CROSS PRESENTATION OF ORAL ANTIGENS FOR INDUCTION OF CD8+ T CELL RESPONSES by Ana Luisa Chávez Steenbock B.Sc., The University of British Columbia, 2009 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in The Faculty of Graduate Studies (Zoology) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) May 2012 © Ana Luisa Chávez Steenbock, 2012 ii Abstract The biggest challenge of intestinal immunity is inducing tolerance towards harmless antigens derived from food and commensal bacteria while maintaining protection against dangerous pathogens. The dendritic cells (DCs) found in the intestine are crucial to maintaining intestinal homeostasis. DCs are constantly sampling antigens, such as food antigens, found in the intestine and present them to CD8+ T cells, which are generally known for killing infected cells. CD8+ T cells have been shown to be involved in pathogenesis of some cases of inflammatory bowel disease. Little is known about how these T cells respond towards the presentation of oral antigens derived from food. Therefore, it is important to understand the mechanisms involved in preventing the induction of harmful CD8+ T cell responses towards food derived antigens. For this purpose, CD8+ T cells with a transgenic T cell receptor that recognizes the food derived antigen ovalbumin were analyzed. It was demonstrated that ovalbumin given to mice orally induced CD8+ T cell activation and proliferation. Although the responding CD8+ T cells lacked cytolytic activity, they were induced to produce IFN-γ. This observed CD8+ T cell activation and proliferation is known to require cross presentation of oral antigens by DCs. Two cross presentation deficient mouse models, ΔYKb and CD74-/-, were used to assess mechanisms in the cross presentation pathway of oral antigens. The ΔYKb mice have a mutated endocytosis motif on MHC-I required for cross presentation and CD74-/- mice lack the CD74 chaperone protein. A low dose of oral ovalbumin resulted in reduced antigen specific CD8+ T cell proliferation and activation in the ΔYKb mouse strain only. Given the importance of the endocytosis motif for efficient cross presentation of oral antigens, human MHC-I alleles were examined for possible polymorphism in this motif. All analyzed human MHC-I alleles were conserved in the endocytosis motif. Collectively, this thesis demonstrates that cross presentation is an important pathway for antigen presentation and required for induction of CD8+ T cells activation and proliferation towards oral antigens. This mechanism might contribute to intestinal homeostasis and oral tolerance. iii Preface All experiments, with the exception of those described below, were performed by Ana Luisa Chávez Steenbock. I conceived of and designed all of the experiments and analyzed the data, wrote and edited the thesis. The data shown in Figure 4.5B was generated with the help of Dr. Gence Basha and Dr. Kyla Omilusik, which has been published and reproduced in Chapter 4 with permissions. Basha G, Omilusik K, Chavez-Steenbock A, Reinicke AT, Lack N, Choi KB, Jefferies WA. (2012) “A CD74-dependent MHC class I endolysosomal cross- presentation pathway.” Nature Immunology. 13(3): 237-245. All studies were performed following the guidelines set by both the University of British Columbia’s Animal Care Committee and the Canadian Council on Animal Care. The animal care breeding protocols are: A07-0373, and A09-0824. The animal care protocols for these studies are: A07-0270, A04-0267, A06-0346, and A05-1109. iv Table of Contents Abstract  .......................................................................................................................................  ii   Preface  .......................................................................................................................................  iii   Table  of  Contents  ....................................................................................................................  iv   List  of  Tables  ............................................................................................................................  vi   List  of  Figures  ..........................................................................................................................  vii   List  of  Symbols  and  Abbreviations  ................................................................................  viii   Acknowledgements  .................................................................................................................  x   Chapter  1.   General  Introduction  .....................................................................................  1   1.1   Dendritic  cells  and  T  cells  in  the  intestinal  associated  immune  system  .............  1  1.1.1   The  intestinal  immune  system  ...................................................................................................  1  1.1.2   T  cell  subsets  in  the  intestine  ......................................................................................................  6  1.1.3   Dendritic  cells  in  intestinal  homeostasis  ............................................................................  10  1.1.3.1   Dendritic  cell  antigen  presentation  ...............................................................................................  12  1.1.3.1.1   The  major  histocompatibility  complex  ...............................................................................  13  1.1.3.1.2   MHC-­‐I  cross  presentation  .........................................................................................................  16  1.1.3.1.3   The  MHC-­‐I  cytoplasmic  endocytosis  motif  in  cross  presentation  ..........................  19  1.1.3.1.4   CD74  (invariant  chain)  in  cross  presentation  .................................................................  20  1.1.3.2   MHC-­‐I  antigen  presentation  for  induction  of  tolerance  ........................................................  23  1.1.3.2.1   Role  of  cross  presentation  for  induction  of  tolerance  ..................................................  23  1.1.3.2.2   Antigen  presentation  by  dendritic  cells  for  oral  tolerance  ........................................  24   1.2   Specific  aims  ..........................................................................................................................  26  1.2.1   CD8+  T  cell  responses  towards  oral  antigens  ....................................................................  26  1.2.2   Cross  presentation  of  oral  antigens  for  CD8+  T  cell  activation  ..................................  26   Chapter  2.   Materials  and  Methods  ...............................................................................  27   2.1   Mice  ..........................................................................................................................................  27   2.2   Genotyping  by  PCR  ..............................................................................................................  28   2.3   Antibodies  and  flow  cytometry  .......................................................................................  28   2.4   Phenotyping  by  flow  cytometry  ......................................................................................  29   2.5   Lymphocyte  isolation  from  spleen,  lymph  nodes  and  thymus  .............................  29   2.6   Adoptive  T  cell  transfer  of  CFSE  labeled  cells  ............................................................  29   2.7   Induction  of  oral  priming  .................................................................................................  30   2.8   CFSE  proliferation  assay  and  tetramer  staining  .......................................................  30   2.9   Intracellular  staining  for  Foxp3+  ....................................................................................  30   2.10   Intracellular  IFN-­‐γ  staining  ...........................................................................................  31   2.11   Generation  of  bone  marrow  derived  dendritic  cells  ............................................  31   2.12   Transfection  of  bone  marrow  derived  dendritic  cells  .........................................  32   2.13   Cross  presentation  assay  ................................................................................................  32   2.14   Statistical  analysis  ............................................................................................................  32   Chapter  3.   CD8+  T  Cell  Responses  Towards  Oral  Antigens  ..................................  34   3.1   Rationale  ................................................................................................................................  34   3.2   Results  .....................................................................................................................................  35  3.2.1   CD4  Foxp3+  T  cells  are  generated  in  response  to  orally  administered  OVA.  ......  35   v 3.2.2   Orally  administered  OVA  does  not  induce  Foxp3  expression  in  OT-­‐I  cells.  ........  38  3.2.3   OT-­‐I  cells  express  activation  markers  after  orally  administered  OVA.  ..................  40  3.2.4   OT-­‐I  cells  are  activated  to  secrete  IFN-­‐γ  in  response  to  oral  OVA.  ..........................  44  3.2.5   OT-­‐I  cells  do  not  show  increased  cytolytic  function  in  response  to  oral  OVA.  ...  46  3.2.6   Adoptively  transferred  OT-­‐I  cells  display  activation  markers  in  response  to  oral  OVA   …………………………………………………………………………………………………………………..48  3.2.7   Oral  administration  of  OVA  protein  induces  proliferation  of  OT-­‐I  cells.  ..............  