suppressive mechanisms of 4T1‐induced myeloid cells were valid for Mφs and MDSCs from different tissues. Given that the lung is a key metastatic target organ for 4T1 tumors we were particular interested in pulmonary Mφs. Importantly, the lungs of 4T1 mice contain both CD11b‐CD11c+F4/80+ resident alveolar Mφs and CD11b+CD11c‐F4/80+ infiltrating Mφs. Although both of these populations are isolated by F4/80 positive selection, the majority of lung Mφs in 4T1 mice are likely induced by tumor cells, given the high levels of splenomegaly (Fig. 5.1B) and myelopoiesis (Fig. 5.3) we observed in these mice. We first tested whether 4T1 PMφs and lung Mφs suppressed T cell proliferation by the same mechanism. We found that, like PMφs, the suppressive properties of lung Mφs were significantly decreased by NAC, but not catalase (Fig. 5.26A). However, it is interesting that NAC completely eliminated PMφ suppression, but only partially reduced the suppressive properties of lung Mφs (Fig. 5.26A). Importantly, neither NAC, nor any of the other inhibitors we tested significantly lessened the suppressive effects of 4T1 TAMs (data not shown), which suggests that TAMs inhibit T cells by a different mechanism than PMφs and lung Mφs. These data may suggest that Mφs at the tumor site function differently than Mφs in the periphery. We also tested whether MDSCs from different tissues utilized the same suppressive mechanism and found that catalase, but not NAC, significantly lessened the suppressive effects of both splenic and pulmonary 4T1 MDSCs, and did so to an equal extent (Fig. 5.26B). All together, our results are consistent with a model in which 4T1 Mφs exert extremely potent immunosuppressive functions via contact‐dependent ROS production, control Mφs strongly inhibit T cell proliferation via IFN‐γ‐induced NO production, and 4T1 MDSCs suppress T cells via contact‐independent ROS production, albeit to a lesser extent than Mφs (Fig. 5.27).  130  Figure 5.26 4T1 Mφs and MDSCs suppress T cell proliferation via different ROS­ mediated mechanisms. A, Lung Mφs or PMφs were isolated from 4T1 mice and co‐cultured with polyclonal‐ stimulated splenocytes (1 Mφ: 8 splenocytes) ± 10 mM NAC (black bars) or 1 mg/ml catalase (grey bars). T cell proliferation was measured and percent of RC proliferation calculated. B, Splenic or pulmonary Gr1+ MDSCs were isolated from 4T1 mice and co‐cultured with polyclonal‐stimulated splenocytes (1 MDSC: 2 splenocytes) ± 10 mM NAC (black bars), or 1 mg/ml (light grey bars) or 2 mg/ml (dark grey bars) catalase. T cell proliferation was measured and percent of RC proliferation calculated. Data are representative of two independent experiments performed in triplicate. *,p<0.05; **,p<0.01; ***,p< 0.001 relative to untreated cells  131  Figure 5.27 Myeloid cells suppress T cell proliferation via different mechanisms and to different degrees. 4T1 Mφs suppress T cell proliferation via contact‐dependent ROS production that is inhibited by treatment with NAC. Mφs from control mice exert their immunosuppressive effects via contact‐dependent NO production, which is prevented by treatment with L‐ NMMA, L‐NIL, and/or α‐IFNγ. Unlike Mφs, 4T1 MDSCs suppress T cell responses via a contact‐independent mechanism that is inhibited by treatment with catalase, indicating a role for H2O2. These cells also differ in their relative suppressive strengths; 4T1 Mφs exhibit the most potent immunosuppressive abilities, followed by control Mφs, and then 4T1 MDSCs.  132  5.2.14 ATRA increases lung metastasis in 4T1 tumor­bearing mice After  thoroughly  characterizing  the  phenotype  of  tumor‐induced  immunosuppressive myeloid cells, we next examined the functions of these cells in vivo. Specifically, we investigated the distinct roles of Mφs and MDSCs in promoting primary tumor growth and metastasis. We conducted a series of experiments using the drug ATRA, which induces the differentiation of MDSCs into Mφs and DCs (Kusmartsev et al., 2003; Nefedova et al., 2007). We treated 4T1 mice with or without placebo or slow‐ release ATRA pellets over a three week timeframe. We found that ATRA treatment did not have an effect on the growth rate of primary tumors, which was very similar in all groups (Fig. 5.28A). Since the lungs are a principal site of metastasis in the 4T1 model, we assayed the number of metastatic tumor cells in the lungs and found that ATRA treatment, especially the higher 10 mg dose, greatly increased lung metastases (Fig. 5.28B). Furthermore, we performed these same studies using the 4TO7 tumor model, which is a much less aggressive metastatic cell line compared to 4T1 cells (i.e. although 4TO7 cells can be recovered from the blood and lungs, visible metastases never develop) (Aslakson and Miller, 1992; Heppner et al., 2000), and found that ATRA treatment increased metastatic lung cells to an even greater extent (Fig. 5.28C). These data suggest that although ATRA treatment does not impact primary tumor growth, it enhances metastasis in both the 4T1 and 4TO7 models.  133  Figure 5.28 ATRA treatment does not alter primary tumor growth but increases lung metastasis. 4T1 mice were treated ± ATRA (5 mg or 10 mg) or placebo pellets and A, primary tumor growth and B, number of tumor cells in the lungs were measured over time. C, 4T07 mice were treated ± ATRA (5 mg or 10 mg) or placebo pellets and the number of tumor cells in the lungs were measured over time. Data are the mean ± SEM of two independent experiments, n>5 mice per group. 134  5.2.15 ATRA increases the proportion of Mφs, which are much more immunosuppressive than MDSCs in the lungs of 4T1 tumor­bearing mice We next performed a number of tests to elucidate the mechanisms by which ATRA treatment increased lung metastasis. First, we considered whether ATRA might influence the immunosuppressive properties of myeloid cells. We assayed the abilities of pulmonary and splenic MDSCs isolated from 4T1 mice treated with or without ATRA to suppress T cell proliferation and found that ATRA treatment did not drastically alter the immunosuppressive function of pulmonary or splenic MDSCs (Fig. 5.29A). Similarly, ATRA treatment did not have an effect on the immunosuppressive abilities of lung Mφs (Fig. 5.29B). However, in performing these experiments we observed that, consistent with our previous results, lung Mφs from mice in all treatment groups were much more potent suppressors of T cell responses than lung MDSCs isolated from the same mice (Fig. 5.29C). In fact, when assayed at the same ratio, pulmonary Mφs were up to 50‐fold more suppressive than pulmonary MDSCs (Fig. 5.29C).  135  Figure 5.29 ATRA does not change the immunosuppressive properties of MDSCs or Mφs. A, Pulmonary or splenic MDSCs were isolated from 4T1 mice treated ± ATRA (5 mg) or placebo pellets and co‐cultured with polyclonal‐stimulated splenocytes (1 MDSC: 1 splenocyte). T cell proliferation was measured and percent of RC proliferation calculated. Significance is relative to untreated cells. B, Lung Mφs were isolated from 4T1 mice treated ± ATRA (5 mg) or placebo pellets and co‐cultured with polyclonal‐stimulated splenocytes at different ratios (Mφ: splenocytes; 1:1 or 1:2). T cell proliferation was measured and percent of RC proliferation calculated. Significance is relative to untreated cells. C, MDSCs or Mφs were isolated from the lungs of 4T1 mice treated ± ATRA (5 mg or 10 mg) or placebo pellets and co‐cultured with polyclonal‐stimulated splenocytes (1 myeloid cell: 1 splenocyte). T cell proliferation was measured and percent of RC proliferation calculated. Data are the mean ± SEM of three independent experiments performed in triplicate. *,p<0.05; ***,p<0.001; NS, not significant 136  Because ATRA induces MDSC differentiation, and thus alters the proportions of different myeloid cells, we examined the impact of ATRA treatment on myelopoiesis. As expected, control and placebo‐treated 4T1 mice exhibited extreme splenomegaly (Fig. 5.30A). However, mice treated with ATRA exhibited much smaller spleens than control mice with the same size primary tumors (Fig. 5.30A), indicating that ATRA treatment decreased tumor‐induced myelopoiesis. Furthermore, we found that the ratio of MDSCs to Mφs in the lungs was significantly reduced by ATRA treatment (Fig. 5.30B), consistent with reports that ATRA treatment induces MDSC differentiation, and consequently reduces the number of MDSCs and increases the number of Mφs (Kusmartsev et al., 2003; Nefedova et al., 2007). Taken together, our data are consistent with a model in which ATRA enhances metastasis by promoting the differentiation of MDSCs to Mφs, which are much more potent suppressors of immune responses (Fig. 5.31).  Figure 5.30 ATRA increases the number of Mφs and reduces the number of MDSCs in 4T1 mice. A, 4T1 mice were treated ± ATRA (5 mg or 10 mg) or placebo pellets and primary tumor weight and spleen weight were measured over time. B, 4T1 mice were treated ± ATRA (5 mg or 10 mg) or placebo pellets, and lung MDSCs and Mφs were harvested and quantified. Data are the mean ± SEM of two independent experiments performed in duplicate. *,p<0.05; **,p<0.01; NS, not significant relative to untreated mice 137  Figure 5.31 ATRA induces the differentiation of MDSCs into more immunosuppressive Mφs and promotes lung metastasis in the 4T1 and 4TO7 tumor models.  138  5.3 Discussion In this chapter, we characterized the phenotypes of immunosuppressive myeloid cells from normal and neoplastic tissues, elucidated the mechanisms by which they inhibit immune responses, and investigated the roles they play in tumorigenesis. Our results reveal that both Mφs and MDSCs possess immunosuppressive abilities. However, we also found these myeloid cells differ with regard to the potency and mechanisms by which they exert their suppressive effects, as well as the functions they perform in vivo. Our data suggest that these characteristics are governed by a number of contextual factors including exposure to different tumor‐derived factors, location in different tissues, and activation state. In general, the studies performed in this chapter reveal a number of key findings regarding the different roles Mφs and MDSCs play both ex vivo and in vivo and highlight the importance of Mφs in immune suppression and metastasis. One important trend that we saw throughout our studies was the contrasting effects of different tumor models, specifically metastatic 4T1 tumors versus non‐ metastatic 67NR tumors, on the phenotype and function of Mφs and MDSCs. We found that 4T1 tumors induce immense expansion of myeloid cells, including both Mφs and MDSCs (Fig. 5.3), and skew their phenotype towards M2 activation (Fig. 5.8). Furthermore, 4T1 tumors enhanced the suppressive function of all cell types we investigated (Fig. 5.14). In contrast, 67NR tumors did not induce myelopoiesis (Fig. 5.3) or alter the phenotype of myeloid cells. The only exception to this trend was with regard to TAMs, which we found were more immunosuppressive from 67NR mice than 4T1 mice (Fig. 5.10). Taken together, these data support two ideas: 1) Different tumors have contrasting effects on myeloid cell induction and immune suppression and 2) there may be important differences between the effect of tumors on cells located within the tumor microenvironment versus the periphery. Although we report significant differences in the effect of 4T1 and 67NR tumors on immunosuppressive myeloid cells, the underlying principles remain unclear. It has  139  been reported that MDSCs are induced by almost all types of human and murine tumors (Youn et al., 2008); however, we found that 67NR tumors did not induce either the accumulation or activation of myeloid cells (Fig. 5.3, 5.12). Myelopoiesis is part of a normal immune response and is induced by any number of chronic and acute inflammatory conditions, including autoimmune disease, trauma, burns, and sepsis (Cuenca et al., 2011). Indeed, pro‐inflammatory factors produced by the tumor, tumor stroma, and infiltrating T cells, including cytokines, chemokines, and S100 proteins, induce accelerated myelopoiesis and drive the mobilization of mature Mφs, PMNs, and immature populations from the BM and blood to inflammatory sites (Ueda et al., 2009). Interestingly, it has also been shown that these same factors not only drive myeloid cell expansion and activation, but also prevent the differentiation of immature myeloid cells, thus promoting MDSC accumulation. Related to this, studies by Kusmartsev et al. have revealed that MDSCs from tumor‐bearing mice injected into healthy animals quickly lose their immunosuppressive phenotype and terminally differentiate into Mφs and DCs (Kusmartsev et al., 2003), indicating that MDSCs have the potential to mature, but are trapped by their environment in an immature state. Therefore, our finding that 67NR tumors do not promote myelopoiesis has a number of interesting implications. First, it suggests that, unlike the vast majority of tumors, 67NR tumors do not induce an inflammation response. Further, it suggests that 67NR tumor cells do not produce factors that promote myeloid cell expansion