51  3.2.8   OT-­‐I  cells  in  C57Bl/6  mice  are  not  induced  to  express  Foxp3  in  response  to  oral  OVA   …………………………………………………………………………………………………………………..55   3.3   Discussion  ..............................................................................................................................  57  3.3.1   CD8+  T  cell  responses  towards  oral  antigens  ....................................................................  57   Chapter  4.   CD8+  T  Cell  Responses  Towards  Oral  Antigens  in  Cross   Presentation  Deficient  Mouse  Models  ...........................................................................  60   4.1   Rationale  ................................................................................................................................  60   4.2   Results  .....................................................................................................................................  61  4.2.1   ΔYKb  breeding  strategy  ...............................................................................................................  61  4.2.2   Genotyping  and  phenotyping  of  the  Kb  Tg+  mice  .............................................................  65  4.2.3   Phenotype  of  CD74-­‐/-­‐  mice  ........................................................................................................  69  4.2.4   Requirement  of  CD74  in  bone  marrow  derived  dendritic  cells  for  cross  presentation  ....................................................................................................................................................  71  4.2.5   The  role  of  cross  presentation  in  oral  tolerance.  ............................................................  73  4.2.6   Cross  presentation  motifs  found  in  cytoplasmic  domains  of  MHC-­‐I  molecules  are  conserved  in  human  MHC-­‐I  alleles  ................................................................................................  77   4.3   Discussion  ..............................................................................................................................  84  4.3.1   Endocytosis  motifs  on  MHC-­‐I  cytoplasmic  tails  of  HLA  alleles  .................................  84   Chapter  5.   Conclusions  and  Future  Directions  ........................................................  89   5.1   General  conclusions  ...........................................................................................................  89  5.1.1   CD8+  T  cell  responses  towards  oral  antigen  ......................................................................  89  5.1.2   Cross  presentation  ........................................................................................................................  90   5.2   Future  directions  .................................................................................................................  91  5.2.1   CD8+  T  cell  responses  towards  oral  antigen  ......................................................................  91  5.2.2   Cross  presentation  ........................................................................................................................  92   References  ..............................................................................................................................  95   vi List of Tables Table 1.1 DC subsets in PPs and MLNs. .......................................................................... 11   Table 1.2. Trafficking motifs on cytoplasmic tails of MHC-I molecules. ........................ 19   vii List of Figures Figure 1.1 Anatomy of the intestinal immune system. ....................................................... 3   Figure 1.2 Antigen uptake in the intestine for presentation to T cells. ............................... 5   Figure 1.3. T cell responses in the intestinal immune system ............................................ 9   Figure 1.4. MHC-I and MHC-II antigen presentation. ..................................................... 15   Figure 1.5. Pathways of cross presentation. ...................................................................... 17   Figure 1.6. Model of MHC-I trafficking for cross presentation in DCs. .......................... 22   Figure 3.1 OT-II mice have a higher frequency of Foxp3+ CD4+ T cells after oral administration of OVA protein. ................................................................................ 37   Figure 3.2 The Foxp3+ CD8+ T cell population in OT-I mice does not increase after oral administration of OVA protein. ................................................................................ 39   Figure 3.3 Activation markers are increased on surface of OT-I cells following oral administration of OVA. ............................................................................................ 43   Figure 3.4 CD8+ T cells secrete IFN-γ in response to oral administration of OVA protein. ...................................................................................................................... 45   Figure 3.5 No increased CTL responses of OT-I cells after oral OVA. ........................... 47   Figure 3.6 Oral administration of OVA induces expression of activation markers in OT-I T cells. ....................................................................................................................... 50   Figure 3.7 Oral administration of OVA induces proliferation of OT-I cells. ................... 52   Figure 3.8 IFN-γ expression in OT-I cells transferred into C57Bl/6 mice is not significantly different after oral administration of OVA. ......................................... 54   Figure 3.9 OT-I T cells do not express Foxp3 in C57Bl/6 mice before or after oral administration of OVA protein. ................................................................................ 56   Figure 4.1. ΔYKb and WTKb breeding strategy schematic ............................................. 64   Figure 4.2. MHC-I genes for genotyping by PCR ............................................................ 66   Figure 4.3. KbTg+ mice express H-2Kb and lack H-2Kk and Dk surface expression. ....... 68   Figure 4.4. Skewed CD8+ T to CD4+ T cell ratios in CD74-/- mice. ................................. 70   Figure 4.5. CD74 is required for efficient DC cross presentation. ................................... 72   Figure 4.6. ΔYKb mice have impaired CD8+ T cell proliferation and activation after oral administration of OVA protein. ................................................................................ 75   Figure 4.7. CD74-/- mice exhibit a trend towards reduced proliferation and activation of OT-I cells after oral administration of OVA protein. ............................................... 76   Figure 4.8. Multiple sequence alignment of transmembrane and cytoplasmic domain of HLA molecules ......................................................................................................... 79   Figure 4.9. Polymorphism in transmembrane and cytoplasmic domains of HLA-A, -B and –C. ............................................................................................................................. 83   viii List of Symbols and Abbreviations APCs Antigen  presenting  cells BmDCs   Bone  marrow  derived  dendritic  cells  CCR7 C-­‐C  chemokine  receptor  type  7 CLIP MHC  Class  II-­‐associated  invariant  chain  peptide CTL Cytotoxic  T  cells DCs Dendritic  cells DNA Deoxyribonucleic  acid EAE   Experimental autoimmune encephalomyelitis  ELC Endolysosomal  compartment ER Endoplasmatic  Reticulum ERAD   Endoplasmic-­‐reticulum-­‐associated  protein  degradation  Foxp3 Forkhead  box  P3 GALT Gut  associated  lymphoid  tissue IBD Inflammatory  Bowel  Disease IDO Indoleamine  2,3-­‐dioxygenase IEC Intestinal  epithelial  cells IEL Intraepithelial  lymphocyte IFN-­‐γ Interferon-­‐gamma ILF Isolated  lymphoid  follicles IPEX Immune  dysregulation,  polyendocrinopathy,  enteropathy,  and  X-­‐Linked  syndrome Kb  Tg Knocked  in  transgenic  H-­‐2Kb  molecule LAP Latency-­‐associated  peptide Listeria-­‐  OVA Recombinant  Listeria  strain    that  expresses  OVA LP Lamina  propria MBP Myelin  basic  protein MHC Major  Histocompatibility  Complex MLN Mesenteric  lymph  nodes NK  cell Natural  killer  cell OVA Ovalbumin PAMPs Pathogen  Associated  molecular  patterns PCR Polymerase  chain  reaction PD-­‐1 Programmed  death  1  receptor PDL Programed  death  cell  ligand PE Phycoerythrin PP Peyer’s  Patches PRRs Pattern  Recognition  Receptors Rac Ras-­‐related  C3  botulinum  toxin  substrate  1 RIP Rat  insulin  promoter RIP-­‐mOVA Transgenic  mouse  which  expresses  membrane  bound  OVA  (mOVA)  under  the  rat  insulin  promotor  (RIP) SD Standard  deviation ix Sec Second TCR T  cell  receptor Teff Effector  T  cells Treg Regulatory  T  cell TSA Tissue  specific  antigens TSLP Thymic  stromal  lymphopoietin WTKb Wild  type  MHC-­‐I  allele  H-­‐2Kb β2m Beta-­‐2-­‐microglobulin ΔYKb H-­‐2Kb  containing  a  single  substitution  of  phenylalanine  to  a  tyrosine x Acknowledgements First of all I want to thank all my family and friends for always giving me unconditional love and support. In particular I want to thank my mom and step dad, Lilian Steenbock and Emilio Zorrilla as well as my attentive and loving siblings, Jorge Chavez Steenbock, Natalia, Chavez Steenbock, Paola Zorrilla Steenbock and Isabel Zorrilla Steenbock. I appreciate my mom and my stepdad who encouraged my curiosity and love for Biology. They both came to Vancouver to drop me off when I started my Bachelor in 2005 at UBC and have never stopped supporting me regardless of the distance. It is because of them I have had the opportunity to pursue and follow my interests and continue this journey of never-ending learning. I want to thank Alejandra Lopez for being a great friend, confidante and companion since the beginnings of my new adventure at UBC. Over the years we have endured long hours of studying, all-nighters, frustration and tears as well as laughs, excitement and fun times. Thank