or activation. It seems unusual that 67NR tumors can grow and thrive in the mammary fat pad, yet not induce a systemic inflammatory response. One possible explanation could be related to the non‐ metastatic nature of 67NR tumors; perhaps the same factors that promote metastasis also induce myelopoiesis or perhaps disseminated tumor cells are required to initiate an inflammatory response. Alternatively, the inability of 67NR tumors to induce myeloid cells or modify their phenotypes may be related to the high degree of vascularization in 67NR tumors(Fig. 5.2). Perhaps the lack of myeloid cell induction and infiltration in 67NR tumors is due to the lack of hypoxic cells in 67NR tumors, which have been shown to induce the production of factors that stimulate angiogenesis and myeloid cell infiltration (Corzo et al., 2010; Doedens et al., 2010; Laoui et al., 2011; Qian  140  and Pollard, 2010). Additional studies using a panel of tumor cell lines, including metastatic and non‐metastatic tumors, as well as tumors that exhibit different degrees of vascularization and hypoxia will be integral to gaining a better understanding of the factors that regulate myeloid cell infiltration and function. As mentioned earlier, there appear to be substantial tissue‐specific differences in the phenotype of Mφs and MDSCs. Most notably, the phenotype of TAMs and tumor MDSCs differs from their peripheral tissue counterparts. For example, we found that MDSCs in peripheral lymphoid organs did not express iNOS or Arg1, while tumor MDSCs expressed both Arg1 and Ym1 (Fig. 5.8B). We also discovered that Mφs in the periphery (i.e. PMφs and lung Mφs) suppressed T cell responses via a different mechanism than TAMs. Intriguingly, most studies investigating MDSCs are performed using cells isolated from peripheral lymphoid organs (i.e. spleen), while much of what we know about the role of Mφs in cancer has been based on studies with TAMs. Consequently, the role of tumor‐associated MDSCs and peripheral Mφs, as well as how they compare to their differently‐located counterparts, remains largely unknown (Corzo et al., 2010; Torroella‐Kouri et al., 2009). In the past few years, a number of groups have investigated the phenotypes of immunosuppressive myeloid cells in the peripheral lymphoid organs versus the tumor site and published results that are consistent with our own data (Haverkamp et al., 2011), including a recent report from Corzo et al. that presented a number of important findings. Using an EL‐4 ascites model, they noted major differences in the immunosuppressive function between tumor and splenic MDSCs. Specifically, they found splenic MDSCs could only suppress Ag‐specific CD8+ T cell function, while tumor MDSCs could suppress both Ag‐specific and non‐ specific T cell responses, and that these differences were principally regulated by hypoxia and induction of hypoxia‐inducible factor (HIF)‐1α (Corzo et al., 2010). They also found that HIF‐1α expression induced the expression of iNOS and Arg1 in MDSCs, decreased ROS production, and promoted the differentiation of MDSCs into TAMs (Corzo et al., 2010). These data shed light on a number of issues relevant to our work and may help explain the relatively low potency of MDSC immunosuppression we  141  observed. Since the necessary mice are not available on the BALB/c background, we were unable to assay the ability of different myeloid cell types to inhibit Ag‐specific CD8+ T cell responses, and instead only tested suppression of Ag‐specific CD4+ and polyclonal (non‐specific) T cell functions. Therefore, it is possible that our finding that splenic and lung MDSCs are less potent immune suppressors than Mφs may not be applicable to tumor MDSCs and/or CD8+ Ag‐specific‐stimulated T cell responses. Furthermore, the finding that hypoxia regulates immunosuppressive function, as well as MDSC differentiation, may explain some of the differences we observed in suppressive ability between 4T1‐ and 67NR‐ induced myeloid cells. Importantly, this phenomenon does not appear to be restricted to MDSCs, since hypoxia and HIF‐1α expression have also been implicated in regulating the suppressive functions of Mφs, as demonstrated by a recent study from Johnson’s group (Doedens et al., 2010). Taken together, these findings may highlight an integral role for hypoxia in regulating the phenotype and functions of myeloid cells, and, in turn, tumor progression and metastasis. Unlike TAMs, the roles of peripheral Mφs in tumorigenesis are not yet clearly elucidated (Torroella‐Kouri et al., 2009). Our results demonstrate that the presence of 4T1 tumors dramatically alters the phenotype and enhances the immunosuppressive function of Mφs in peripheral tissues, such as the PC and lung (Fig. 5.14). Our finding that increased numbers of lung Mφs correlate with increased metastatic growth in the lungs raises questions about whether Mφs contribute to the formation of the pre‐ metastatic niche. To date, most groups have focused on the involvement of VEGF‐R1 expressing BM‐derived HPCs in niche development (Kaplan et al., 2005), but our data suggest a potential role for Mφs, and future studies will be integral to determine their contribution to this process. In addition, the role of PMφs in the context of cancer remains unclear. It is possible that PMφs function to suppress activated T cells in the PC, and may contribute to ascites formation in this manner. Alternatively, the increased immunosuppressive abilities of PMφs may be representative of a systemic immunosuppressive state. Taken together, these results suggest that both TAMs and  142  peripheral Mφs are central players in the immunosuppressive network and their roles should not be neglected or minimized. Rather, further studies should be performed to clarify the roles Mφs play in different stages of carcinogenesis and validate these cells as potential therapeutic targets. As we have discussed, tumor‐induced myeloid cells display immense heterogeneity. As Figure 5.7 shows, we analyzed MDSC expression of CD11b and Gr1 and found some degree of variation in Gr1 expression. However, these subpopulations were most evident in control and 67NR Gr1+ cells, while 4T1 MDSCs appeared to be a more uniform population (Fig. 5.7). Moreover, we analyzed splenic and pulmonary 4T1 MDSCs by flow cytometry and found that only a very small proportion (<5%) expressed Ly6C and the vast majority were Ly6G+ (data not shown), which are consistent with our morphological analysis presented in Chapter 4 (Fig. 4.3). Consequently, we did not analyze different subpopulations of 4T1 MDSCs, but rather chose to study them as a single ‘bulk’ population.  However, there are reports indicating that different  subpopulations of MDSCs vary with regard to immunosuppressive function (Movahedi et al., 2008). Related to this, a recent study by Bronte’s group investigating the phenotype of three splenic MDSC subpopulations, isolated by their differential expression of Gr1, found that the CD11b+/Gr1int subset, which was comprised of monocytes and myeloid cell progenitors, was highly suppressive, while the CD11b+/Gr1high subset, made up of mostly granulocytes, was only immunosuppressive in some tumor models and under certain conditions (Dolcetti et al., 2010). Interestingly, they found that the development and function of more immunosuppressive subsets was dependent on tumor‐produced GM‐CSF (Dolcetti et al., 2010). These data, along with reports from several other groups (Corzo et al., 2010; Cuenca et al., 2011; Laoui et al., 2011; Peranzoni et al., 2010; Youn and Gabrilovich, 2010), highlight the extreme heterogeneity encompassed by the MDSC classification and reinforce the idea that MDSCs cannot be identified by expression of particular markers, but rather by their immunosuppressive nature (Laoui et al., 2011).  143  Since accurate detection of immunosuppressive function was critical both for the accurate identification of MDSCs, as well as for our studies comparing the relative immunosuppressive potencies of myeloid cells, we used a number of assays to quantify suppression of T cell responses.  For the most part, whether we assayed T cell  proliferation by 3H‐thy incorporation or flow cytometry analysis of CFSE, the results were very consistent. However, we observed two important discrepancies in the results of these assays that reveal important clues about immunosuppressive mechanism. The first is that while PMφs isolated from control mice strongly suppressed T cell proliferation as measured by 3H‐thy incorporation (Fig. 5.9), they did not reduce the proportion of divided T cells (Fig. 5.17). A possible explanation for this disparity relates to the different time frames during which proliferation is measured using each method. Analysis by flow cytometry determines T cell division over the entire 72 h period of co‐ culture while 3H‐thy incorporation only measures proliferation in the final 18 h. Therefore, these data may suggest that control PMφs are not immediately suppressive, but require time and/or activation to acquire suppressive function. This is consistent with our finding that control PMφs are only able to produce NO and exert their suppressive effects following exposure to T cell‐produced IFN‐γ, which induces iNOS expression in PMφs (Fig 3.13). Moreover, Figure 5.18 provides further evidence that control PMφs possess immunosuppressive properties, since control PMφs significantly reduced the proportion of both CD4+ and CD8+ T cells present in the culture after 72 h. The other noteworthy difference between the results from the two methods is that, at first glance, 4T1 MDSCs seem much more immunosuppressive by CFSE‐flow cytometric analysis than by the 3H‐thy assay (Fig. 5.16). However, these data indicate that the majority of cells are dead (98.1%) following 72 h co‐culture with 4T1 MDSCs (Fig. 5.16), which, in theory, would suggest MDSCs are exerting non‐specific cytotoxic activities. However, these results are not consistent with the high levels of 3H‐thy incorporation we detected, since if all the cells were dead there would be little or no incorporation of 3H‐thy during the last 18 h of co‐culture. Furthermore, when we determined the proportion of viable T cells following 4T1 MDSC co‐culture using a non‐  144  CFSE flow cytometric approach, we found that 4T1 MDSCs reduced the proportion of viable CD4+ T cells from 37.4% to 20.3% and CD8+ T cells from 22.3% to 4.35% (Fig. 5.18). These numbers are consistent with the data from the 3H‐thy assay and indicate that 4T1 MDSCs restrict the proliferation of both CD4+ and CD8+ T cells, but do not obliterate all viable T cells. Taken together, these data suggest that CFSE may increase the sensitivity of splenocytes to the cytotoxic effects of 4T1 MDSCs and support our finding that MDSCs are less immunosuppressive on a per cell basis than Mφs. Further studies specifically investigating the effect of CFSE pre‐treatment on the viability of splenocytes co‐cultured with 4T1 MDSCs will reveal the mechanisms involved. Related to this, when we investigated the effect of different Mφ and MDSC populations on bulk splenocyte and T cell viability we made a number of interesting observations. First, we noted that although control and 4T1 PMφs potently suppressed T cell proliferation, they increased the proportion of viable T cells in culture (Fig. 5.19B). In contrast, co‐culture with either control Gr1+ cells (which were not immunosuppressive) or 4T1 MDSCs significantly promoted T cell death (Fig. 5.19B). These data suggest that Mφs and MDSCs have opposing effects on T cell viability and also provide evidence that T cell suppression and death are not always directly related. In general, suppression of T cell responses can be achieved via induction of anergy, apoptosis, or necrosis.  However, further studies will be required to accurately  distinguish which of these mechanisms is utilized by different immunosuppressive myeloid cells. It seems probable that the increase in T cell viability we see is due to the removal of dead or dying cells from the culture by phagocytic Mφs but studies using specific inhibitors of phagocytosis will be required to substantiate this hypothesis. One of the key findings from these studies is the elucidation of the mechanisms by which 4T1 Mφs and MDSCs exert their immunosuppressive functions. We found that both Mφs and MDSCs suppress T cell proliferation by ROS production, but use different mechanisms. Specifically, the suppressive effects of MDSCs were contact‐independent (Fig. 5.20A) and inhibited by catalase (Fig. 5.25C), suggesting MDSCs inhibit T cells by  145  production of extracellular H2O2. Although MDSCs in other tumor models have been reported to produce ROS, this immunosuppressive mechanism has not been specifically reported for 4T1 MDSCs. In contrast, we found that 4T1 PMφ‐induced suppression of T cells was inhibited by NAC (Fig. 5.23D) and that PMφs were much more potent suppressors when in contact with target cells (Fig. 5.20B). NAC contains both a thiol (sulfhydryl; SH) and cysteine group, each of which contributes to its anti‐oxidant properties (Zhang et al., 2011). On the one hand, via its thiol group, NAC serves as a proton donor, reducing unstable molecules such as ROS (Zhang et al., 2011). On the other hand, via its cysteine group, NAC functions as the rate‐limiting factor in GSH synthesis, which is a potent intracellular antioxidant in cells (Pompella et al., 2003). Therefore, although NAC is an important player in the neutralization of free radicals and ROS both directly and indirectly (Zhang et al., 2011), we are unable to surmise from our data the specific ROS produced by 4T1 PMφs. However, since neither catalase nor SOD reduced their suppressive effects, our results indicate that 4T1 PMφs suppress T cells via intracellular ROS other than H2O2 or O2‐. Not only do our data demonstrate that MDSCs and Mφs suppress T cells via different mechanisms, they also suggest that MDSCs and Mφs play distinct roles in tumorigenesis. To discriminate between the effects of these different cell types we used ATRA, which induces the terminal differentiation of MDSCs into Mφs and DCs (Almand et al., 2001; Kusmartsev et al., 2003; Mirza et al., 2006). A recent report by Nefedova et al. investigated the mechanism by which ATRA induced MDSC differentiation and found that ATRA increased levels of glutathione synthase (GSS) in MDSCs, which in turn resulted in increased levels of GSH and decreased ROS production (Nefedova et al., 2007). These findings are consistent with previous reports suggesting that in addition to mediating the suppressive effects of MDSCs, ROS also contributed to the inability of MDSCs to differentiate into mature cells (Kusmartsev and Gabrilovich, 2003; Kusmartsev et al., 2004). Interestingly, ATRA did not stimulate the differentiation of MDSCs induced in a model of sepsis (Cuenca et al., 2011), which is compatible with the idea that different stimuli induce MDSCs with different phenotypes (i.e. sepsis‐induced  146  MDSCs do not produce ROS and thus are not affected by ATRA treatment). Nevertheless, a number of studies have demonstrated that ATRA potently eliminates MDSCs both in vitro (Almand et al., 2001; Gabrilovich et al., 2001) and in vivo, in tumor‐bearing mice and cancer patients (Kusmartsev et al., 2003; Kusmartsev et al., 2008; Mirza et al., 2006), by differentiating MDSCs into mature myeloid cells. Furthermore, there is evidence that ATRA, in combination with cancer vaccines or chemotherapy, increases cancer treatment efficacy (Kusmartsev et al., 2003; Mirza et al., 2006; Sanz and Lo‐Coco, 2011). However, little is known about the effect of ATRA on primary or metastatic growth in the absence of other treatments. Our studies reveal that treating tumor‐ bearing mice with ATRA reduces the number of MDSCs, increases the number of Mφs (Fig. 5.30), and increases lung metastasis, but does not affect primary tumor growth (Fig. 5.28). These data argue against the notion that inducing MDSC differentiation leads to increased anti‐tumor immune responses and decreased tumorigenesis. Rather, they suggest that, at least in some tumor models, inducing MDSC differentiation can promote metastasis. Interestingly, we noted that while ATRA enhanced lung metastasis in both 4T1 and 4TO7 models, it increased 4TO7 metastasis to a greater extent. This may be because 1) the baseline level of 4TO7 tumor cell metastasis is much lower, allowing the metastasis‐enhancing effects of ATRA to be more evident, and/or 2) 4TO7 tumors may be more sensitive to the both pro‐ and anti‐tumor effects of the immune system, i.e., in the absence of ATRA the anti‐tumor immune response is sufficient to inhibit 4TO7 metastasis, but the increase in pro‐tumor Mφs mediated by ATRA profoundly increases 4TO7 metastatic growth. Given our finding that Mφs are more potent immune suppressors than MDSCs on a per cell basis (Fig. 5.14), we propose that inducing the differentiation of MDSCs to Mφs increases overall immune suppression, which in turn fosters the growth of metastatic tumor cells in the lungs. In contrast, altering the ratio of MDSCs to Mφs with ATRA treatment did not affect primary tumor growth (Fig. 5.28A), which may indicate that MDSCs and Mφs play a similar role in promoting primary tumor growth than metastatic growth. This is consistent with our preliminary data demonstrating that tumor MDSCs  147  and TAMs exhibit comparable immunosuppressive abilities (data not shown), unlike MDSCs and Mφs isolated from metastatic sites. The idea that MDSCs and Mφs have different roles in metastasis is further supported by our finding that MDSCs and Mφs are located in very different regions of metastatic tumor nodules, with MDSCs lining the periphery of nodules and Mφs invading deep into the tumor interior (Fig. 5.2B). It is also possible that Mφs promote metastatic growth to a greater extent than MDSCs due to other mechanisms (i.e. production of factors that support cell growth or angiogenesis). Although our hypothesis is supported by the critical role immune suppression is known to play in all stages of tumorigenesis, including metastasis, the relative contribution of the immunosuppressive functions of Mφs to metastatic growth could be more accurately determined using a T cell‐deficient mouse model. As a whole, our studies underscore the vast heterogeneity, as well as the phenotypic similarities, of tumor‐induced myeloid cells. By characterizing the differences between cells from normal tissues and their tumor‐induced counterparts we have increased our understanding of the effects of different tumors on the phenotype and immunosuppressive function of myeloid cells. Importantly, our results highlight that not all tumor models have equivalent effects on the induction and suppressive function of myeloid cells, demonstrate that while most myeloid cells possess immunosuppressive properties the potency varies dramatically between different cell types as do the mechanisms by which cells exert their suppressive effects, and finally, reveal that Mφs may play a critical role in promoting metastasis, and may do so to a greater extent than MDSCs. Together, these findings help clarify the roles of different myeloid cells in cancer, which may contribute to the development of targeted therapeutics.  148  CHAPTER 6 : SUMMARY AND FUTURE DIRECTIONS The overall objective of this thesis was to investigate the factors that regulate the immunosuppressive properties of different myeloid cells and elucidate the roles these cells play in both normal and neoplastic tissues. Although the immunomodulatory properties of myeloid cells have long been recognized (Holda et al., 1985; Strober, 1984), when we embarked on this work many questions remained regarding the specific cell types involved, the mechanisms by which they functioned, and the contexts in which they exerted their effects. Our studies have focused on clarifying the immunosuppressive properties of two mononuclear cell populations, Mφs and MDSCs. Overall, the results of our work reveal a number of important insights into the functions and regulation of these cells in both physiological and pathological environments, highlight the considerable phenotypic and functional heterogeneity exhibited by Mφs and MDSCs, and underscore the considerable therapeutic potential of Mφs. In Chapter 3, we examined the immunosuppressive abilities of myeloid cells under physiological conditions and obtained a number of interesting results (Hamilton et al., 2010). Although CD11b+Gr1+ IMCs exist in various murine tissues, including the BM and spleen, they do not possess immunosuppressive abilities under steady‐state conditions and therefore are not considered MDSCs (Delano et al., 2007; Laoui et al., 2011). We thus focused on exploring the immunosuppressive functions of Mφs. We determined that resident Mφs, isolated from the PC of non‐tumor‐bearing mice, potently suppressed T cell responses via an IFN‐γ and NO‐dependent mechanism (Hamilton et al., 2010). These results are consistent with previous studies from the early 1990s that identified this pathway in Mφs (Albina et al., 1991; Cox et al., 1992; Deng et al., 1993). However, these original studies have been largely overlooked in recent years and the factors that regulate this pathway, and thus determine whether or not Mφs exert their suppressive effects, remained poorly understood. We performed a series of experiments to address this question and discovered that the suppressive abilities of Mφs were eliminated by TLR agonists that signal through the TRIF cascade 149  (i.e. LPS or dsRNA), but not by those that signal exclusively through MyD88 (i.e. CpG or PGN) (Hamilton et al., 2010). Moreover, we found that these effects were mediated by the induction of IFN‐β that occurs following TRIF recruitment and IRF‐3 activation, which decreases the ability of Mφs to respond to IFN‐γ stimulation (Hamilton et al., 2010). These novel findings reveal a role for resident Mφs in the maintenance of immune homeostasis, increase our understanding of the mechanisms by which Mφs contribute to T cell tolerance, and suggest a key role for TLR signaling and IFN‐β in regulating the immunosuppressive functions of Mφs. The findings presented in Chapter 3 offer novel insights into the mechanisms regulating Mφ‐induced T cell tolerance on multiple levels. At the molecular level, our data provide evidence that IFN‐β can antagonize IFN‐γ signaling; however, the precise mechanism(s) by which this occurs remain(s) unclear and will require further work. Specifically, activation levels of different IFN‐γ signaling components (e.g. IFN‐γ receptor, JAK‐1, JAK‐2, STAT1) in IFN‐γ‐stimulated Mφs pre‐treated with or without IFN‐β would contribute to our understanding of this process. Our data support the concept that NO can play divergent roles in immune responses (Prabhu and Guruvayoorappan, 2010) and demonstrate a previously unappreciated link between TLR signaling and regulation of immune suppression. These findings also contribute to our understanding of the factors that regulate Mφ suppression and T cell tolerance at the cellular and organismal levels. Intriguingly, we show for the first time that viral‐ but not bacterial‐derived PAMPs reduce the immunosuppressive properties of Mφs. However, an important caveat to these findings is they are based on in vitro stimulation of Mφs with individual PAMP ligands and this does not accurately reflect the context in which Mφs would encounter intact pathogens in vivo. Future experiments investigating the effect of different pathogens on Mφ immunosuppression using in vivo animal infection models, e.g., Legionella pneumophila (intracellular bacteria), Mycobacterium tuberculosis (extracellular bacteria), influenza (virus), Candida albicans (fungus), Trichuris muris (helminth) would validate our in vitro results and provide insights into the physiological rationale for these findings. 150  Importantly, our data reveal specific factors that regulate the suppressive mechanisms of Mφs and suggest that Mφs may be attractive therapeutic targets. Treatment with IFN‐β, LPS, dsRNA, or small molecule iNOS inhibitors may reduce Mφ suppression, which could be useful following infection or as part of future cancer therapies. Alternatively, small molecule Arg1 or IFN‐β inhibitors, via their enhancement of the immunosuppressive properties of Mφs, may be desirable in patients with autoimmune diseases or GvHD. Future studies will be required to test the effectiveness of these potential therapies in vivo, in animal models of clinical diseases. Overall, the findings presented in Chapter 3 emphasize the dual role of Mφs in the promotion and inhibition of immune responses in normal tissues and highlight their considerable therapeutic potential, as well as the importance of increasing our understanding of the factors that regulate Mφ function. After investigating the regulatory role of Mφs under physiological conditions, we sought to determine their role in a tumor environment, as well as compare the immunosuppressive properties of Mφs and MDSCs in tumor‐bearing mice.  However,  although there are many reports in the literature of tumor‐derived MDSCs inhibiting T cell responses in vitro, we had difficulties reproducing these findings. This prompted us to consider the effect of different cell culture conditions on the immunosuppressive properties of MDSCs and the results of these experiments are presented in Chapter 4. We found that MDSCs isolated from 4T1 tumor‐bearing mice, unlike PMφs or TAMs, could not suppress T cell proliferation in the presence of FCS and, in fact, promoted T cell proliferation in some experiments (Hamilton et al., 2011). Investigation of the underlying mechanisms revealed that serum albumin was a major contributor to the antagonistic effects of FCS on MDSC‐induced immune suppression and did so by restricting ROS production from MDSCs (Hamilton et al., 2011). These novel findings have a number of important implications regarding the accurate detection and identification  of  MDSCs  as  well  as  the  mechanisms  that  regulate  the  immunosuppressive properties of MDSCs.  