you Ale for always hearing me out and providing me with a shoulder to rely on. I want to thank Lisa Murphy for making the beginnings of graduate school less scary and more fun. It was really hard and sad when you left but I am glad you made the choice that was best for you. I want to thank Maria Acevedo and Ashley Sanders for being supportive friends, being so much fun to be around with, also going through graduate school and not minding to talk about science even during social events. Both of you are great and Maria has been a really brave ski partner who I have had really good times with in Whistler. I want to thank all the members of the Biomedical Research Centre (BRC), including Ben Paylor, Sherie Duncan, Melanie Olson, Frann Antignano and Les Rollins who in different ways provided me with a fun and stimulating environment to learn and work at. I want to thank Taka Murakami for performing all the necessary genotyping. I want to thank the former and current members of the Jefferies lab including Dr. Cheryl Pfeifer, KB Choi, Carola Dreier and Laura Ho. In particular I want to thank Dr. Kyla Omilusik for having been and still being an amazing mentor. I am indebted to you for your scientific training, guidance and patience. I esteem you as a scientist and as a person. I am lucky to have had xi the opportunity to learn from you. I want to thank Dr. Kaan Biron for being a good friend, a fun labmate and showing me all the ins and outs of the BRC. I also want to thank you for your feedback and help in the process of writing and completing this thesis. Your advice and care have been tremendously important Cxliii [LU]. I want to thank Dr. Greg Lizée for his previous work on the ΔYKb mouse model and sparking my curiosity and interest in dendritic cells and cross presentation. I want to thank Ray Gopaul and Julian Kaye for their assistance in animal breeding and care at animal unit in South Campus and at the Animal Research Unit (ARU), respectively. I want to thank the ubcFLOW cytometry facility, especially Andy Johnson for his training and help on the flow cytometer. I want to thank my Committee Members Dr. Marc Horwitz and Dr. Thomas Grigliatti for their support and advice. Last but not least I want to acknowledge my supervisor Dr. Wilf Jefferies who has given me the opportunity to join his laboratory. Since the beginning he has challenged me intellectually, helped me grow as a scientist and driven me to accomplish this thesis. I am thankful to having had the opportunity to do research in his lab. The experience of being in graduate school and part of the Jefferies Lab has had a huge influence and contribution in making me become the person I am today. 1 Chapter 1. General Introduction 1.1 Dendritic cells and T cells in the intestinal associated immune system The intestinal mucosal surface encounters more antigens than any other part of the body [1]. Most human pathogens enter the body through mucosal surfaces [1], such as the intestine, so the immune system must be on guard to prevent pathogen invasion. On the other hand, the intestine is in constant contact with commensal bacteria and harmless food antigens. Mechanisms to maintain tolerance are therefore necessary to avoid unwanted immune responses that may lead to inflammatory bowel diseases (IBD) like Crohn’s disease or ulcerative colitis. The ultimate goal of the intestinal immune system is to protect against infection while avoiding destructive responses thereby maintaining intestinal homeostasis. Discriminating between harmful and harmless antigens is a big challenge. 1.1.1 The intestinal immune system The intestine is in contact with many environmental antigens and is also home to a vast number of commensal organisms [2]. Different components of the intestinal immune system (reviewed in [2, 3]) prevent the invasion of pathogens that gain access to the intestinal lumen. The first barrier of protection is provided by intestinal epithelial cells (IECs), which are connected by tight junctions forming an impermeable barrier. The IECs are made up of cells of a common lineage that include enterocytes, enteroendocrine, and paneth that differentiate from epithelial stem cells [4]. Enterocytes import nutrients and produce antimicrobial proteins like defensins [2]. Paneth and enteroendocrine cells secrete enteric hormones and high amounts of antimicrobial peptides such as proteolytic enzymes [2]. IECs express pattern recognition receptors (PRRs) and upon their stimulation produce pro-inflammatory cytokines [2]. Additional specialized IECs cells called goblet cells secrete glycoproteins that assemble into a thick mucus layer [2]. Beneath the IECs is the Lamina propria (LP), which is made up of stromal cells (myofibroblasts) and an extensive network of innate and adaptive immune cells [4]. Plasma cells located here produce IgA antibodies that are transcytosed across the 2 epithelial cell layer and secreted into the luminal space of the gut. Together with the mucus layer and antimicrobial peptides, IgA antibodies make a relatively impermeable sheet on the luminal surface of the epithelium that minimizes bacterial-epithelial cell contact, and therefore prevents pathogen invasion (Figure 1.1) [2, 4]. However, when pathogens manage to breach the first barrier of protection, the immune cell network (including macrophages, dendritic cells, T cells and plasma cells) located in the LP and organized lymphoid tissues along the intestinal tract is waiting to respond. 3 Figure 1.1 Anatomy of the intestinal immune system. A single layer of intestinal epithelial cells (IECs) provides a physical barrier that separates the commensal bacteria and food antigens in the intestinal lumen from the underlying lamina propria. Goblet cells produce mucus that together with antimicrobial peptides and IgA antibodies form a protective layer on the apical surface. Epithelial stem cells proliferate and give rise to the goblet cells and enterocytes, which absorb nutrients (small intestine) and water (colon). Progenitor IECs also differentiate into enteroendocrine cells, which secrete enteric hormones, and Paneth cells at the base of the small intestinal crypts. Beneath the IECs, the lamina propria is made up of stromal cells (myofibroblasts), B cells (especially IgA-producing plasma cells), T cells, macrophages and dendritic cells. Certain subsets of T cells (IEL, intraepithelial lymphocyte) and dendritic cells localize between the IECs. The regions of specialized epithelium in the small intestine associated with lymphoid tissue are termed follicle-associated epithelium. Microfold (M) cells, located in the epithelium associated to the Peyer’s patches, transport antigens across the epithelial layer. Adapted by permission from Macmillan Publishers Ltd: [Nature Reviews Immunology] ([4]), copyright (2010). 4 The intestinal immune system is comprised of effector sites and organized lymphoid tissue. The effector sites consist of lymphocytes scattered throughout the intestinal epithelium and LP. The organized lymphoid tissue is made up of Peyer’s Patches (PPs) and isolated lymphoid follicles (ILFs) [3]. PPs are only found along the small intestine while the isolated lymphoid follicles (ILFs) are distributed along the whole intestinal tract. Both sites are important for the induction of lymphocyte responses towards the antigens found in the intestine [1]. Located in the epithelium associated with PPs, microfold (M) cells sample and transport antigens from the intestinal lumen across the epithelial layer [4]. On the other side, dendritic cells (DCs) and macrophages can process and present the transported antigens [4]. Moreover, goblet cells in the small intestine also deliver low molecular weight soluble antigens from the intestinal lumen to underlying LP DCs [5]. DCs can also sample antigens directly from the intestinal lumen by inserting their dendritic processes between the epithelial cells layer [6]. Antigen-loaded DCs can then either present antigens directly to T cells along the gut lymphoid tissue or migrate in a CCR7-dependent manner to the intestine-draining mesenteric lymph nodes (MLNs) (Figure 1.2) [1, 7]. DCs, sampling intraluminal intestinal antigens from both food and bacteria, interact with naïve T cells in PPs, ILFs and MLNs to induce their activation and differentiation into specific T cell subsets [8]. The various T cell subsets found in the intestinal mucosa respond accordingly providing strong immune responses that eliminate infection to avoid systemic spread to the rest of the body [9]. 5 Figure 1.2 Antigen uptake in the intestine for presentation to T cells. The Peyer’s patches (PPs) and mesenteric lymph nodes (MLNs) are organized lymphoid tissue and the effector sites are scattered throughout the lamina propria and epithelium of the mucosa The subepithelial dome (SED) is the region of PPs rich in dendritic cells (DCs) and the thymus-dependent area (TDA) is the region where T cells reside. PPs and and villus lamina propria are drained by afferent lymphatics that go to the MLNs. (a) Antigen in the intestinal lumen enters through the microfold (M) cells into the PPs and is transferred to DCs. (b) DC can present these antigens directly to T cells in the PPs. (c) Alternatively, antigen-loaded DCs from the PPs can gain access to draining lymph and transport the antigen to the MLNs. (d) At the MLN, DCs can present the antigens to naïve T-cells (e) The antigen-responsive T cells express the chemokine receptor, CCR9, leave the MLN in the efferent lymph and exit into the mucosa through vessels in the lamina propria or disseminate throughout the peripheral immune system. Adapted by permission from Macmillan Publishers Ltd: [Nature Reviews Immunology] ([1]), copyright (2003). 6 1.1.2 T cell subsets in the intestine T cells develop in the thymus and are important cells of the immune system. T cells bear T cell antigenic receptors, also called T-cell receptors (TCR) that recognize antigens bound to major histocompatibility complexes (MHCs) [10]. The TCR can consist of either an α and β chain (TCRαβ) or a γ and δ chain (TCRγδ) [11]. Signaling through the TCR requires co-receptors CD4 or CD8 [10]. T cells expressing co-receptor CD4 differentiate into effector helper T cells, which in turn activate other lymphocytes [10]. Co-receptor CD8 consists of an α and β chain (CD8αβ) or two α chains (CD8αα) [10]. T cells expressing co-receptor CD8 commonly differentiate into armed effector T cells with cytolytic function (CTL cells) to kill infected cells [12]. T cell subsets in the intestine (reviewed in [8, 9]) are found in the epithelium, in aggregations of lymphoid follicles along the intestine, such as ILF and PPs, LP and in the MLN. The various subsets of T cells in the intestinal immune system play an important role in controlling infection, and at the same time in protecting the integrity of the epithelial cell barrier and thus the host. The T cells found in the epithelium located between IECs are called intraepithelial lymphocytes (IELs) [9]. These cells are categorized into two groups, unconventional and conventional. Unconventional IEL cells are comprised of TCRγδ+ and TCRαβ+ cell subsets expressing co-receptor CD8αα and not typical TCR co-receptors CD4 and CD8αβ [9]. Unconventional IEL subsets are thought to contribute to intestinal homeostasis by regulating the turnover of IECs and secreting hormones for epithelial repair [9]. They prevent microbial invasion by eliminating infected IEC through the direct recognition of self-antigen and stress induced receptors. The conventional IEL population is comprised of TCRαβ+ cell subsets expressing co-receptor CD8αβ (CD8+ T) or CD4 (CD4+ T) [9]. Conventional IEL T cells are believed to be primed in the periphery or MLN then migrate to the mucosal site to remain as memory cells rapidly responding upon antigen re-exposure [9]. The conventional IELs are therefore ready to recognize pathogen specific antigens while unconventional IEL are involved in repair and reducing damage. The T lymphocytes found in the LP consist mainly of TCRαβ+ CD4+ (CD4+ T) T cells and to a lesser extent TCRαβ+ CD8αβ+ (CD8+ T) cells. The LP also contains low 7 numbers of unconventional lymphocyte subsets such as invariant Natural Killer T cell (iNK) and mucosal associated invariant T cells, which interact with non-classical MHC [9]. Upon activation, CD4+ T cells can be induced to differentiate into Th1, Th2, Th17 and regulatory T cell subsets (Figure 1.5). The Th1 subset drives cell mediated immune responses towards intracellular pathogens. This is accomplished by the secretion of cytokines, like IFN-γ and TNF-α, which enhance macrophage activation and production of IgG [8]. The Th2 subset is involved in clearing parasitic infections, like helminthes. Cytokines secreted by the Th2 subset lead to the activation of mast cells, eosinophils and the production of IgE [8]. The Th17 subset provides defense against fungi and bacteria through the secretion of pro-inflammatory cytokines such as IL-17 and IL-22 [8]. Uncontrolled production of cytokines by T helper effector cells of the Th1, Th2 and Th17 subsets in the intestine have been shown to cause murine inflammatory bowel disease [8, 13]. Therefore, cytokine production of T helper effector cells has to be tightly regulated to avoid unwanted inflammation and intestinal damage. The CD4+ regulatory T (Treg) cells help to maintain intestinal homeostasis by suppressing cytokine production and proliferation of effector cells (reviewed in [3, 8, 13]). CD4+ Tregs are mainly identified by the expression of the transcription factor forkhead box P3 (Foxp3) and play a key role in regulating and controlling immune responses, especially in the intestine. This is highlighted by clinical studies where mutations in the Foxp3 gene have been shown to cause immune dysregulation, polyendocrinopathy, enteropathy, and X-linked (IPEX) syndrome [14]. Patients of IPEX suffer from multi-organ immune disorders including intestinal inflammation and damage [3]. The suppressive function of CD4+ Tregs in the intestine is also apparent in a T cell transfer model of chronic colitis used to study the regulation and the induction of intestinal inflammation [3, 15]. In this model, transfer of effector T cells into lymphopenic mice, such as Rag-deficient mice, induces intestinal damage and chronic colitis whereas transfer of Foxp3+ CD4+ Tregs suppresses intestinal inflammation and damage. This is presumably due to the suppressive function of Tregs which includes production of regulatory cytokines such as IL-10 and TGF-β [3]. The regulatory function of CD4+ Foxp3+ cells is crucial in suppressing unwanted inflammation and maintaining intestinal homeostasis [3, 8, 13]. 8 Most of the mouse models used to study intestinal inflammation have investigated the role of CD4+ T cells. However, little is known of the lesser abundant CD8+ T cells’ role in intestinal homeostasis. CD8+ T cells can be divided into two categories, armed effector T cells and regulatory cells. The armed effector function in the intestine can induce intestinal inflammation mediated by the secretion of IFN-γ and TNF-α, which results in the loss of epithelial barrier integrity [16, 17]. Effector CD8+ T cells are thought to be suppressed by antigen specific CD4+ Tregs [18]. However, the detailed role that CD8+ T cells play during intestinal homeostasis has not been widely studied and requires our attention. The intestinal CD8+ Treg cell population (reviewed in [19, 20]) is much smaller than CD4+ Treg population. Nevertheless, the CD8+ Treg cells that have been identified display similar suppressive functions as their CD4+ Treg counterparts [21]. However, contrary to CD4+ Tregs that are widely identified by expression of Foxp3, no specific markers have been established for CD8+ Tregs, making them hard to characterize. In the context of the intestine, a cell population with suppressive function, identified by the surface markers CD8+CD28-, was shown to prevent the development of colitis in a adoptive T transfer model of colitis [22]. Furthermore, a CD8+ Treg population expressing Foxp3+ was identified in a transgenic mouse strain, which contains hemagglutinin (HA) specific CD8+ T cells and enterocytes endogenously expressing HA [17, 21]. The authors suggested that gut specific antigen presentation induced CD8+ Foxp3+ Tregs in vivo. In turn, CD8+ Foxp3+ Tregs maintained intestinal homeostasis by down-modulating effector functions of T cells. CD8+ Foxp3+ cells have also been isolated from human patients with increasing numbers in patients with colorectal tumors [23]. It was suggested that in these cases increased CD8+ Foxp3+ cells in humans might contribute to tumor immune escape in the intestinal mucosa. Altogether, the various T cell subsets found in the intestine are important in maintaining a fine balance between effector and regulatory responses to provide protection while keeping intestinal integrity and function. 9 Figure 1.3. T cell responses in the intestinal immune system The T cells located between IECs are the intraepithelial lymphocytes (IELs) and are comprised of unconventional TCRγδ+CD8αα+, TCRαβ+CD8αα+ and conventional TCRαβ+CD8αβ+ and TCRαβ+CD4+ cells. DCs sample intestinal antigen found in the intestinal lumen and upon exposure to certain signals, they migrate to specialized immune regions such as MLNs, where they may interact with naïve TCRαβ+CD4+ (CD4+) or TCRαβ+CD8αβ+ (CD8+) T cell. The nature of the presented antigen and the cytokine milieu affects the balance of specific effector and regulatory CD4+ and CD8+ T cell subsets. Th1, T helper 1; Th2, T helper 2; Th17, T helper 17; Treg, regulatory T cells; CTL, cytotoxic T cells. Adapted from International Journal of Medical Microbiology, Volume 300, Issue 1, Astrid M. Westendorf, Diana Fleissner, Wiebke Hansen, Jan Buer, T cells, dendritic cells and epithelial cells in intestinal homeostasis, 11-18, Copyright (2010), with permission from Elsevier. 10 1.1.3 Dendritic cells in intestinal homeostasis The biggest challenge of intestinal immunity is inducing tolerance towards harmless antigens derived from food and commensal bacteria while maintaining protection against dangerous pathogens. Loss of this balance would initiate unwanted intestinal inflammation and lead to the development of IBDs [24]. DCs found in the intestine have a crucial role in maintaining this intestinal homeostasis (reviewed in [25, 26]). DCs are a heterogeneous population of professional antigen presenting cells (APCs) of haematopoietic origin that are specialized for capturing, processing and presenting antigens to T cells [27]. Through direct interaction, DCs stimulate T cells to differentiate into T cell subsets with specific functions. The DC subtype and activation state are important dictators of the fate of the T cell differentiation. Therefore, DCs can control induction of either immunity or tolerance and unresponsiveness to self and harmless foreign antigens [28]. DCs are found in all lymphoid structures of the intestine (PPs and ILFs), the MLN, and are scattered throughout the subepithelial LP. In the intestinal immune system, DCs can be classified depending on expression of certain surface markers (Table 1.1). The conventional DC subsets in the intestinal immune system are CD11chiCD11b+CD8α-, CD11chiCD11b-CD8α+ and CD11chiCD11b-CD8α-, all of which have distinct histological distributions and functional responses. Furthermore, DCs that originate in the intestine LP have been characterized by surface expression of integrin CD103 (αE integrin) (reviewed in [26]). CD103+ DCs constantly sample luminal antigens before migrating in a CCR7-dependent manner to the MLN where they interact with naive lymphocytes. CD103+ DCs are known to have a tolerogenic profile and play a crucial role in tolerance to commensal bacteria and food antigens [26]. 