151  MDSCs represent an extremely heterogeneous population of cells. In fact, some have argued that identifying MDSCs by co‐expression of CD11b and Gr1 has led to ambiguity in the literature since this definition is neither specific nor inclusive (Cuenca et al., 2011). In reality, cells of the mononuclear phagocyte lineage can be particularly difficult to identify by cell surface markers since the phenotype of these cells depends on a number of factors including inflammatory status (i.e. steady‐state versus inflammatory conditions), type and/or duration of inflammation, and tissue location (Laoui et al., 2011). Therefore, MDSCs must be identified functionally, which requires assays to accurately assess MDSC immunosuppression ex vivo. Our data, which demonstrate for the first time that different culture conditions have a profound effect on the suppressive functions of MDSCs, are of critical importance for researchers regarding the appropriate design and interpretation of immunosuppression assays. One caveat is that our studies were limited to 4T1‐induced MDSCs and we did not test the effects of serum on MDSCs induced in any other models. Validation of these results in different models of MDSC induction (e.g. different tumor models, trauma, parasitic infections, sepsis) will reveal whether this phenomenon is unique to 4T1‐ MDSCs or more generally applicable to other tumor or inflammatory models. Similarly, we  did  not  separate  different  populations  of  MDSCs,  but  studied  the  immunosuppressive features of bulk pulmonary or splenic Gr1+ cells. Although we found that 4T1‐MDSCs were a relatively uniform population by flow cytometric and morphological analyses (>95% of cells were G‐MDSCs), further studies comparing the effects of serum and albumin on the suppressive functions and ROS production of different subpopulations of MDSCs will reveal the potential applicability and/or limitations of our findings. Despite the technical nature of the work presented in Chapter 4, our results may shed light on the physiological roles of MDSCs in tumor biology. Our finding that MDSC‐ induced immunosuppression is not a fixed property, but varies in response to different culture conditions is consistent with reports that, in certain contexts, MDSCs can exert pro‐inflammatory functions (Cuenca et al., 2011; Nausch et al., 2008; Pastula and 152  Marcinkiewicz, 2011). Briefly, there is evidence that the same properties of MDSCs that enable them to inhibit T cell responses, such as ROS and inflammatory cytokine production, play a role in anti‐microbial responses and increased immune surveillance (Cuenca et al., 2011). Moreover, there are some reports that MDSCs, under some circumstances, can be immunostimulatory (Nausch et al., 2008; Pastula and Marcinkiewicz, 2011), which is consistent with results obtained in some of our experiments in which serum was present. In addition, our finding that albumin blunts ROS production and, as a result, immune suppression, demonstrates for the first time that 4T1‐induced MDSCs can inhibit T cell responses via ROS production. Although these experiments did not reveal the particular ROS involved, our findings from Chapter 5 suggest a possible role for H2O2, and future studies measuring production of H2O2 and other ROS should be performed to validate this finding. As previously mentioned, these data may suggest a physiological function for albumin in restricting the activation and function of MDSCs. While we have not identified the specific mechanism(s) at play, we have some evidence that albumin‐associated FAs may be involved. It would be interesting to further elucidate the involvement of FAs in regulating MDSC function and test the effect of altering the ratio of different FAs (i.e. omega‐3 versus ‐6) as this might be of therapeutic value. In summary, the data presented in Chapter 4 clearly highlight the importance of testing different in vitro culture conditions on MDSC function to ensure that the presence of serum is not masking the full immunosuppressive properties of MDSCs. These results will enable more accurate identification of MDSCs based on their immunosuppressive properties and consequently advance our understanding of the roles that MDSCs perform in promoting primary and metastatic tumor growth. Once we established culture conditions that allowed immune suppression to be accurately assayed in vitro, we embarked on a series of studies comparing the regulatory properties of different subtypes of myeloid cells from control, 4T1, and 67NR tumor‐bearing mice. These experiments, communicated in Chapter 5, revealed several striking results. We demonstrated that, contrary to currently accepted notions, not all tumors induce myelopoiesis and the development of immunosuppressive cells. 153  We noted a significant difference in the vascularization of 4T1 and 67NR tumors, consistent with a number of reports suggesting a link between hypoxia and induction and/or activation of immunosuppressive myeloid cells (Erler et al., 2009; Laoui et al., 2011). In general, hypoxia and HIF‐1α stabilization promotes the immunosuppressive function of MDSCs (Corzo et al., 2010) and Mφs (Doedens et al., 2010). HIF‐1α also stimulates the pro‐angiogenic functions of Mφs, by recruiting Mφs to hypoxic regions (i.e. in the tumor) and driving Mφ production of VEGF and matrix metalloproteinase (MMP)9 (Grimshaw et al., 2002; Murdoch et al., 2008). Further studies using myeloid‐ specific HIF‐1α mice (i.e. HIF‐1α deletion under the control of the M‐CSFR promoter) would be helpful in further elucidating the role of hypoxia on the induction and activation of suppressive myeloid cells. In addition, proteomic studies identifying the different factors produced by 4T1 and 67NR tumors cultured in vitro under either normoxic or hypoxic conditions will be critical to understanding the specific tumor‐ derived factors, as well as the role of hypoxia, in inducing myelopoiesis and promoting metastasis. Although the majority of current research is focused on targeting MDSCs, when we compared the suppressive properties of Mφs and MDSCs isolated from the same tissues, we found that Mφs suppressed T cell proliferation to a much greater extent. Moreover, we found that 4T1 Mφs and MDSCs inhibited T cell responses via different mechanisms; although both Mφs and MDSCs produced ROS, Mφ suppression was contact‐dependent and inhibited by NAC while MDSC suppression was contact‐ independent and inhibited by catalase. There a few potential caveats that must be considered related to these in vitro results: 1) due to the lack of appropriate genetically modified mice we were unable to compare the ability of Mφs and MDSCs to specifically suppress CD8+ T cell Ag‐specific responses, 2) due to low cell yield we were not able to compare the suppressive abilities of tumor‐associated MDSCs and Mφs, and 3) rather than examining precise MDSC and Mφ subsets we examined total cell populations. Identification of individual subpopulations of myeloid cells is difficult since these cells are highly related and exhibit variable phenotypes depending on the stimuli to which  154  they are exposed (Laoui et al., 2011). Related to this, we observed major differences in both the strength and mechanism of suppression of cells in peripheral versus tumor tissues. For example, we found that, unlike peritoneal and pulmonary Mφs, TAM‐ mediated immune suppression could not be reversed by NAC, or any other inhibitors that we tested in our assay. These findings are consistent with multiple studies that have reported significant functional differences between cells located in the tumor and periphery (Corzo et al., 2010). It is well established that cancer is a systemic disease, inducing wide‐spread immune deficiency in addition to local immune suppression (Torroella‐Kouri et al., 2009). However, while much emphasis has been placed on elucidating the roles of TAMs in tumor progression, the role of peripheral Mφs remains poorly understood (Torroella‐Kouri et al., 2009). Our data clearly show that the presence of certain tumors changes the phenotype of peripheral Mφs, increasing their suppressive capabilities. Studies by the Lopez laboratory investigating the phenotype of PMφs from mice bearing D1‐DMBA‐3 mammary tumors showed that these cells exhibit decreased APC ability (Watson and Lopez, 1995) and produce lower levels of IL‐1β, IL‐6, IL‐12, TNF‐α, NO, CCL2, and M‐CSF compared to control PMφs (Dinapoli et al., 1996; Handel‐Fernandez et al., 1997; Torroella‐Kouri et al., 2009). Recently, they published that the decreased pro‐ inflammatory functions of tumor PMφs are due, at least in part, to lower expression of NF‐κB and CCAAT/enhancer binding protein (C/EBP) TFs and that these cells, which cannot be classified as either M1 or M2, are more sensitive to apoptosis and express lower levels of Mφ markers (i.e. F4/80, CD11b, CD68, CD116). (Torroella‐Kouri et al., 2005; Torroella‐Kouri et al., 2009). These studies from the Lopez laboratory provide a mechanistic basis for the effect of tumors on peripheral Mφs and it would be useful to perform these experiments in 4T1 tumor‐bearing mice to reveal whether these mechanisms are consistent between different mammary tumor models, as well as to explore the functions of these cells in vivo.  155  As well, in Chapter 5 we show for the first time that ATRA treatment, which induces the differentiation of MDSCs to terminally differentiated Mφs and DCs, enhances metastatic growth. These results are particularly intriguing given that ATRA has been proposed as a possible cancer therapy because of its ability to decrease MDSC levels (Kusmartsev et al., 2008). ATRA is currently used to treat APL in combination with chemotherapy (Sanz and Lo‐Coco, 2011); however, a key difference is that, in APL, ATRA targets the malignant cells directly, rather than acting indirectly through MDSCs. Our results suggest that, at least in certain tumor models, Mφs promote metastasis to a greater degree than MDSCs. Although this finding is consistent with our data demonstrating that, in vitro, Mφs suppress T cell responses to a greater degree than MDSCs, these experiments are not sufficient to determine causality. Related to this, there is substantial evidence that Mφs can promote metastasis via non‐ immunosuppressive functions including production of IL‐1, TNF, IL‐6, VEGF, and proteases that breakdown the ECM and aid tumor cell escape and migration such as cathepsins, MMPs, and serine proteases (Laoui et al., 2011; Mantovani and Sica, 2010; Qian and Pollard, 2010). On the other hand, Mφs have also been reported to promote carcinogenesis via multiple immune regulatory mechanisms, such as production of IL‐ 10, PGE2, and/or TGF‐β (Kuang et al., 2009; Torroella‐Kouri et al., 2009), upregulation of PD‐1L on monocytes (Kuang et al., 2009), induction of Tregs via CCL22 production (Curiel et al., 2004), and based on our results, ROS production. Given the key role of immune suppression in fostering tumorigenesis, it is possible that Mφs promote 4T1 metastatic growth, at least in part, via suppression of anti‐tumor T cell‐mediated immunity. Further experiments using T cell‐deficient mice may be helpful to test this hypothesis; however, initial studies comparing the immunosuppressive properties of Mφs isolated from wild‐type and T‐cell deficient mice will be required to assess the validity of this approach. Another question that should be addressed is the role of DCs in  ATRA‐treated  mice.  Studies  analyzing  the  relative  proportions  and  immunosuppressive functions of Mφs and DCs in 4T1 mice treated with or without ATRA, as well as studies specifically targeting Mφs or DCs, should be performed to clarify the distinct roles of these cells in metastasis.  