11 indicated that most of these are not genuine DCs, as defined by the ability to prime naı̈ve T cells after migrating to the draining lymph nodes. The only cells that can do this in the resting mucosa express CD103 and not the fractalk- ine receptor CX3C chemokine (CX3CR)1 [15,19,22,23] (Box 1). These CD103+ DCs have many unique properties and are the focus of this review. A minor population of plasmacytoid DCs (pDCs) is also present in the LP. These C–C chemokine receptor (CCR)9+ cells have been sug- gested to play a role in driving the migration of DCs to the MLNs [24], but their role in antigen presentation and the regulation of mucosal immunity is unclear. CD103+ DCs in the LP CD103 (aE integrin), which binds the integrin b7 to form the aEb7 complex, was first detected as a marker of intrae- pithelial CD8+ T cells in the gut [25,26], but its function remains unclear. Its best-known ligand, E-cadherin is expressed by intestinal epithelial cells (IECs) and is sug- gested to function in maintaining CD103+ T cells and CD103+ DCs in the intestine [26–28]. However, to the best of our knowledge, this has yet to be proven. An additional, uncharacterised ligand for CD103 has also been identified on vascular endothelium in the intestine [29], but the nature of this interaction remains to be elucidated. CD103+ DCs make up 2–3% of total leukocytes in the small intestinal LP of normal mice, where they display a rapid turnover, migrate constitutively to the MLNs (800 000 DCs per day in rats [30]) and are replenished continually by blood-borne precursors [31]. Smaller num- bers of CD103– expressing DCs have also been identified in PPs, MLNs and lymph [32], as well as in non-intestinal tissues including the skin, lungs, spleen and peripheral lymph nodes [22,33,34]. Recent work has suggested that regardless of their localisation, all CD103+ DCs might share a common lineage. Like CD103– DCs, CD103+ DCs are derived from a CX3CR1+ c-kit+ bone marrow precursor in an Fms-like tyrosine kinase 3 (Flt3) ligand-dependent manner. Despite this, CD103+ and CD103– DCs are genet- ically distinct. CD103+ DCs express higher levels of CCR6, CCR7, Toll-like receptor (TLR)5 and TLR9, but lower amounts of other TLRs, co-stimulatory molecules and proinflammatory mediators than CD103– DCs express [35]. It is not clear how non-intestinal CD103+ DCs relate to those in the gut. Non intestinal CD103+ DCs appear to be related to the CD8a+ lineage of conventional DCs, which requires the transcription factors inhibitor of DNA-binding 2 (Id2), interferon regulatory factor 8 (Irf8) and basic leucine zipper transcription factor ATF-like 3 (Batf3) for their development, and are particularly effective at cross- presenting exogenous antigen to CD8+ T cells [36]. Al- though CD103+ LP DCs can also cross-present exogenous antigens to T cells [36] and express high levels of CCR7 [37], the expression and function of TLR by mucosal CD103+ DCs remains an unresolved issue. Although mi- grating CD103+ DCs in rat intestinal lymph appear to express all TLRs except TLR4 [38], early studies of TLR expression by LP DCs in mice used heterogeneous popula- tions of mononuclear cells, and are now difficult to inter- pret [39,40]. A further important difference between CD103+ DCs in the LP compared with other tissues is that the LP population is heterogeneous, and contains subsets of CD11b+ CD8a– and CD11b– CD8a+ DCs that are found among CD103– DCs elsewhere [41,42]. Of these, only the CD11b–CD8a+ subset of CD103+ LP DCs requires Id2, Irf8 and Batf3 for its development [43]. Thus, more work is needed to determine if and how intestinal and non-intesti- nal populations of CD103+ DCs are related. Functions of LP DCs in resting intestine CD103+ LP DCs have several unique features that distin- guish them from other DCs. The first documented of these is the ability to imprint the expression of the gut homing markers CCR9 and a4b7 on interacting naı̈ve T and B cells and to induce expression of FoxP3 by naı̈ve CD4+ T cells [22,44–46]. This occurs in the MLNs rather than in the mucosa itself, where naı̈ve T lymphocytes are rare [47,48]. Although CD103+ DCs appear to be the only cells that can present intestinal protein or bacterial antigens to T cells [15,49], it is unknown how they acquire antigen. They are probably not the cells that can extend processes through the epithelium into the lumen, as was originally thought, because this seems to be a property of mucosal CX3CR1+ macrophages [15,49]. These macrophages might subse- quently transfer antigen to CD103+ DCs in the LP. After acquiring antigen, CD103+ LPDCsmigrate to theMLNs in Table 1. DC subsets in PPs and MLNs. Phenotype Location Function Reference CD11b+ CD8a– PP Th2 polarising ability IgA class switching IL-10 production [41] CD11b– CD8a+ PP Th1 polarising ability IL-12p70 production [41] CD11b– CD8a– PP TH1 polarising ability IL-12p70 production [41] CD103+ MLN Treg polarising Gut-homing T cell imprinting [22,27,46] CD103– MLN Proinflammatory Th1/Th17 polarising ability [22,102] Box 1. Definition of intestinal DCs Based on work in non-intestinal lymphoid organs, murine DCs in the intestine were originally defined simply as CD11c+ class II MHC+ cells, and several functionally specialised subsets were described on this basis. Recently, it has become apparent that this is not sufficient to distinguish DCs from other myeloid cells such as macrophages, particularly in non-lymphoid tissues such as the gut [42,49]. There is an emerging consensus that many of the CD11c+ class II MHC+ cells in the intestinal mucosa do not fulfil the functional requirements of DCs, and two subsets of mononuclear phagocytes have now been defined in the murine gut on the basis of the expression of the mutually exclusive markers CD103 and the fractalkine receptor CX3CR1 [23]. Of these, only the CD103+ CX3CR1– subset can migrate from the LP to the MLNs and present locally administered antigen to naı̈ve CD4+ T cells [15,22]. The CD103+ subset is derived from the common DC precursor and its development depends on the DC- specific growth factor Flt3 ligand [49]. These CD11c+ class II MHC+ CD103+ cells therefore appear to be bona fide DCs. Conversely, the CD103– CX3CR1+ subset appears to be sessile in the mucosa and has little or no ability to prime naı̈ve T cells. Furthermore, these cells express the macrophage marker F4/80; their development is controlled by macrophage CSFs; and they are derived from Ly6Chi blood monocytes [49]. For these reasons, the latter cells are now considered to be macrophages [19]. Review Trends in Immunology September 2011, Vol. 32, No. 9 413 Table 1.1 DC subsets in PPs and MLNs. Adapted from Trends in Immunology, Volume 32, Issue 9, Charlotte L. Scott, Aude M. Aumeunier and Allan McI. Mowat, Intestinal CD103+ dendritic cells: master regulators of tolerance?, 412–419, Copyright (2011), with permission from Elsevier. The signals that DCs receive to induce tolerance versus an immune response are currently under study (r viewed in [25]). The anatomical location and local microenvironment of the DC during exposure to antigen is believed to be a key determinant of the type of immune response generated. Various f ct rs present in the intestine are important for DCs to promote tolerogenic responses. These include two important soluble factors, TGF-β and IL-10, that are produced by a variety of different cells types in the intestinal mucosa [25]. TGF-β or IL-10 knockout mice develop spontaneous intestinal inflammation illustrating the importance for these two cytokines in maintaining intestinal homeostasis [29, 30]. DCs ar also conditioned by IECs [31]. The expression of a broad range of cytokines, like thymic stromal lymphopoietin (TSLP), by IECs influences DCs to downregulate responses to b cterial stimulation and mount Th2 type responses [25, 32]. For instance, the IECs of some patients with Crohn’s disease showed impaired production of TSLP and thereby failed to condition DCs to become tolerogenic [31]. This presumably has fatal consequences in avoiding uncontrolled Th1 driven responses and would potentially be the cause for intestinal inflammation seen in these patients. Tolerogenic DCs express low l vels of MHC-II molecules, co-stimulatory molecules and certain TLRs [25]. They are therefore resumed to be less re ponsive to pathogen associated molecular patters (PAMPs) compared to DCs from the blood or spleen. Hart et al. showed that the DCs isolated from Crohn’s disease patients had higher 12 expression of TLR2 and TLR4 and produced higher levels of cytokine IL-12 [33]. It was speculated that the higher expression of TLRs in DCs resulted in poor DC conditioning. Therefore, these cells were more prone to drive Th1 type responses and promote inflammation. Taken together, conditioning of DCs in their local environment is essential to establishing tolerance and in fact has been argued by some to have a more important role than the subset of DC [25, 26] Intestinal DCs conditioned to be tolerogenic migrate to the PP or MLN to interact with B and T cells. As a result, IgA-secreting plasma cells and non-inflammatory Th2 type responses are induced. Intestinal DCs, mainly of the CD103+ subset, have also shown to promote tolerogenic responses by inducing CD4+ Foxp3+ Tregs from naïve CD4+ Foxp3- T cells. Induction of Tregs, which is dependent on TGF-β and retinoic acid (RA), is of critical importance to prevent naïve T cells from reacting to harmless food or commensal bacteria antigens [34-36]. The RA is derived from Vitamin A (retinol), which can be catabolized by IECs making it available for use by DCs. Furthermore, conditioned CD103+ DCs also express the appropriate enzymes to convert RA from retinol [26]. Conditioned DCs in the intestine are also induced to produce the enzyme indoleamine 2,3-dioxygenase (IDO) that is important in tryptophan metabolism. When expressed by DCs in the intestine, IDO depletes tryptophan from the microenvironment favoring Treg induction and inhibiting effector T cells generation [37]. Overall the microenvironment in the intestine plays a big role in conditioning DCs to promote tolerance and shape the nature of the immune response in the gut. [25]. 1.1.3.1 Dendritic cell antigen presentation DCs establish immune responses in the intestine by interacting with T cells and inducing their differentiation into the appropriate effector cell. This initial T cell differentiation is initiated by the interaction of MHCs on the surface of APCs with TCRs on the cell surface of T cells [38]. DCs are constantly processing antigens from exogenous and endogenous sources and loading antigenic peptides on MHCs for presentation to T cells [39]. The nature of the antigen as well as expression of co- stimulatory signals influence the type of immune response generated. 13 1.1.3.1.1 The major histocompatibility complex MHC molecules are vital in generating a proper immune response. These molecules are expressed on the cell surface as MHC-peptide complexes [39]. Peptides bind to the MHC peptide-binding cleft and are displayed on the cell surface for immune surveillance. The amino acids forming the peptide-binding cleft of the MHC molecule are highly polymorphic, meaning there is a high degree of genetic variability for the display of an array of different peptides on the cell surface [39]. Expressed by APCs like DCs, the MHC-peptide complex interacts with T cells to initiate an immune response [39]. There are two classes of MHC molecules: MHC Class I (MHC-I) and Class II (MHC- II). MHC Class I molecules consist of a heavy chain with three α-domains, a transmembrane domain and a cytoplasmic domain [40]. The heavy chain is non- covalently associated to β2-microglobulin (β2-m) [39]. MHC-II molecules are composed of an α and β chain linked non-covalently [39]. Each MHC-II chain contains both a transmembrane and cytoplasmic domain [41]. MHC-I and MHC-II are synthesized in the ER but have different pathways and mechanisms for being loaded with peptides and trafficked to the cell surface (reviewed in [39]). In the case for MHC-I (reviewed in [39]), the molecule functions as part of an immune surveillance mechanism to detect cells infected with intracellular pathogens. Endogenous proteins in the cytosol are degraded into small peptides by the proteasome and subsequently pumped into the ER by TAP where they bind to MHC-I molecules (Figure 1.4A). Proper folding and peptide loading of MHC-I molecules in the ER is assisted by a variety of chaperones including calnexin, tapasin and ERp57 [39]. The MHC-I molecules are retained in the ER until they have bound a peptide. Together with β2-microglobulin, the MHC-peptide complex is transported through the Golgi to the plasma membrane [39]. At the cell surface, the MHC-peptide complex can interact with the T cell receptor (TCR) of CD8+ T cells [39]. The MHC-TCR interaction, together with costimulation, leads to T cell differentiation, activation and proliferation. If the peptide displayed on MHC-I molecules is pathogen derived, activated CD8+ T cells can differentiate into CTLs to kill infected cells [42]. MHC-II molecules present antigen peptides from engulfed extracellular pathogens (reviewed in [43]). In the ER, MHC-II molecules bind with the CD74 (invariant chain) 14 chaperone protein (Figure 1.4B). The binding of CD74 prevents loading of MHC-II with endogenous peptides found in the ER. The cytoplasmic domain of CD74 contains two leucine based sorting motifs [43]. This aids in directing the CD74-MHC-II complex through the endomembrane system to the compartments containing the exogenous derived peptides [43]. Low pH levels and proteases, like cathepsin S, found in these endocytic compartments cleave and degrade CD74 [44]. However, a short CD74 fragment called CLIP is left bound to the MHC-II molecule [43]. Within the same compartment, the chaperone protein HLA-DM binds, stabilizes and aides MHC-II in the exchange of CLIP for exogenous antigenic peptides [43]. The MHC-II-peptide complex is then trafficked to the cell surface for interaction with the TCR of CD4+ T cells. Upon MHC-TCR interaction, together with co-stimulation, CD4+ T cells are activated and differentiate into effector helper T cells (Th1, Th2, Th17 and Treg subsets), which secrete cytokines to engage other lymphocytes to fight infection [8]. 15 Figure 1.4. MHC-I and MHC-II antigen presentation. (A) Classical MHC-I antigen presentation pathway. Endogenous antigen is loaded onto MHC I. The proteasome degrades proteins found in the cytosol and peptides are delivered to the endoplasmic reticulum (ER) through the TAP transporter. In the ER the processed antigenic peptides are loaded into the binding groove of nascent MHC-I molecules. MHC-I is trafficked to the cell surface for recognition by CD8+ T cells. (B) Classical MHC-II antigen presentation pathway. MHC-II αβ heterodimers are assembled in the ER where CD74 (not shown) associates with the peptide binding groove to prevent premature binding of peptides in the ER. CD74 chaperones MHC-II through the endocytic route to the compartments containing internalized antigens. Endosomal proteases degrade the exogenous antigen as well as CD74 and exogenous antigens are loaded on MHC-II. Upon loading, MHC-II traffics to the cell surface for recognition by CD4+ T cells. (C) MHC-I cross presentation pathway. MHC-I molecules gain access to peptides derived from exogenously derived antigens. Peptides are loaded onto MHC-I which then traffic to the cell surface for recognition by CD8+ T cells. Reprinted by permission from Macmillan Publishers Ltd: [Nature Reviews Immunology] ([45]), copyright (2007). 16 1.1.3.1.2 MHC-I cross presentation In the classical pathway, MHC-I presents endogenously derived antigens. Nevertheless, DCs can mount an immune response towards intracellular pathogens even if they are not themselves infected. Through the cross presentation pathway, an alternate MHC-I pathway, DCs can acquire, process and present antigenic peptides derived from extracellular sources (Figure 1.4C). This pathway is essential for CD8+ T cell-mediated responses against viruses, bacteria and tumours [42]. Three non-mutually exclusive models have been proposed to explain the cross presentation mechanism: the cytosolic model, the phago ER model and the vacuolar model (Figure 1.5) (reviewed in [46]). In the cytosolic model, two protein channels related to the ER-associated degradation (ERAD) pathway (Sec 61 and Derlin 1), have been proposed to transclocate endocytosed antigens from the phagosome into the cytosol [46]. Once in the cytosol, antigens can be processed by the classical MHC-I pathway. The cytosolic pathway would completely rely on the proteasome and TAP for the processing of exogenously derived peptides. The phago-ER model proposes that phagocytic and ER vesicles fuse to form a specialized compartment referred to as the “ergosome”. This vesicle would contain endocytosed antigens, MHC-I loading machinery and ERAD components like Sec 61 and therefore would act as a self-sufficient organelle for cross presentation [46]. Exogenous antigens translocated out of the ergosome by Sec 61 into the cytosol would be degraded by associated proteasome complexes and the peptides would subsequently be transported back into the ergosome by TAP. The exogenously-derived proteasomal peptide products would replace peptides that bound to MHC-I in the ER. This model remains controversial because the existence of the ER-phagosome fusion has been called into question [47]. 17 Figure 1.5. Pathways of cross presentation. (A) Cytosolic model: Exogenous protein (red) is internalized into the endolysosomal pathway through early endosomes (EE) to late endosome compartments (LE) and transported from this compartment to the cytosol. There it enters the classical MHC-I pathway along with endogenous antigen (green). The antigen is degraded by cytosolic proteasome complexe and peptides are transported by TAP into the ER. Loaded MHC I complexes are then transported from the ER through the trans-Golgi Network (TGN) and to the plasma membrane (PM) by the secretory pathway. (B) Vacuolar model: Endocytosed extracellular protein antigens are transported through EEs to LE compartments, where they are degraded into antigenic peptides by resident proteases. Cell surface MHC-I molecules gain access to LEs and peptide exchange takes place. MHC-I bound to exogenously derived peptide is transported to the cell surface for cross- presentation to T cells. (C) Phago-ER model: Phagocytosis involves fusion of the ER with the PM to form an ER–phagosome mix compartment which contains ER components, such as TAP, tapasin and Sec 61. Phagocytosed exogenous protein antigens are partially degraded then transported into the cytosol by the Sec61 transporter, where they are further degraded by the associated proteosome complexes and transported back into the same ER–phagosome by TAP. Exogenously derived peptides are then loaded onto MHC-I molecules derived from the ER or PM. If the phagosome is contiguous with the ER, the peptides that re-enter the phagosome-ER can also be loaded onto the newly synthesized MHC class I molecules. Adapted by permission from Macmillan Publishers Ltd: [Nature Immunology] ([48]), copyright (2012). 18 In the vacuolar pathway, exogenous proteins are internalized by phagocytosis or endocytosis by the cell [42, 46]. The resulting internalized proteins are then degraded by proteases, like cathepsin S, generating antigen peptides. The vacuolar model proposes MHC-I molecules are directed to these endocytic compartments and then loaded with the antigen peptides. The loaded MHC-I molecules are transported to the cell surface for antigen presentation. In this model, exogenous proteins are processed independently of the proteasome and TAP [42, 46]. It is still unclear how MHC-I molecules gain access to this compartment for loading of peptides of exogenous origin. However, studies from the Jefferies lab suggest two mechanisms. MHC-I molecules are directed from the cell surface to endocytic compartments by way of a tyrosine internalization motif found in the MHC-I cytoplasmic tail [40]. Alternatively, MHC-I are trafficked directly from the ER in a CD74 dependent manner to the endocytic compartment [49, 50]. These studies will be further discussed. 