156  While we have demonstrated that Mφs enhance metastatic growth, we have not investigated at which stage Mφs are involved. Metastasis requires a number of steps: release of tumor cells from the primary tumor, migration through the circulatory or lymphatic system, extravasation in distant tissues, and subsequent survival and growth (Qian and Pollard, 2010). Recently, important work has revealed a direct and critical role for Mφs in tumor cell extravasation, survival, and metastatic growth (Qian et al., 2009). Moreover, Mφs likely play a role in the establishment of the pre‐metastatic niche, since they are recruited to future sites of metastasis prior to the arrival of metastatic tumor cells and greatly enhance metastatic seeding and growth (Qian et al., 2009). However, the precise timing, cell types, and mechanisms involved are subjects of current debate and future investigation (Dawson et al., 2009; Mantovani and Sica, 2010; Psaila and Lyden, 2009; Qian et al., 2009). While we successfully elucidated a number of pathways that regulate the immunosuppressive properties of myeloid cells, many questions still remain. In particular, we did not address the specific tumor‐derived factors that induce the accumulation and function of Mφs. Previous studies have demonstrated roles for a number of different factors (i.e. IL‐1β, IL‐6 (Bunt et al., 2006; Bunt et al., 2007), S100A8/A9 proteins (Sinha et al., 2008), PGE2 (Eruslanov et al., 2010; Sinha et al., 2007b), VEGF (Fricke et al., 2007; Gabrilovich et al., 1998), M‐CSF (Laoui et al., 2011), GM‐CSF (Dolcetti et al., 2010)), chemokines (i.e. CCL2 (Laoui et al., 2011; Qian et al., 2011)), and signaling pathway components (i.e. NF‐κB, STAT‐1, STAT‐3, C‐EBPβ, HIF‐1α (Karin and Greten, 2005; Laoui et al., 2011; Mantovani and Sica, 2010)) in both murine and human cancers. In particular, inhibition of the CCL2‐CCR2 axis (Qian et al., 2011; Si et al., 2010) and M‐CSF (Abraham et al., 2010; Aharinejad et al., 2009; Kubota et al., 2009; Lin et al., 2001) have proven particularly effective in limiting the cancer‐ promoting functions of myeloid cells. It is likely that the specific factors involved will vary with different tumor types, disease stages, and at the person‐to‐person level. Given the immense therapeutic potential of targeting these pathways using small molecule inhibitors it would be extremely beneficial to develop an in vitro assay that could be  157  used to screen for efficacy (i.e. reduction in immunosuppressive cell development and/or function) on an individual basis. Although the work presented in each chapter of this thesis reveals individual interesting and novel findings, there are some common themes that exist throughout the entirety of our studies. Firstly, the regulatory properties of myeloid cells are not fixed, but vary depending on the factors to which these cells are exposed. Examples of this include regulation of Mφ immunosuppression by different TLR agonists, regulation of MDSC immunosuppression by albumin, and regulation of both Mφ and MDSC immunosuppression by different tumor‐derived factors. Importantly, the mechanisms that control myeloid cell immunosuppression are complex and precise, i.e. not all TLRs or tumors have the same effect on myeloid cell function. While the studies presented in this thesis reveal some of the specific mechanisms that regulate myeloid cell‐induced immune suppression, there is a great deal that we still do not understand; further elucidation of the factors that regulate myeloid cell immunosuppression is one of the most important challenges in our field. Secondly, our findings emphasize the enormous heterogeneity of myeloid cells. The phenotypic and functional characteristics of both Mφs and MDSCs are incredibly diverse and vary depending on many factors (e.g. cell lineage, maturation state, activation state, duration and/or strength of stimulation by specific factors, cell‐cell interactions, tissue location, inflammatory status, etc.). Additional studies characterizing the heterogeneity of myeloid cells and identifying meaningful ways to define these cells will be essential for advancing research. Finally, the findings presented in this thesis demonstrate the important regulatory functions of Mφs, which have been largely overlooked by the field to date. Although MDSCs are considered one of the main contributors to immune suppression in cancer, our work reveals that Mφs are more potent immune suppressors than MDSCs and suggests Mφs play key roles in both peripheral tolerance and tumor‐induced immune suppression. While we do not dispute the influence of MDSCs, we propose that Mφs are important regulatory cells that should be explored as potential therapeutic targets.  158  Taken together, the findings presented in this thesis provide further evidence of the important regulatory roles of Mφs in the maintenance of homeostasis and in cancer, and reveal new insights into the factors that control the suppressive functions of these cells. The relationship between Mφs and tumors is complex. There is evidence that Mφs play both positive and negative roles and participate in all three stages of the cancer immunoediting process (i.e. elimination, equilibrium, and escape) (Qian and Pollard, 2010; Schreiber et al., 2011). However, data from both animal models and the clinic indicate that in most cases Mφs promote carcinogenesis, and in fact, due to their extreme diversity and plasticity, contribute to all phases of tumorigenesis (Qian and Pollard, 2010). Further characterization of the molecular and functional heterogeneity of tumor‐induced myeloid cells, using studies analogous to those presented herein, will be critical in the development of targeted therapies. Nevertheless, the potent immunosuppressive properties of Mφs, combined with our enhanced understanding of the factors that regulate these functions, make them extremely attractive therapeutic targets. Although these findings will need to be validated in patients, it is our hope that the findings presented in this thesis will inform the development of enhanced immunotherapies to treat human disease. Strategies that inhibit immune suppression could be utilized to promote anti‐tumor immunity and also to treat infectious diseases. Moreover, therapies that enhance immune suppression could be harnessed to reduce autoimmune disease and increase transplantation tolerance. In summary, the ability of mononuclear phagocytes to contribute to immunity and tolerance, as well as to the promotion and suppression of anti‐tumor responses, emphasizes their considerable therapeutic potential and highlights the importance of increasing our understanding of the factors that regulate myeloid cell function under normal and neoplastic conditions.  159  REFERENCES Abraham, D., K. Zins, M. Sioud, T. Lucas, R. Schafer, E.R. Stanley, and S. Aharinejad. 2010. Stromal cell‐derived CSF‐1 blockade prolongs xenograft survival of CSF‐1‐ negative neuroblastoma. Int J Cancer 126:1339‐1352. Aharinejad, S., M. Sioud, T. Lucas, and D. Abraham. 2009. Targeting stromal‐cancer cell interactions with siRNAs. Methods Mol Biol 487:243‐266. Akira, S., and K. Takeda. 2004. Toll‐like receptor signalling. Nat Rev Immunol 4:499‐511. Albina, J.E., J.A. Abate, and W.L. Henry, Jr. 1991. Nitric oxide production is required for murine resident peritoneal macrophages to suppress mitogen‐stimulated T cell proliferation. Role of IFN‐gamma in the induction of the nitric oxide‐synthesizing pathway. J Immunol 147:144‐148. Alcon, V.L., C. Luther, D. Balce, and F. Takei. 2009. B‐cell co‐receptor CD72 is expressed on NK cells and inhibits IFN‐gamma production but not cytotoxicity. Eur J Immunol 39:826‐832. Allison, A.C. 1978. Mechanisms by which activated macrophages inhibit lymphocyte responses. Immunol Rev 40:3‐27. Almand, B., J.I. Clark, E. Nikitina, J. van Beynen, N.R. English, S.C. Knight, D.P. Carbone, and D.I. Gabrilovich. 2001. Increased production of immature myeloid cells in cancer patients: a mechanism of immunosuppression in cancer. Journal of immunology 166:678‐689. Antignano, F., M. Hamilton, S. Patterson, V. Ho, C. Cohen, M.K. Levings, and G. Krystal. 2011. SHIP‐Deficient Dendritic Cells, Unlike Wild Type Dendritic Cells, Suppress T Cell Proliferation via a Nitric Oxide‐Independent Mechanism. PLoS One 6:e21893. Aslakson, C.J., and F.R. Miller. 1992. Selective events in the metastatic process defined by analysis of the sequential dissemination of subpopulations of a mouse mammary tumor. Cancer Res 52:1399‐1405. Barnes, D., and G. Sato. 1980. Serum‐free cell culture: a unifying approach. Cell 22:649‐ 655. Bastien, J., and C. Rochette‐Egly. 2004. Nuclear retinoid receptors and the transcription of retinoid‐target genes. Gene 328:1‐16. Bettini, M., and D.A. Vignali. 2009. Regulatory T cells and inhibitory cytokines in autoimmunity. Curr Opin Immunol 21:612‐618. Bhardwaj, N. 2007. Harnessing the immune system to treat cancer. The Journal of clinical investigation 117:1130‐1136. 160  Bingisser, R.M., P.A. Tilbrook, P.G. Holt, and U.R. Kees. 1998. Macrophage‐derived nitric oxide regulates T cell activation via reversible disruption of the Jak3/STAT5 signaling pathway. J Immunol 160:5729‐5734. Bingle, L., N.J. Brown, and C.E. Lewis. 2002. The role of tumour‐associated macrophages in tumour progression: implications for new anticancer therapies. J Pathol 196:254‐265. Bogdan, C. 2001. Nitric oxide and the immune response. Nat Immunol 2:907‐916. Brem‐Exner, B.G., C. Sattler, J.A. Hutchinson, G.E. Koehl, K. Kronenberg, S. Farkas, S. Inoue, C. Blank, S.J. Knechtle, H.J. Schlitt, F. Fandrich, and E.K. Geissler. 2008. Macrophages driven to a novel state of activation have anti‐inflammatory properties in mice. J Immunol 180:335‐349. Bronte, V., P. Serafini, A. Mazzoni, D.M. Segal, and P. Zanovello. 2003. L‐arginine metabolism in myeloid cells controls T‐lymphocyte functions. Trends Immunol 24:302‐306. Bunt, S.K., P. Sinha, V.K. Clements, J. Leips, and S. Ostrand‐Rosenberg. 2006. Inflammation induces myeloid‐derived suppressor cells that facilitate tumor progression. Journal of immunology 176:284‐290. Bunt, S.K., L. Yang, P. Sinha, V.K. Clements, J. Leips, and S. Ostrand‐Rosenberg. 2007. Reduced inflammation in the tumor microenvironment delays the accumulation of myeloid‐derived suppressor cells and limits tumor progression. Cancer research 67:10019‐10026. Burnet, M. 1957. Cancer; a biological approach. I. The processes of control. Br Med J 1:779‐786. Cesta, M.F. 2006. Normal structure, function, and histology of the spleen. Toxicol Pathol 34:455‐465. Chitu, V., and E.R. Stanley. 2006. Colony‐stimulating factor‐1 in immunity and inflammation. Curr Opin Immunol 18:39‐48. Coghlin, C., and G.I. Murray. 2010. Current and emerging concepts in tumour metastasis. J Pathol 222:1‐15. Composto, G., D. Gonzalez, A. Bucknum, D. Silberman, J. Taylor, M. Kozlowski, T. Bloomfield, T. Bartlett, and J. Riggs. 2011. Peritoneal T lymphocyte regulation by macrophages. Immunobiology 216:256‐264. Condeelis, J., and J.W. Pollard. 2006. Macrophages: obligate partners for tumor cell migration, invasion, and metastasis. Cell 124:263‐266.  161  Corzo, C.A., T. Condamine, L. Lu, M.J. Cotter, J.I. Youn, P. Cheng, H.I. Cho, E. Celis, D.G. Quiceno, T. Padhya, T.V. McCaffrey, J.C. McCaffrey, and D.I. Gabrilovich. 2010. HIF‐1alpha regulates function and differentiation of myeloid‐derived suppressor cells in the tumor microenvironment. The Journal of experimental medicine 207:2439‐2453. Corzo, C.A., M.J. Cotter, P. Cheng, F. Cheng, S. Kusmartsev, E. Sotomayor, T. Padhya, T.V. McCaffrey, J.C. McCaffrey, and D.I. Gabrilovich. 2009. Mechanism regulating reactive oxygen species in tumor‐induced myeloid‐derived suppressor cells. Journal of immunology 182:5693‐5701. Cox, G.W., G. Melillo, U. Chattopadhyay, D. Mullet, R.H. Fertel, and L. Varesio. 1992. Tumor necrosis factor‐alpha‐dependent production of reactive nitrogen intermediates mediates IFN‐gamma plus IL‐2‐induced murine macrophage tumoricidal activity. J Immunol 149:3290‐3296. Cuenca, A.G., M.J. Delano, K.M. Kelly‐Scumpia, C. Moreno, P.O. Scumpia, D.M. Laface, P.G. Heyworth, P.A. Efron, and L.L. Moldawer. 2011. A paradoxical role for myeloid‐ derived suppressor cells in sepsis and trauma. Mol Med 17:281‐292. Curiel, T.J., G. Coukos, L. Zou, X. Alvarez, P. Cheng, P. Mottram, M. Evdemon‐Hogan, J.R. Conejo‐Garcia, L. Zhang, M. Burow, Y. Zhu, S. Wei, I. Kryczek, B. Daniel, A. Gordon, L. Myers, A. Lackner, M.L. Disis, K.L. Knutson, L. Chen, and W. Zou. 2004. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat Med 10:942‐949. Dainiak, N. 1985. Role of defined and undefined serum additives to hematopoietic stem cell culture. In Hematopoietic Stem Cell Physiology. E.P. Cronkite, N. Dainiak, R.P. McCaffrey, J. Palek, and P.J. Quesenberry, editors. Alan R. Liss, New York. 59‐76. Damen, J.E., H. Wakao, A. Miyajima, J. Krosl, R.K. Humphries, R.L. Cutler, and G. Krystal. 1995. Tyrosine 343 in the erythropoietin receptor positively regulates erythropoietin‐induced cell proliferation and Stat5 activation. EMBO J 14:5557‐ 5568. Dawson, M.R., D.G. Duda, S.S. Chae, D. Fukumura, and R.K. Jain. 2009. VEGFR1 activity modulates myeloid cell infiltration in growing lung metastases but is not required for spontaneous metastasis formation. PLoS One 4:e6525. Delano, M.J., P.O. Scumpia, J.S. Weinstein, D. Coco, S. Nagaraj, K.M. Kelly‐Scumpia, K.A. O'Malley, J.L. Wynn, S. Antonenko, S.Z. Al‐Quran, R. Swan, C.S. Chung, M.A. Atkinson, R. Ramphal, D.I. Gabrilovich, W.H. Reeves, A. Ayala, J. Phillips, D. Laface, P.G. Heyworth, M. Clare‐Salzler, and L.L. Moldawer. 2007. MyD88‐dependent expansion of an immature GR‐1(+)CD11b(+) population induces T cell suppression and Th2 polarization in sepsis. The Journal of experimental medicine 204:1463‐1474.  162  Deng, W., B. Thiel, C.S. Tannenbaum, T.A. Hamilton, and D.J. Stuehr. 1993. Synergistic cooperation between T cell lymphokines for induction of the nitric oxide synthase gene in murine peritoneal macrophages. J Immunol 151:322‐329. Dexter, D.L., H.M. Kowalski, B.A. Blazar, Z. Fligiel, R. Vogel, and G.H. Heppner. 1978. Heterogeneity of tumor cells from a single mouse mammary tumor. Cancer research 38:3174‐3181. Di Napoli, M.R., M. Torroella‐Kouri, G. Perry, and D.M. Lopez. 2005. Diminished PKC activity and decreased binding of transcription factors are involved in the impaired production of nitric oxide by macrophages from tumor‐bearing mice. Int J Mol Med 15:503‐511. Diaz‐Montero, C.M., M.L. Salem, M.I. Nishimura, E. Garrett‐Mayer, D.J. Cole, and A.J. Montero. 