19 1.1.3.1.3 The MHC-I cytoplasmic endocytosis motif in cross presentation The cytoplasmic region of MHC-I, encoded by exons 5 to 8 [41], contains important trafficking motifs (Table 1.2) necessary for ER export, endocytosis from the cell surface, recycling and targeting for degradation (reviewed in [51]). The trafficking motifs may explain how MHC-I is targeted to cellular compartments in the context of the vacuolar model of cross presentation. MHC-I cytoplasmic motif Amino acids Shown for Refs Target lysines for ubiquitylation K HLA-A [52], [53] ER export motif RXR HLA-F [54] Tyrosine based endocytosis motif YXXØ HLA-B [55] H-2Kb (murine) [40] Dileucine endocytosis motif DXSLI HLA-C [56] MHC-I dimers linked by disulfide bonds between cytoplasmic residues C HLA-B [57] C-terminal valine for ER export V HLA-A [58] HLA-F [54] C-terminal alanine for ER export A HLA-B, HLA-C [59] Table 1.2. Trafficking motifs on cytoplasmic tails of MHC-I molecules. Cytoplasmic domains of MHC-I molecules contain important trafficking motifs. Ø indicates a bulky hydrophobic amino acid (Leu, Met, Ile, Phe or Val) and X is any amino acid. 20 MHC in humans are called Human Leukocyte Antigen (HLA) and there are six class I HLA genes called HLA-A, -B, -C, -E, -F and –G. Of these six genes HLA-A, -B, -E and –F contain a cytoplasmic tyrosine-based endocytosis motif, YXXØ. This motif has been extensively characterized in the context of the transferrin receptor, which undergoes clathrin-mediated endocytosis [60, 61]. In MHC-I, the YXXØ motif is conserved across species [62]. Substitution of the tyrosine for a phenylalanine in the YXXØ motif on murine (H-2Kb) or human (HLA-B27) MHC-I molecules abrogated MHC-I endocytosis and trafficking to endocytic compartments [40, 55, 62]. HLA-C lacks the YXXØ motif and relies on a dileucine motif (DSXLI) for endocytosis and translocation to the lysosomal compartments [56]. Both YXXØ and DXSLI motifs have been proposed to be important for cross presentation [40, 55, 56, 62]. The ΔYKbC3H mouse strain has been used to assess the importance of the YXXØ motif for cross presentation and the induction of immunity in vivo [40, 62, 63]. Made on a C3H background, the ΔYKbC3H mice express a mutant MHC-I molecule containing a single substitution of phenylalanine for tyrosine in the YXXØ motif. DCs in ΔYKbC3H mice mount inferior CD8+ T cell responses against viruses and intracellular bacteria [40, 62, 63]. Cross presentation efficiency by DCs was further studied using ovalbumin (OVA) as the model antigen. DCs derived from the mutant mice had fewer H- 2Kb/OVA257-264 complexes in endolysosomal compartments and at the cell surface [40, 62, 63]. Confocal microscopy on DCs revealed that wild-type H-2Kb molecules were endocytosed and trafficked to endolysosomal vesicles. However, ΔYKb molecules accumulated in the DC cell surface [40, 62]. These studies support the vacuolar model of cross presentation and illustrate the importance of the MHC-I YXXØ motif. 1.1.3.1.4 CD74 (invariant chain) in cross presentation The chaperone protein CD74 is primarily known for its role in MHC-II antigen presentation [43]. As mentioned earlier, CD74 directs MHC-II from the ER through the endomembrane system to compartments containing endocytosed antigens. Recent studies by the Jefferies Lab have suggested a role for CD74 in cross presentation [49, 50]. Studies using a CD74 deficient mouse strain (CD74-/-) demonstrated that DCs in this mouse have reduced MHC-I loading in endolysosomal compartments. Furthermore, there 21 was an impaired ability to present MHC-I-antigen complexes at the cell surface. As a result, the DCs from CD74-/- mice were unable to prime CD8+ T cells and generate efficient primary antiviral immune responses [49, 50]. The authors proposed a CD74- mediated cross presentation pathway. Studies supporting the CD74-mediated cross presentation pathway suggest CD74 binds to a fraction of MHC-I molecules in the ER [49, 50]. Signals in the cytoplasmic domain of CD74 direct trafficking of the CD74-MHC-I complex to endolysosomal compartments, similar to MHC-II. At this compartment, MHC-I dissociate from CD74 allowing for loading of exogenous derived antigenic peptides. The loaded MHC-I is then trafficked from the endocytic compartment to the cell surface (Figure 1.6). In support of this model, CD74 peptides corresponding to CLIP can be eluted from MHC-I molecules [64]. Altogether, the involvement of CD74 in cross presentation pathway provides support for the vacuolar model. The CD74 mediated cross presentation pathway also provides an additional mechanism by which MHC-I molecules are directed to intracellular compartments for loading of exogenously derived peptides. 22 Figure 1.6. Model of MHC-I trafficking for cross presentation in DCs. Schematic representation of trafficking routes for MHC-I molecules in DCs for acquisition of exogenously derived antigens. Exogenous protein antigens (red) are internalized into endocytic vesicles and transported into endolysosomal compartments (ELC) where they are degraded. MHC-I molecules can get access to ELC via two pathways: (a) MHC-I molecules in the cell surface are internalized by the cytoplasmic tyrosine based endocytosis motif (Y) and directed to ELC. Here endogenously derived peptides (green) are exchanged with high affinity exogenous derived peptides (red). MHC-I molecules are then directed back to the cell surface. (b) Alternatively a fraction of MHC-I molecules at the ER interact with CD74 chaperone proteins, which direct them through the endocytitic pathway to ELC. Here CD74 is degraded and exogenous derived proteins can be loaded into MHC-I molecules. Then MHC-I molecules are directed to the cell surface for interaction with TCRs of CD8+ T cell. Adapted by permission from Macmillan Publishers Ltd: [Nature Immunology] ([48]), copyright (2012). 23 1.1.3.2 MHC-I antigen presentation for induction of tolerance 1.1.3.2.1 Role of cross presentation for induction of tolerance Antigen processing and presentation by DCs is crucial for generating immune responses. Specifically, cross presentation of pathogen or tumour derived antigens are necessary for inducing differentiation of CD8+ T cells into CTLs [42]. However, DCs are also important mediators of tolerance. In circumstances where DCs induce tolerance they are referred to as tolerogenic DCs (reviewed in [28, 65]). DCs with a tolerogenic phenotype present antigen but lack co-stimulatory molecules (CD80 and CD86) and express high levels of inhibitory molecules, such as programmed death cell ligand (PDL) [3]. Tolerance towards antigens endogenously expressed by DCs depends on the expression of programmed death 1 (PD-1) and CTLA-4 receptors on CD8+ T cells [66]. Both receptors are known to be negative regulators on T cells function [67, 68]. Tolerogenic DCs can induce several T cell fates including: death, anergy, or differentiation of naïve T cells into regulatory T (Treg) cells rather than inducing differentiation of effector T cells (Teffs) [28, 65]. This prevents T cells with TCRs that recognize self-antigens or external harmless antigens to induce an unwanted immune response [65]. Cross presentation by DCs for induction of CD8+ T cell tolerance is called cross tolerance. The mechanisms and factors involved in cross tolerance still remain unclear. Nevertheless, this process is of particular importance to induce CD8+ tolerance towards self-antigens from apoptotic cells found in the periphery as well as harmless environmental antigens like food [69-72]. Three studies strongly suggest DCs play an important role in inducing cross tolerance to avoid generation of CD8+ T cells with effector function. Kurst et al. used the transgenic mouse model RIP-mOVA mice, which expresses membrane-bound OVA (mOVA) under the rat insulin promoter (RIP) [72]. Transfer of CD8+ T cells that recognize H-2 Kb/OVA(257–264) complexes (OT-I cells) into RIP-mOVA mice did not induce T cell mediated responses towards the pancreatic cells expressing the mOVA [72]. The authors suggested DCs cross presented the OVA derived from apoptotic pancreatic cells, rendering OT-I cells unresponsive and thereby inducing tolerance. Luckashenak et al. used a transgenic mouse strain which express a dominant-negative mutant of the 24 GTPase Rac1 under control of the DC-selective CD11c-promoter, referred to as Rac mice [69]. DCs from this strain, are unable to cross present apoptotic-derived antigens to initiate CD8+ T cell responses [69]. Rac mice were crossed to RIP-mOVA mice to generate a mouse strain that contained DCs deficient in cross presentation as well as pancreatic cells expressing mOVA. When OT-I cells were transferred into these mice the transferred OT-I cells mounted a cell mediated immune response against pancreatic cells [69]. This result verified that DCs cross present antigens derived from apoptotic cells to induce tolerance. Moreover, Rac mice accumulated self-reactive T cells that when transferred to lymphopenic mice (Rag1-/-) caused various symptoms of immune-mediated disease [69]. Schildknecth et al. examined cross tolerance with a more complex system [70]. This study used a mouse model that had inducible expression of CD8+ T cell epitopes of the lymphocytic choriomeningitis virus (LCMV) in the nervous system. Challenge with LCMV failed to induce neural inflammation and to activate specific CD8+ T cells that recognize the particular LCMV epitopes expressed by oligodendrocytes and Schwann cells. The generation of bone marrow chimeras confirmed that bone marrow derived cells mediated the tolerance. The authors suggested DCs cross present the LCMV epitopes expressed by oligodendrocytes and Schwann to induce cross tolerance. These three studies illustrate the importance of cross presentation for the induction of CD8+ tolerance. 