2009. Increased circulating myeloid‐derived suppressor cells correlate with clinical cancer stage, metastatic tumor burden, and doxorubicin‐ cyclophosphamide chemotherapy. Cancer Immunol Immunother 58:49‐59. Dinapoli, M.R., C.L. Calderon, and D.M. Lopez. 1996. The altered tumoricidal capacity of macrophages isolated from tumor‐bearing mice is related to reduce expression of the inducible nitric oxide synthase gene. The Journal of experimental medicine 183:1323‐1329. Doedens, A.L., C. Stockmann, M.P. Rubinstein, D. Liao, N. Zhang, D.G. DeNardo, L.M. Coussens, M. Karin, A.W. Goldrath, and R.S. Johnson. 2010. Macrophage expression of hypoxia‐inducible factor‐1 alpha suppresses T‐cell function and promotes tumor progression. Cancer research 70:7465‐7475. Dolcetti, L., E. Peranzoni, S. Ugel, I. Marigo, A. Fernandez Gomez, C. Mesa, M. Geilich, G. Winkels, E. Traggiai, A. Casati, F. Grassi, and V. Bronte. 2010. Hierarchy of immunosuppressive strength among myeloid‐derived suppressor cell subsets is determined by GM‐CSF. European journal of immunology 40:22‐35. Dunn, G.P., A.T. Bruce, H. Ikeda, L.J. Old, and R.D. Schreiber. 2002. Cancer immunoediting: from immunosurveillance to tumor escape. Nat Immunol 3:991‐ 998. Ehrlich, P. 1909. Über den jetzigen stand der karzinomforschung. Ned Tijdschr Geneeskd 5:273‐290. Elkabets, M., V.S. Ribeiro, C.A. Dinarello, S. Ostrand‐Rosenberg, J.P. Di Santo, R.N. Apte, and C.A. Vosshenrich. 2010. IL‐1beta regulates a novel myeloid‐derived suppressor cell subset that impairs NK cell development and function. European journal of immunology 40:3347‐3357.  163  Erler, J.T., K.L. Bennewith, T.R. Cox, G. Lang, D. Bird, A. Koong, Q.T. Le, and A.J. Giaccia. 2009. Hypoxia‐induced lysyl oxidase is a critical mediator of bone marrow cell recruitment to form the premetastatic niche. Cancer Cell 15:35‐44. Eruslanov, E., I. Daurkin, J. Ortiz, J. Vieweg, and S. Kusmartsev. 2010. Pivotal Advance: Tumor‐mediated induction of myeloid‐derived suppressor cells and M2‐ polarized macrophages by altering intracellular PGE catabolism in myeloid cells. Journal of leukocyte biology 88:839‐848. Eruslanov, E., and S. Kusmartsev. 2010. Identification of ROS using oxidized DCFDA and flow‐cytometry. Methods Mol Biol 594:57‐72. Fasano, M., S. Curry, E. Terreno, M. Galliano, G. Fanali, P. Narciso, S. Notari, and P. Ascenzi. 2005. The extraordinary ligand binding properties of human serum albumin. IUBMB life 57:787‐796. Fidler, I.J. 2003. The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited. Nat Rev Cancer 3:453‐458. Fitzner, N., S. Clauberg, F. Essmann, J. Liebmann, and V. Kolb‐Bachofen. 2008. Human skin endothelial cells can express all 10 TLR genes and respond to respective ligands. Clinical and vaccine immunology : CVI 15:138‐146. Fricke, I., N. Mirza, J. Dupont, C. Lockhart, A. Jackson, J.H. Lee, J.A. Sosman, and D.I. Gabrilovich. 2007. Vascular endothelial growth factor‐trap overcomes defects in dendritic cell differentiation but does not improve antigen‐specific immune responses. Clinical cancer research : an official journal of the American Association for Cancer Research 13:4840‐4848. Gabrilovich, D., T. Ishida, T. Oyama, S. Ran, V. Kravtsov, S. Nadaf, and D.P. Carbone. 1998. Vascular endothelial growth factor inhibits the development of dendritic cells and dramatically affects the differentiation of multiple hematopoietic lineages in vivo. Blood 92:4150‐4166. Gabrilovich, D.I., M.P. Velders, E.M. Sotomayor, and W.M. Kast. 2001. Mechanism of immune dysfunction in cancer mediated by immature Gr‐1+ myeloid cells. Journal of immunology 166:5398‐5406. Gallina, G., L. Dolcetti, P. Serafini, C. De Santo, I. Marigo, M.P. Colombo, G. Basso, F. Brombacher, I. Borrello, P. Zanovello, S. Bicciato, and V. Bronte. 2006. Tumors induce a subset of inflammatory monocytes with immunosuppressive activity on CD8+ T cells. J Clin Invest 116:2777‐2790. Gandhi, R., D.E. Anderson, and H.L. Weiner. 2007. Cutting Edge: Immature human dendritic cells express latency‐associated peptide and inhibit T cell activation in a TGF‐beta‐dependent manner. J Immunol 178:4017‐4021.  164  Geissmann, F., C. Auffray, R. Palframan, C. Wirrig, A. Ciocca, L. Campisi, E. Narni‐ Mancinelli, and G. Lauvau. 2008. Blood monocytes: distinct subsets, how they relate to dendritic cells, and their possible roles in the regulation of T‐cell responses. Immunol Cell Biol 86:398‐408. Geissmann, F., M.G. Manz, S. Jung, M.H. Sieweke, M. Merad, and K. Ley. 2010. Development of monocytes, macrophages, and dendritic cells. Science 327:656‐ 661. Godfrey, D.I., S. Stankovic, and A.G. Baxter. 2010. Raising the NKT cell family. Nat Immunol 11:197‐206. Gordon, S. 2003. Alternative activation of macrophages. Nat Rev Immunol 3:23‐35. Gordon, S., and F.O. Martinez. 2010. Alternative activation of macrophages: mechanism and functions. Immunity 32:593‐604. Gordon, S., and P.R. Taylor. 2005. Monocyte and macrophage heterogeneity. Nature reviews. Immunology 5:953‐964. Gough, D.J., N.L. Messina, L. Hii, J.A. Gould, K. Sabapathy, A.P. Robertson, J.A. Trapani, D.E. Levy, P.J. Hertzog, C.J. Clarke, and R.W. Johnstone. 2010. Functional crosstalk between type I and II interferon through the regulated expression of STAT1. PLoS Biol 8:e1000361. Greenwald, R.J., G.J. Freeman, and A.H. Sharpe. 2005. The B7 family revisited. Annu Rev Immunol 23:515‐548. Greifenberg, V., E. Ribechini, S. Rossner, and M.B. Lutz. 2009. Myeloid‐derived suppressor cell activation by combined LPS and IFN‐gamma treatment impairs DC development. European journal of immunology 39:2865‐2876. Grimshaw, M.J., J.L. Wilson, and F.R. Balkwill. 2002. Endothelin‐2 is a macrophage chemoattractant: implications for macrophage distribution in tumors. European journal of immunology 32:2393‐2400. Guerra, N., Y.X. Tan, N.T. Joncker, A. Choy, F. Gallardo, N. Xiong, S. Knoblaugh, D. Cado, N.M. Greenberg, and D.H. Raulet. 2008. NKG2D‐deficient mice are defective in tumor surveillance in models of spontaneous malignancy. Immunity 28:571‐580. Guth, A.M., W.J. Janssen, C.M. Bosio, E.C. Crouch, P.M. Henson, and S.W. Dow. 2009. Lung environment determines unique phenotype of alveolar macrophages. Am J Physiol Lung Cell Mol Physiol 296:L936‐946.  165  Habibi, M., M. Kmieciak, L. Graham, J.K. Morales, H.D. Bear, and M.H. Manjili. 2009. Radiofrequency thermal ablation of breast tumors combined with intralesional administration of IL‐7 and IL‐15 augments anti‐tumor immune responses and inhibits tumor development and metastasis. Breast Cancer Res Treat 114:423‐ 431. Hagemann, T., T. Lawrence, I. McNeish, K.A. Charles, H. Kulbe, R.G. Thompson, S.C. Robinson, and F.R. Balkwill. 2008. "Re‐educating" tumor‐associated macrophages by targeting NF‐kappaB. J Exp Med 205:1261‐1268. Hamilton, M.J., F. Antignano, A. von Rossum, J.L. Boucher, K.L. Bennewith, and G. Krystal. 2010. TLR agonists that induce IFN‐beta abrogate resident macrophage suppression of T cells. J Immunol 185:4545‐4553. Hamilton, M.J., J.P. Banath, V. Lam, N.E. Lepard, G. Krystal, and K.L. Bennewith. 2011. Serum inhibits the immunosuppressive function of myeloid‐derived suppressor cells isolated from 4T1 tumor‐bearing mice. Cancer Immunol Immunother In Press. Hanahan, D., and R.A. Weinberg. 2011. Hallmarks of cancer: the next generation. Cell 144:646‐674. Handel‐Fernandez, M.E., X. Cheng, L.M. Herbert, and D.M. Lopez. 1997. Down‐regulation of IL‐12, not a shift from a T helper‐1 to a T helper‐2 phenotype, is responsible for impaired IFN‐gamma production in mammary tumor‐bearing mice. Journal of immunology 158:280‐286. Hanson, E.M., V.K. Clements, P. Sinha, D. Ilkovitch, and S. Ostrand‐Rosenberg. 2009. Myeloid‐derived suppressor cells down‐regulate L‐selectin expression on CD4+ and CD8+ T cells. Journal of immunology 183:937‐944. Haverkamp, J.M., S.A. Crist, B.D. Elzey, C. Cimen, and T.L. Ratliff. 2011. In vivo suppressive function of myeloid‐derived suppressor cells is limited to the inflammatory site. Eur J Immunol 41:749‐759. Heppner, G.H., F.R. Miller, and P.M. Shekhar. 2000. Nontransgenic models of breast cancer. Breast Cancer Res 2:331‐334. Ho, V.W., and L.M. Sly. 2009. Derivation and characterization of murine alternatively activated (M2) macrophages. Methods Mol Biol 531:173‐185. Holcomb, I.N., R.C. Kabakoff, B. Chan, T.W. Baker, A. Gurney, W. Henzel, C. Nelson, H.B. Lowman, B.D. Wright, N.J. Skelton, G.D. Frantz, D.B. Tumas, F.V. Peale, Jr., D.L. Shelton, and C.C. Hebert. 2000. FIZZ1, a novel cysteine‐rich secreted protein associated with pulmonary inflammation, defines a new gene family. EMBO J 19:4046‐4055.  166  Holda, J.H., T. Maier, and H.N. Claman. 1985. Murine graft‐versus‐host disease across minor barriers: immunosuppressive aspects of natural suppressor cells. Immunological reviews 88:87‐105. Hoves, S., S.W. Krause, C. Schutz, D. Halbritter, J. Scholmerich, H. Herfarth, and M. Fleck. 2006. Monocyte‐derived human macrophages mediate anergy in allogeneic T cells and induce regulatory T cells. J Immunol 177:2691‐2698. Huang, B., P.Y. Pan, Q. Li, A.I. Sato, D.E. Levy, J. Bromberg, C.M. Divino, and S.H. Chen. 2006. Gr‐1+CD115+ immature myeloid suppressor cells mediate the development of tumor‐induced T regulatory cells and T‐cell anergy in tumor‐ bearing host. Cancer Res 66:1123‐1131. Husemann, Y., J.B. Geigl, F. Schubert, P. Musiani, M. Meyer, E. Burghart, G. Forni, R. Eils, T. Fehm, G. Riethmuller, and C.A. Klein. 2008. Systemic spread is an early step in breast cancer. Cancer Cell 13:58‐68. Ilkovitch, D., and D.M. Lopez. 2009. The liver is a site for tumor‐induced myeloid‐ derived suppressor cell accumulation and immunosuppression. Cancer research 69:5514‐5521. Inaba, K., M. Kitaura, T. Kato, Y. Watanabe, Y. Kawade, and S. Muramatsu. 1986. Contrasting effect of alpha/beta‐ and gamma‐interferons on expression of macrophage Ia antigens. J Exp Med 163:1030‐1035. Joyce, J.A., and J.W. Pollard. 2009. Microenvironmental regulation of metastasis. Nat Rev Cancer 9:239‐252. Kaplan, R.N., B. Psaila, and D. Lyden. 2006. Bone marrow cells in the 'pre‐metastatic niche': within bone and beyond. Cancer Metastasis Rev 25:521‐529. Kaplan, R.N., R.D. Riba, S. Zacharoulis, A.H. Bramley, L. Vincent, C. Costa, D.D. MacDonald, D.K. Jin, K. Shido, S.A. Kerns, Z. Zhu, D. Hicklin, Y. Wu, J.L. Port, N. Altorki, E.R. Port, D. Ruggero, S.V. Shmelkov, K.K. Jensen, S. Rafii, and D. Lyden. 2005. VEGFR1‐positive haematopoietic bone marrow progenitors initiate the pre‐metastatic niche. Nature 438:820‐827. Karin, M., and F.R. Greten. 2005. NF‐kappaB: linking inflammation and immunity to cancer development and progression. Nature reviews. Immunology 5:749‐759. Kennedy, A.D., and F.R. DeLeo. 2009. Neutrophil apoptosis and the resolution of infection. Immunol Res 43:25‐61. Kleinbongard, P., T. Rassaf, A. Dejam, S. Kerber, and M. Kelm. 2002. Griess method for nitrite measurement of aqueous and protein‐containing samples. Methods Enzymol 359:158‐168.  167  Ko, J.S., P. Rayman, J. Ireland, S. Swaidani, G. Li, K.D. Bunting, B. Rini, J.H. Finke, and P.A. Cohen. 2010. Direct and differential suppression of myeloid‐derived suppressor cell subsets by sunitinib is compartmentally constrained. Cancer Res 70:3526‐ 3536. Kodumudi, K.N., K. Woan, D.L. Gilvary, E. Sahakian, S. Wei, and J.Y. Djeu. 2010. A novel chemoimmunomodulating property of docetaxel: suppression of myeloid‐ derived suppressor cells in tumor bearers. Clin Cancer Res 16:4583‐4594. Kornstein, M.J., J.S. Brooks, and D.E. Elder. 1983. Immunoperoxidase localization of lymphocyte subsets in the host response to melanoma and nevi. Cancer research 43:2749‐2753. Kryczek, I., L. Zou, P. Rodriguez, G. Zhu, S. Wei, P. Mottram, M. Brumlik, P. Cheng, T. Curiel, L. Myers, A. Lackner, X. Alvarez, A. Ochoa, L. Chen, and W. Zou. 2006. B7‐ H4 expression identifies a novel suppressive macrophage population in human ovarian carcinoma. J Exp Med 203:871‐881. Kuang, D.M., Q. Zhao, C. Peng, J. Xu, J.P. Zhang, C. Wu, and L. Zheng. 2009. Activated monocytes in peritumoral stroma of hepatocellular carcinoma foster immune privilege and disease progression through PD‐L1. The Journal of experimental medicine 206:1327‐1337. Kubota, Y., K. Takubo, T. Shimizu, H. Ohno, K. Kishi, M. Shibuya, H. Saya, and T. Suda. 2009. M‐CSF inhibition selectively targets pathological angiogenesis and lymphangiogenesis. The Journal of experimental medicine 206:1089‐1102. Kung, J.T., S.B. Brooks, J.P. Jakway, L.L. Leonard, and D.W. Talmage. 1977. Suppression of in vitro cytotoxic response by macrophages due to induced arginase. J Exp Med 146:665‐672. Kuroda, E., V. Ho, J. Ruschmann, F. Antignano, M. Hamilton, M.J. Rauh, A. Antov, R.A. Flavell, L.M. Sly, and G. Krystal. 2009. SHIP represses the generation of IL‐3‐ induced M2 macrophages by inhibiting IL‐4 production from basophils. J Immunol 183:3652‐3660. Kusmartsev, S., F. Cheng, B. Yu, Y. Nefedova, E. Sotomayor, R. Lush, and D. Gabrilovich. 2003. All‐trans‐retinoic acid eliminates immature myeloid cells from tumor‐ bearing mice and improves the effect of vaccination. Cancer research 63:4441‐ 4449. Kusmartsev, S., and D.I. Gabrilovich. 2003. Inhibition of myeloid cell differentiation in cancer: the role of reactive oxygen species. Journal of leukocyte biology 74:186‐ 196.  168  Kusmartsev, S., Y. Nefedova, D. Yoder, and D.I. Gabrilovich. 2004. Antigen‐specific inhibition of CD8+ T cell response by immature myeloid cells in cancer is mediated by reactive oxygen species. Journal of immunology 172:989‐999. Kusmartsev, S., Z. Su, A. Heiser, J. Dannull, E. Eruslanov, H. Kubler, D. Yancey, P. Dahm, and J. Vieweg. 2008. Reversal of myeloid cell‐mediated immunosuppression in patients with metastatic renal cell carcinoma. Clin Cancer Res 14:8270‐8278. Laoui, D., E. Van Overmeire, K. Movahedi, J. Van den Bossche, E. Schouppe, C. Mommer, A. Nikolaou, Y. Morias, P. De Baetselier, and J.A. Van Ginderachter. 2011. Mononuclear phagocyte heterogeneity in cancer: Different subsets and activation states reaching out at the tumor site. Immunobiology Le, H.K., L. Graham, E. Cha, J.K. Morales, M.H. Manjili, and H.D. Bear. 2009. Gemcitabine directly inhibits myeloid derived suppressor cells in BALB/c mice bearing 4T1 mammary carcinoma and augments expansion of T cells from tumor‐bearing mice. Int Immunopharmacol 9:900‐909. Leek, R.D., C.E. Lewis, R. Whitehouse, M. Greenall, J. Clarke, and A.L. Harris. 1996. Association of macrophage infiltration with angiogenesis and prognosis in invasive breast carcinoma. Cancer research 56:4625‐4629. Levings, M.K., and M.G. Roncarolo. 2000. T‐regulatory 1 cells: a novel subset of CD4 T cells with immunoregulatory properties. J Allergy Clin Immunol 106:S109‐112. Lin, E.Y., A.V. Nguyen, R.G. Russell, and J.W. Pollard. 2001. Colony‐stimulating factor 1 promotes progression of mammary tumors to malignancy. The Journal of experimental medicine 193:727‐740. Ling, P.D., M.K. Warren, and S.N. Vogel. 1985. Antagonistic effect of interferon‐beta on the interferon‐gamma‐induced expression of Ia antigen in murine macrophages. J Immunol 135:1857‐1863. Lorsbach, R.B., and S.W. Russell. 1992. A specific sequence of stimulation is required to induce synthesis of the antimicrobial molecule nitric oxide by mouse macrophages. Infect Immun 60:2133‐2135. Luzzi, K.J., I.C. MacDonald, E.E. Schmidt, N. Kerkvliet, V.L. Morris, A.F. Chambers, and A.C. Groom. 1998. Multistep nature of metastatic inefficiency: dormancy of solitary cells after successful extravasation and limited survival of early micrometastases. Am J Pathol 153:865‐873. Lyons, A.B. 2000. Analysing cell division in vivo and in vitro using flow cytometric measurement of CFSE dye dilution. J Immunol Methods 243:147‐154.  169  Mace, E.M., S.J. Monkley, D.R. Critchley, and F. Takei. 2009. A dual role for talin in NK cell cytotoxicity: activation of LFA‐1‐mediated cell adhesion and polarization of NK cells. J Immunol 182:948‐956. Maltby, S., K. Khazaie, and K.M. McNagny. 2009. Mast cells in tumor growth: angiogenesis, tissue remodelling and immune‐modulation. Biochim Biophys Acta 1796:19‐26. Mandruzzato, S., S. Solito, E. Falisi, S. Francescato, V. Chiarion‐Sileni, S. Mocellin, A. Zanon, C.R. Rossi, D. Nitti, V. Bronte, and P. Zanovello. 2009. IL4Ralpha+ myeloid‐ derived suppressor cell expansion in cancer patients. J Immunol 182:6562‐6568. Mantovani, A. 2009. Cancer: Inflaming metastasis. Nature 457:36‐37. Mantovani, A., M.A. Cassatella, C. Costantini, and S. Jaillon. 2011. Neutrophils in the activation and regulation of innate and adaptive immunity. Nature reviews. Immunology 11:519‐531. Mantovani, A., and A. Sica. 2010. Macrophages, innate immunity and cancer: balance, tolerance, and diversity. Curr Opin Immunol 22:231‐237. Mantovani, A., A. Sica, S. Sozzani, P. Allavena, A. Vecchi, and M. Locati. 2004. The chemokine system in diverse forms of macrophage activation and polarization. Trends Immunol 25:677‐686. Mantovani, A., S. Sozzani, M. Locati, P. Allavena, and A. Sica. 2002. Macrophage polarization: tumor‐associated macrophages as a paradigm for polarized M2 mononuclear phagocytes. Trends Immunol 23:549‐555. Martin, C.W., and I.F. Muir. 1990. The role of lymphocytes in wound healing. Br J Plast Surg 43:655‐662. Martinez, F.O., S. Gordon, M. Locati, and A. Mantovani. 2006. Transcriptional profiling of the human monocyte‐to‐macrophage differentiation and polarization: new molecules and patterns of gene expression. Journal of immunology 177:7303‐ 7311. Martinez, F.O., A. Sica, A. Mantovani, and M. Locati. 2008. Macrophage activation and polarization. Front Biosci 13:453‐461. Matzinger, P. 1994. Tolerance, danger, and the extended family. Annual review of immunology 12:991‐1045. McGrath, M.S., and V. Kodelja. 1999. Balanced macrophage activation hypothesis: a biological model for development of drugs targeted at macrophage functional states. Pathobiology 67:277‐281.  170  McKee, A.S., M.W. Munks, and P. Marrack. 2007. How do adjuvants work? Important considerations for new generation adjuvants. Immunity 27:687‐690. Metzger, Z., J.T. Hoffeld, and J.J. Oppenheim. 1980. Macrophage‐mediated suppression. I. Evidence for participation of both hdyrogen peroxide and prostaglandins in suppression of murine lymphocyte proliferation. J Immunol 124:983‐988. Miller, F.R., B.E. Miller, and G.H. Heppner. 1983. Characterization of metastatic heterogeneity among subpopulations of a single mouse mammary tumor: heterogeneity in phenotypic stability. Invasion Metastasis 3:22‐31. Mirza, N., M. Fishman, I. Fricke, M. Dunn, A.M. Neuger, T.J. Frost, R.M. Lush, S. Antonia, and D.I. Gabrilovich. 2006. All‐trans‐retinoic acid improves differentiation of myeloid cells and immune response in cancer patients. Cancer research 66:9299‐ 9307. Monette, F.C., and G. Sigounas. 1990. Some observations on the growth requirements of multi‐potent stem cells under defined culture conditions. In The Biology of Hematopoiesis. N. Dainiak, E.P. Cronkite, R. McCaffrey, and R.K. Shadduck, editors. Wiley‐Liss, Inc., New York. 37‐48. Moresco, E.M., D. Lavine, and B. Beutler. 2011. Toll‐like receptors. Curr Biol 21:R488‐ 493. Morrison, D.C., and D.M. Jacobs. 1976. Binding of polymyxin B to the lipid A portion of bacterial lipopolysaccharides. Immunochemistry 13:813‐818. Mosser, D.M., and J.P. Edwards. 2008. Exploring the full spectrum of macrophage activation. Nature reviews. Immunology 8:958‐969. Movahedi, K., M. Guilliams, J. Van den Bossche, R. Van den Bergh, C. Gysemans, A. Beschin, P. De Baetselier, and J.A. Van Ginderachter. 2008. Identification of discrete tumor‐induced myeloid‐derived suppressor cell subpopulations with distinct T cell‐suppressive activity. Blood 111:4233‐4244. Movahedi, K., D. Laoui, C. Gysemans, M. Baeten, G. Stange, J. Van den Bossche, M. Mack, D. Pipeleers, P. In't Veld, P. De Baetselier, and J.A. Van Ginderachter. 2010. Different tumor microenvironments contain functionally distinct subsets of macrophages derived from Ly6C(high) monocytes. Cancer research 70:5728‐ 5739. Munn, D.H., and A.L. Mellor. 2003. Macrophages and the regulation of self‐reactive T cells. Curr Pharm Des 9:257‐264. Munn, D.H., E. Shafizadeh, J.T. Attwood, I. Bondarev, A. Pashine, and A.L. Mellor. 1999. Inhibition of T cell proliferation by macrophage tryptophan catabolism. J Exp Med 189:1363‐1372. 171  Murdoch, C., M. Muthana, S.B. Coffelt, and C.E. Lewis. 2008. The role of myeloid cells in the promotion of tumour angiogenesis. Nat Rev Cancer 8:618‐631. Murphy, K.M., A.B. Heimberger, and D.Y. Loh. 1990. Induction by antigen of intrathymic apoptosis of CD4+CD8+TCRlo thymocytes in vivo. Science 250:1720‐1723. Nagaraj, S., and D.I. Gabrilovich. 2010. Myeloid‐derived suppressor cells in human cancer. Cancer J 16:348‐353. Nagaraj, S., K. Gupta, V. Pisarev, L. Kinarsky, S. Sherman, L. Kang, D.L. Herber, J. Schneck, and D.I. Gabrilovich. 2007. Altered recognition of antigen is a mechanism of CD8+ T cell tolerance in cancer. Nat Med 13:828‐835. Nausch, N., I.E. Galani, E. Schlecker, and A. Cerwenka. 2008. Mononuclear myeloid‐ derived "suppressor" cells express RAE‐1 and activate natural killer cells. Blood 112:4080‐4089. Nefedova, Y., M. Fishman, S. Sherman, X. Wang, A.A. Beg, and D.I. Gabrilovich. 2007. Mechanism of all‐trans retinoic acid effect on tumor‐associated myeloid‐derived suppressor cells. Cancer research 67:11021‐11028. Nurieva, R.I., X. Liu, and C. Dong. 2011. Molecular mechanisms of T‐cell tolerance. Immunological reviews 241:133‐144. Ochoa, A.C., A.H. Zea, C. Hernandez, and P.C. Rodriguez. 2007. Arginase, prostaglandins, and myeloid‐derived suppressor cells in renal cell carcinoma. Clin Cancer Res 13:721s‐726s. Ostrand‐Rosenberg, S. 2010. Myeloid‐derived suppressor cells: more mechanisms for inhibiting antitumor immunity. Cancer Immunol Immunother 59:1593‐1600. Ostrand‐Rosenberg, S., and P. Sinha. 2009. Myeloid‐derived suppressor cells: linking inflammation and cancer. Journal of immunology 182:4499‐4506. Paget, S. 1889. The distribution of secondary growths in cancer of the breast. The Lancet 133:571‐573. Parkin, J., and B. Cohen. 2001. An overview of the immune system. Lancet 357:1777‐ 1789. Pastula, A., and J. Marcinkiewicz. 2011. Myeloid‐derived suppressor cells: a double‐ edged sword? Int J Exp Pathol 92:73‐78. Peranzoni, E., S. Zilio, I. Marigo, L. Dolcetti, P. Zanovello, S. Mandruzzato, and V. Bronte. 2010. Myeloid‐derived suppressor cell heterogeneity and subset definition. Curr Opin Immunol 22:238‐244.  172  Pollard, J.W. 2009. Trophic macrophages in development and disease. Nature reviews. Immunology 9:259‐270. Pompella, A., A. Visvikis, A. Paolicchi, V. De Tata, and A.F. Casini. 2003. The changing faces of glutathione, a cellular protagonist. Biochem Pharmacol 66:1499‐1503. Porta, C., M. Rimoldi, G. Raes, L. Brys, P. Ghezzi, D. Di Liberto, F. Dieli, S. Ghisletti, G. Natoli, P. De Baetselier, A. Mantovani, and A. Sica. 2009. Tolerance and M2 (alternative) macrophage polarization are related processes orchestrated by p50 nuclear factor kappaB. Proc Natl Acad Sci U S A 106:14978‐14983. Prabhu, V., and C. Guruvayoorappan. 2010. Nitric oxide: pros and cons in tumor progression. Immunopharmacol Immunotoxicol 32:387‐392. Prehn, R.T., and J.M. Main. 1957. Immunity to methylcholanthrene‐induced sarcomas. J Natl Cancer Inst 18:769‐778. Priceman, S.J., J.L. Sung, Z. Shaposhnik, J.B. Burton, A.X. Torres‐Collado, D.L. Moughon, M. Johnson, A.J. Lusis, D.A. Cohen, M.L. Iruela‐Arispe, and L. Wu. 2010. Targeting distinct tumor‐infiltrating myeloid cells by inhibiting CSF‐1 receptor: combating tumor evasion of antiangiogenic therapy. Blood 115:1461‐1471. Psaila, B., R.N. Kaplan, E.R. Port, and D. Lyden. 2006. Priming the 'soil' for breast cancer metastasis: the pre‐metastatic niche. Breast Dis 26:65‐74. Psaila, B., and D. Lyden. 2009. The metastatic niche: adapting the foreign soil. Nat Rev Cancer 9:285‐293. Pulaski, B.A., and S. Ostrand‐Rosenberg. 2001. Mouse 4T1 breast tumor model. Curr Protoc Immunol Chapter 20:Unit 20 22. Punturieri, A., R.S. Alviani, T. Polak, P. Copper, J. Sonstein, and J.L. Curtis. 2004. Specific engagement of TLR4 or TLR3 does not lead to IFN‐beta‐mediated innate signal amplification and STAT1 phosphorylation in resident murine alveolar macrophages. Journal of immunology 173:1033‐1042. Qian, B., Y. Deng, J.H. Im, R.J. Muschel, Y. Zou, J. Li, R.A. Lang, and J.W. Pollard. 2009. A distinct macrophage population mediates metastatic breast cancer cell extravasation, establishment and growth. PLoS One 4:e6562. Qian, B.Z., J. Li, H. Zhang, T. Kitamura, J. Zhang, L.R. Campion, E.A. Kaiser, L.A. Snyder, and J.W. Pollard. 2011. CCL2 recruits inflammatory monocytes to facilitate breast‐tumour metastasis. Nature 475:222‐225. Qian, B.Z., and J.W. Pollard. 2010. Macrophage diversity enhances tumor progression and metastasis. Cell 141:39‐51.  173  Raes, G., P. De Baetselier, W. Noel, A. Beschin, F. Brombacher, and G. Hassanzadeh Gh. 2002. Differential expression of FIZZ1 and Ym1 in alternatively versus classically activated macrophages. J Leukoc Biol 71:597‐602. Rauh, M.J., V. Ho, C. Pereira, A. Sham, L.M. Sly, V. Lam, L. Huxham, A.I. Minchinton, A. Mui, and G. Krystal. 2005. SHIP represses the generation of alternatively activated macrophages. Immunity 23:361‐374. Rayamajhi, M., J. Humann, K. Penheiter, K. Andreasen, and L.L. Lenz. 2010. Induction of IFN‐alphabeta enables Listeria monocytogenes to suppress macrophage activation by IFN‐gamma. J Exp Med 207:327‐337. Roche, M., P. Rondeau, N.R. Singh, E. Tarnus, and E. Bourdon. 2008. The antioxidant properties of serum albumin. FEBS letters 582:1783‐1787. Rodriguez, P.C., C.P. Hernandez, D. Quiceno, S.M. Dubinett, J. Zabaleta, J.B. Ochoa, J. Gilbert, and A.C. Ochoa. 2005. Arginase I in myeloid suppressor cells is induced by COX‐2 in lung carcinoma. J Exp Med 202:931‐939. Rodriguez, P.C., A.H. Zea, K.S. Culotta, J. Zabaleta, J.B. Ochoa, and A.C. Ochoa. 2002. Regulation of T cell receptor CD3zeta chain expression by L‐arginine. J Biol Chem 277:21123‐21129. Rodriguez, P.C., A.H. Zea, J. DeSalvo, K.S. Culotta, J. Zabaleta, D.G. Quiceno, J.B. Ochoa, and A.C. Ochoa. 2003. L‐arginine consumption by macrophages modulates the expression of CD3 zeta chain in T lymphocytes. J Immunol 171:1232‐1239. Sansone, P., and J. Bromberg. 2011. Environment, inflammation, and cancer. Curr Opin Genet Dev 21:80‐85. Sanz, M.A., and F. Lo‐Coco. 2011. Modern approaches to treating acute promyelocytic leukemia. J Clin Oncol 29:495‐503. Schmidt‐Kittler, O., T. Ragg, A. Daskalakis, M. Granzow, A. Ahr, T.J. Blankenstein, M. Kaufmann, J. Diebold, H. Arnholdt, P. Muller, J. Bischoff, D. Harich, G. Schlimok, G. Riethmuller, R. Eils, and C.A. Klein. 2003. From latent disseminated cells to overt metastasis: genetic analysis of systemic breast cancer progression. Proc Natl Acad Sci U S A 100:7737‐7742. Schreiber, R.D., L.J. Old, and M.J. Smyth. 2011. Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion. Science 331:1565‐1570. Schroder, K., P.J. Hertzog, T. Ravasi, and D.A. Hume. 2004. Interferon‐gamma: an overview of signals, mechanisms and functions. J Leukoc Biol 75:163‐189.  174  Seifert, C.F., and B.H. Resman‐Targoff. 2006. Clinical laboratory tests and interpretation. In Textbook of Therapeutics: Drug and Disease Management. R.A. Helms, D.J. Quan, E.T. Herfindal, and D.R. Gourley, editors. Lippincott Williams & Wilkins, Philadelphia. 91‐115. Serafini, P. 2010. Editorial: PGE2‐producing MDSC: a role in tumor progression? Journal of leukocyte biology 88:827‐829. Serafini, P., S. Mgebroff, K. Noonan, and I. Borrello. 2008. Myeloid‐derived suppressor cells promote cross‐tolerance in B‐cell lymphoma by expanding regulatory T cells. Cancer research 68:5439‐5449. Shankaran, V., H. Ikeda, A.T. Bruce, J.M. White, P.E. Swanson, L.J. Old, and R.D. Schreiber. 2001. IFNgamma and lymphocytes prevent primary tumour development and shape tumour immunogenicity. Nature 410:1107‐1111. Shiku, H., P. Kisielow, M.A. Bean, T. Takahashi, E.A. Boyse, H.F. Oettgen, and L.J. Old. 1975. Expression of T‐cell differentiation antigens on effector cells in cell‐ mediated cytotoxicity in vitro. Evidence for functional heterogeneity related to the surface phenotype of T cells. The Journal of experimental medicine 141:227‐ 241. Si, Y., C.L. Tsou, K. Croft, and I.F. Charo. 2010. CCR2 mediates hematopoietic stem and progenitor cell trafficking to sites of inflammation in mice. The Journal of clinical investigation 120:1192‐1203. Sims, G.P., D.C. Rowe, S.T. Rietdijk, R. Herbst, and A.J. Coyle. 2010. HMGB1 and RAGE in inflammation and cancer. Annual review of immunology 28:367‐388. Singh, D., P.G. Febbo, K. Ross, D.G. Jackson, J. Manola, C. Ladd, P. Tamayo, A.A. Renshaw, A.V. D'Amico, J.P. Richie, E.S. Lander, M. Loda, P.W. Kantoff, T.R. Golub, and W.R. Sellers. 2002. Gene expression correlates of clinical prostate cancer behavior. Cancer Cell 1:203‐209. Sinha, P., V.K. Clements, S.K. Bunt, S.M. Albelda, and S. Ostrand‐Rosenberg. 2007a. Cross‐ talk between myeloid‐derived suppressor cells and macrophages subverts tumor immunity toward a type 2 response. Journal of immunology 179:977‐983. Sinha, P., V.K. Clements, A.M. Fulton, and S. Ostrand‐Rosenberg. 2007b. Prostaglandin E2 promotes tumor progression by inducing myeloid‐derived suppressor cells. Cancer research 67:4507‐4513. Sinha, P., V.K. Clements, S. Miller, and S. Ostrand‐Rosenberg. 2005a. Tumor immunity: a balancing act between T cell activation, macrophage activation and tumor‐ induced immune suppression. Cancer immunology, immunotherapy : CII 54:1137‐1142.  175  Sinha, P., V.K. Clements, and S. Ostrand‐Rosenberg. 2005b. Interleukin‐13‐regulated M2 macrophages in combination with myeloid suppressor cells block immune surveillance against metastasis. Cancer Res 65:11743‐11751. Sinha, P., V.K. Clements, and S. Ostrand‐Rosenberg. 2005c. Reduction of myeloid‐ derived suppressor cells and induction of M1 macrophages facilitate the rejection of established metastatic disease. J Immunol 174:636‐645. Sinha, P., C. Okoro, D. Foell, H.H. Freeze, S. Ostrand‐Rosenberg, and G. Srikrishna. 2008. Proinflammatory S100 proteins regulate the accumulation of myeloid‐derived suppressor cells. J Immunol 181:4666‐4675. Siveen, K.S., and G. Kuttan. 2009. Role of macrophages in tumour progression. Immunol Lett 123:97‐102. Sly, L.M., M.J. Hamilton, E. Kuroda, V.W. Ho, F.L. Antignano, S.L. Omeis, C.J. van Netten‐ Thomas, D. Wong, H.K. Brugger, O. Williams, M.E. Feldman, B.T. Houseman, D. Fiedler, K.M. Shokat, and G. Krystal. 2009. SHIP prevents lipopolysaccharide from triggering an antiviral response in mice. Blood 113:2945‐2954. Sly, L.M., J. Kalesnikoff, V. Lam, D. Wong, C. Song, S. Omeis, K. Chan, C.W. Lee, R.P. Siraganian, J. Rivera, and G. Krystal. 2008. IgE‐induced mast cell survival requires the prolonged generation of reactive oxygen species. J Immunol 181:3850‐3860. Soehnlein, O., and L. Lindbom. 2010. Phagocyte partnership during the onset and resolution of inflammation. Nature reviews. Immunology 10:427‐439. Solinas, G., S. Schiarea, M. Liguori, M. Fabbri, S. Pesce, L. Zammataro, F. Pasqualini, M. Nebuloni, C. Chiabrando, A. Mantovani, and P. Allavena. 2010. Tumor‐ conditioned macrophages secrete migration‐stimulating factor: a new marker for M2‐polarization, influencing tumor cell motility. Journal of immunology 185:642‐652. Srivastava, M.K., P. Sinha, V.K. Clements, P. Rodriguez, and S. Ostrand‐Rosenberg. 2010. Myeloid‐derived suppressor cells inhibit T‐cell activation by depleting cystine and cysteine. Cancer Res 70:68‐77. Stein, M., S. Keshav, N. Harris, and S. Gordon. 1992. Interleukin 4 potently enhances murine macrophage mannose receptor activity: a marker of alternative immunologic macrophage activation. The Journal of experimental medicine 176:287‐292. Strickland, D., U.R. Kees, and P.G. Holt. 1996. Regulation of T‐cell activation in the lung: alveolar macrophages induce reversible T‐cell anergy in vitro associated with inhibition of interleukin‐2 receptor signal transduction. Immunology 87:250‐ 258. 176  Strober, S. 1984. Natural suppressor (NS) cells, neonatal tolerance, and total lymphoid irradiation: exploring obscure relationships. Annual review of immunology 2:219‐237. Stutman, O. 1974. Tumor development after 3‐methylcholanthrene in immunologically deficient athymic‐nude mice. Science 183:534‐536. Stutman, O. 1975. Immunodepression and malignancy. Adv Cancer Res 22:261‐422. Suit, H.D., R.S. Sedlacek, G. Silver, and D. Dosoretz. 1985. Pentobarbital anesthesia and the response of tumor and normal tissue in the C3Hf/sed mouse to radiation. Radiat Res 104:47‐65. Taylor‐Robinson, A.W., F.Y. Liew, A. Severn, D. Xu, S.J. McSorley, P. Garside, J. Padron, and R.S. Phillips. 1994. Regulation of the immune response by nitric oxide differentially produced by T helper type 1 and T helper type 2 cells. Eur J Immunol 24:980‐984. Torroella‐Kouri, M., X. Ma, G. Perry, M. Ivanova, P.J. Cejas, J.L. Owen, V. Iragavarapu‐ Charyulu, and D.M. Lopez. 2005. Diminished expression of transcription factors nuclear factor kappaB and CCAAT/enhancer binding protein underlies a novel tumor evasion mechanism affecting macrophages of mammary tumor‐bearing mice. Cancer research 65:10578‐10584. Torroella‐Kouri, M., R. Silvera, D. Rodriguez, R. Caso, A. Shatry, S. Opiela, D. Ilkovitch, R.A. Schwendener, V. Iragavarapu‐Charyulu, Y. Cardentey, N. Strbo, and D.M. Lopez. 2009. Identification of a subpopulation of macrophages in mammary tumor‐bearing mice that are neither M1 nor M2 and are less differentiated. Cancer research 69:4800‐4809. Ueda, Y., D.W. Cain, M. Kuraoka, M. Kondo, and G. Kelsoe. 2009. IL‐1R type I‐dependent hemopoietic stem cell proliferation is necessary for inflammatory granulopoiesis and reactive neutrophilia. Journal of immunology 182:6477‐6484. Ueha, S., F.H. Shand, and K. Matsushima. 2011. Myeloid cell population dynamics in healthy and tumor‐bearing mice. International immunopharmacology 11:783‐ 788. Valenti, R., V. Huber, P. Filipazzi, L. Pilla, G. Sovena, A. Villa, A. Corbelli, S. Fais, G. Parmiani, and L. Rivoltini. 2006. Human tumor‐released microvesicles promote the differentiation of myeloid cells with transforming growth factor‐beta‐ mediated suppressive activity on T lymphocytes. Cancer Res 66:9290‐9298. van 't Veer, L.J., H. Dai, M.J. van de Vijver, Y.D. He, A.A. Hart, M. Mao, H.L. Peterse, K. van der Kooy, M.J. Marton, A.T. Witteveen, G.J. Schreiber, R.M. Kerkhoven, C. Roberts, P.S. Linsley, R. Bernards, and S.H. Friend. 2002. Gene expression profiling predicts clinical outcome of breast cancer. Nature 415:530‐536. 177  Van Ginderachter, J.A., A. Beschin, P. De Baetselier, and G. Raes. 2010. Myeloid‐derived suppressor cells in parasitic infections. European journal of immunology 40:2976‐2985. Van Pel, A., and T. Boon. 1982. Protection against a nonimmunogenic mouse leukemia by an immunogenic variant obtained by mutagenesis. Proc Natl Acad Sci U S A 79:4718‐4722. Vesely, M.D., M.H. Kershaw, R.D. Schreiber, and M.J. Smyth. 2011. Natural innate and adaptive immunity to cancer. Annual review of immunology 29:235‐271. Vieweg, J., Z. Su, P. Dahm, and S. Kusmartsev. 2007. Reversal of tumor‐mediated immunosuppression. Clin Cancer Res 13:727s‐732s. von Bergwelt‐Baildon, M.S., A. Popov, T. Saric, J. Chemnitz, S. Classen, M.S. Stoffel, F. Fiore, U. Roth, M. Beyer, S. Debey, C. Wickenhauser, F.G. Hanisch, and J.L. Schultze. 2006. CD25 and indoleamine 2,3‐dioxygenase are up‐regulated by prostaglandin E2 and expressed by tumor‐associated dendritic cells in vivo: additional mechanisms of T‐cell inhibition. Blood 108:228‐237. von Kleist, S., J. Berling, W. Bohle, and C. Wittekind. 1987. Immunohistological analysis of lymphocyte subpopulations infiltrating breast carcinomas and benign lesions. Int J Cancer 40:18‐23. Waltenbaugh, C., T. Doan, R. Melvold, and S. Viselli. 2008. Immunology. Wolters Kluwer Health/Lippincott Williams & Wilkins, Philadelphia. Wang, Q., H. Ni, L. Lan, X. Wei, R. Xiang, and Y. Wang. 2010. Fra‐1 protooncogene regulates IL‐6 expression in macrophages and promotes the generation of M2d macrophages. Cell Res 20:701‐712. Watson, G.A., and D.M. Lopez. 1995. Aberrant antigen presentation by macrophages from tumor‐bearing mice is involved in the down‐regulation of their T cell responses. Journal of immunology 155:3124‐3134. Wei, X.Q., I.G. Charles, A. Smith, J. Ure, G.J. Feng, F.P. Huang, D. Xu, W. Muller, S. Moncada, and F.Y. Liew. 1995. Altered immune responses in mice lacking inducible nitric oxide synthase. Nature 375:408‐411. Weinstein, S.L., A.J. Finn, S.H. Dave, F. Meng, C.A. Lowell, J.S. Sanghera, and A.L. DeFranco. 2000. Phosphatidylinositol 3‐kinase and mTOR mediate lipopolysaccharide‐stimulated nitric oxide production in macrophages via interferon‐beta. Journal of leukocyte biology 67:405‐414. Whiteside, T.L. 2006. Immune suppression in cancer: effects on immune cells, mechanisms and future therapeutic intervention. Semin Cancer Biol 16:3‐15.  178  Whiteside, T.L. 2008. The tumor microenvironment and its role in promoting tumor growth. Oncogene 27:5904‐5912. Whiteside, T.L. 2010. Immune responses to malignancies. The Journal of allergy and clinical immunology 125:S272‐283. Xia, Y., L.J. Roman, B.S. Masters, and J.L. Zweier. 1998. Inducible nitric‐oxide synthase generates superoxide from the reductase domain. The Journal of biological chemistry 273:22635‐22639. Yang, L., L.M. DeBusk, K. Fukuda, B. Fingleton, B. Green‐Jarvis, Y. Shyr, L.M. Matrisian, D.P. Carbone, and P.C. Lin. 2004. Expansion of myeloid immune suppressor Gr+CD11b+ cells in tumor‐bearing host directly promotes tumor angiogenesis. Cancer Cell 6:409‐421. Yang, L., J. Huang, X. Ren, A.E. Gorska, A. Chytil, M. Aakre, D.P. Carbone, L.M. Matrisian, A. Richmond, P.C. Lin, and H.L. Moses. 2008. Abrogation of TGF beta signaling in mammary carcinomas recruits Gr‐1+CD11b+ myeloid cells that promote metastasis. Cancer Cell 13:23‐35. Yoshida, R., H.W. Murray, and C.F. Nathan. 1988. Agonist and antagonist effects of interferon alpha and beta on activation of human macrophages. Two classes of interferon gamma receptors and blockade of the high‐affinity sites by interferon alpha or beta. J Exp Med 167:1171‐1185. Youn, J.I., and D.I. Gabrilovich. 2010. The biology of myeloid‐derived suppressor cells: the blessing and the curse of morphological and functional heterogeneity. Eur J Immunol 40:2969‐2975. Youn, J.I., S. Nagaraj, M. Collazo, and D.I. Gabrilovich. 2008. Subsets of myeloid‐derived suppressor cells in tumor‐bearing mice. J Immunol 181:5791‐5802. Young, M.R., and D.M. Lathers. 1999. Myeloid progenitor cells mediate immune suppression in patients with head and neck cancers. Int J Immunopharmacol 21:241‐252. Zea, A.H., P.C. Rodriguez, M.B. Atkins, C. Hernandez, S. Signoretti, J. Zabaleta, D. McDermott, D. Quiceno, A. Youmans, A. O'Neill, J. Mier, and A.C. Ochoa. 2005. Arginase‐producing myeloid suppressor cells in renal cell carcinoma patients: a mechanism of tumor evasion. Cancer Res 65:3044‐3048. Zhang, F., S.S. Lau, and T.J. Monks. 2011. The cytoprotective effect of N‐acetyl‐L‐cysteine against ROS‐induced cytotoxicity is independent of its ability to enhance glutathione synthesis. Toxicol Sci 120:87‐97.  179  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0072452/manifest

Comment

Related Items