1.1.3.2.2 Antigen presentation by dendritic cells for oral tolerance It has been observed that antigens acquired through the oral route suppress humoral and cell mediated immune response following immunization with the same antigen. This hyporesponsiveness is a phenomenon known as “oral tolerance” (reviewed in [73, 74]). With this concept in mind, studies reviewed by Faria et al. have explored the possibility of inducing tolerance by administering antigens via the oral route [74]. One study showed that oral administration of insulin prevented diabetes in specific diabetes mouse models. In the experimental autoimmune encephalomyelitis (EAE) model, widely used to study inflammatory responses towards the central nervous system and to better understand the mechanisms in the disease of multiple sclerosis, T cells mount an immune response towards myelin antigens. Oral administration of myelin basic protein (MBP) to rats and mice prior to disease induction resulted in suppression of clinical signs of disease [75, 25 76]. Furthermore, in humans, failure to induce tolerance towards the dietary gluten is thought to be the main cause of a Th1 type derived hypersensitive reaction seen in coeliac disease [77]. Oral tolerance towards dietary antigens is therefore an important part of intestinal homeostasis and it is important to understand how DCs and T cells respond to dietary antigens. Mediated by DCs, tolerance to oral antigens takes place in the MLNs [78]. After DCs acquire oral antigens from the intestinal lumen, they travel to the MLN in a CCR7 dependent manner [78, 79]. In the MLN, DCs present the orally derived antigens to CD4+ and CD8+ T cells inducing them to proliferate [80]. When DCs present oral antigens to CD4+ T cells, they are induced to differentiate into CD4+ Foxp3+ Tregs [34, 35, 81]. Gut homing receptors CCR9 and α4β7 integrin are upregualted by CD4+ Tregs. This allows CD4+ Tregs induced in the MLN to migrate to the inestinal mucosa where they can fully differentiate and exert their regulatory functions [73]. It is still not clear what differentiation profile CD8+ T cells take when cross primed by DCs. The mechanisms involved in avoiding unwanted CD8+ T cell immune responses towards oral antigens remains controversial. Some studies, reviewed by Faria et al. [74], show that in response to oral antigens CD8+ T cells differentiate into CTLs (Figure 1.5). Specifically, administration of OVA was said to generate OVA specific CTLs, which induced diabetes in a mouse model expressing OVA in the pancreas [82, 83]. Conversely, other studies showed orally administered OVA inhibited induction of OVA specific CTLs [71, 84]. It was shown that the intestinal CD11b+ CD8α- DC subset cross presented oral antigens to CD8+ T cells in the MLN for induction of oral tolerance [71]. Mice lacking the CD11b+ cells have a defect in establishing oral tolerance, which further demonstrates the importance of this subset [85]. Despite the conflicting results, the latter studies suggest an important role for cross presentation of oral antigens for suppressing generation of harmful CD8+ T cell responses. It can be speculated that DCs induce CD8+ T cell tolerance towards harmless oral antigens in a similar manner described for oral antigen specific CD4+ Foxp3+ Tregs. 26 1.2 Specific aims 1.2.1 CD8+ T cell responses towards oral antigens DCs play an important role in maintaining intestinal homeostasis and inducing tolerance towards oral antigens. These cells acquire, process and present oral antigens to naïve T cells in MLN. Moreover, DC antigen presentation to naïve CD4+ T cells induces their proliferation and differentiation into Tregs [34, 35, 81]. Cross presentation of oral antigens induces proliferation of CD8+ T cells [4] but aspects of this process are unclear. For example, what do CD8+ T cells differentiate when presented with orally derived antigens? What is the role of CD8+ T cells in oral tolerance and intestinal homeostasis? The first portion of my thesis addresses the CD8+ T cell responses generated towards oral antigens in vivo. This chapter demonstrates that CD8+ T cells do not express regulatory markers in response to oral antigens in sharp contrast to their CD4+ Treg counterparts but instead contribute to oral tolerance in a distinct manner. 1.2.2 Cross presentation of oral antigens for CD8+ T cell activation To induce CD8+ T cell immunity or tolerance towards exogenous antigens, DCs perform cross presentation. Trafficking of MHC-I through the cross presentation to access exogenous antigen is mediated by the YXXØ internalization motif in the cytoplasmic tail of MHC-I molecules or the CD74 chaperone protein [40, 50, 86]. The second portion of my thesis addresses the in vivo role of CD74 and the MHC-I YXXØ motif in cross presentation of orally derived antigens for CD8+ T cell proliferation and activation. The studies conducted here identify the DC antigen processing pathways necessary for cross tolerance. Specifically, the cytoplasmic domain of MHC-I is required for full activation and proliferation of CD8+ T cells in response to oral antigens. 27 Chapter 2. Materials and Methods 2.1 Mice C57Bl/6 mice were purchased from Charles River and The Jackson Laboratory. OT-I T cell transgenic and B6.Rag1-/- were acquired from The Jackson Laboratory. H-2Kb2Db double knockout (H-2K-/-D-/-) [87] were a kind gift from Dr. John W. Chamberlain (The Hospital for Sick Children, Toronto, Canada). OT-II T cell transgenic mice were a kind gift from Dr. Jan Dutz (Child and Family Research Institute, University of British Columbia) and Dr. Colby Zaph (Biomedical Research Centre, University of British Columbia). C3H/He mice expressing transgenic wild-type MHC I allele H-2Kb (WTKbC3H) or cytoplasmic tail tyrosine mutant (ΔYKbC3H) have been previously described [40, 86]. CD74-/- transgenic mice were a gift of Diane Mathis (C.U. Strasbourg, France and The Harvard Stem Cell Institute, Boston, MA). All mice were bred and maintained at the University of British Columbia Small Animal Facility at South Campus and Animal Resource Unit (ARU). Mice were fed standard lab chow and water ad libitum and kept under a 12 hour light/dark cycle. All studies followed guidelines set by both the University of British Columbia’s Animal Care Committee and the Canadian Council on Animal Care. ΔYKb H2D-/- and WTKb H2D-/- strains were generated for this thesis by crossing H2Kb2Db double knockout (H-2K-/-D-/-) with WTKbC3H and ΔYKbC3H strains, respectively. WTKbC3H and ΔYKbC3H strains were maintained on a C3H/He background and express MHC-I genes H-2Kk and 2Dk as well as the knocked in transgenic H-2Kb gene (ΔYKb or WTKb). Progeny of the crosses were genotyped by PCR and phenotyped by flow cytometry (as described below) to establish the presence of MHC I alleles H-2Kk, Kb, Dk and Db as well as the disrupted H-2Kb and Db genes. The mice were backcrossed onto a C57BL/6 background by breeding them with H-2K-/-D-/- strain (H2Kb2Db double knockout). 28 2.2 Genotyping by PCR Mice were genotyped by PCR using DNA isolated from earclips. Briefly, earclips were incubated for 90 minutes at 55°C in 20 µL of digestion buffer (50 mM Tris-Cl [pH 8.3], 100 mM NaCl, 1% SDS, 100 mM EDTA pH 8.0) and 10 µg/ml proteinase K (Fermentas). Samples were vortexed every 30 min. The reaction was diluted by addition of 980 µL MilliQ H2O and incubated at 95°C for 5 min to inactivate Proteinase K; the DNA was then used as template for a PCR. Primers to detect the presence of disrupted MHC-I alleles H-2Kb and Db genes have been previously described [87]. PCR to detect presence of knocked in transgenic H-2Kb (ΔYKb or WTKb) used H-2Kb specific oligonucleotides, 5’-TCGCTGAGGTATTTCGTC-3’ and 5’-TTGCCCTTGGCTTTCTG T-3’, as previously described [86]. Primers used to detect MHC-I alleles H-2Kk and Dk were H2K&D(k)_fwd: 5'- GGAAGCCCCGGTTCATCTCT -3', H2K(k)_rev: 5'- ACAGCCGTACATCCGTTGGAAC - 3' and H2D(k)_rev: 5’-CCGGACAACCGCTGG ATC-3’. 2.3 Antibodies and flow cytometry Antibodies against H-2Kb (AF.6-88.5), I-Ab (AF6-120.1), CD86 (GL1), CD4 (GK1.5), CD8a (53-6.7), CD11b (M1/70), CD11c (HL3), CD40 (3/23), CD44 (IM7), CD25 (PC61.5), CD62L (MEL-14), CD69 (H1.2F3), PD-1 (J43) were obtained from BD Biosciences. Antibodies against Foxp3 (FJK-16s), CD8a (53-6.7), CD4 (GK1.5) were purchased from eBioscience. CD3 (145-2C11) and CD44 (IM7) antibodies were obtained from the UBC Antibody Facility (AbLab). The H-2Kb/OVA257–264 (25.D1.16) antibody was purified from the supernatant of the 28.14.8S hybridoma and directly conjugated to Alexa-649 (A649) by the AbLab. Cells stained with antibodies against surface proteins were incubated on ice with saturating amounts of antibody for 30 minutes. The cells were then washed twice with cold PBS containing 2% FBS. For intracellular staining with anti-IFN-γ (3F11; BD Biosciences) and Foxp3, the Foxp3-Staining Buffer Set (Fixation/Permeabilization and Permeabilization buffers; eBioscience) was used according to the manufacturer's protocol. Data were acquired using either FACScan or FACSCalibur/CellQuest software 29 (BD Biosciences) or LSRII/FACSDiVa software (BD Biosciences). Data were analyzed with Flowjo software (Treestar, Inc). 2.4 Phenotyping by flow cytometry Expression of specific MHC-I molecules on the cell surface of lymphocytes was assessed by flow cytometry. Murine blood was collected from the saphenous vein in heparin coated microvette tubes (Sarstedt). To isolate peripheral blood leukocytes (PBLs), whole blood was transferred with 200 uL of PBS to BD Falcon 5 mL polystyrene round-bottom tubes. An 800 µL Ficoll gradient was applied and tubes were spun for 15 minutes at 500g. PBLs were recovered and washed twice in RMPI completed media (RPMI-1640 supplemented with 10% fetal calf serum (FCS)). The cells were incubated with appropriate antibodies as outlined in Section 2.3. PBLS from progeny of H2Kb2Db double knockout (H2K-/-D-/-) with WTKbC3H and ΔYKbC3H strains were stained for antibodies specific to extracelluar domains of H-2Kb, Kk and Dk, respectively. Phenotyping of CD74-/- mice was performed on PBLs using antibodies specific to CD4 and CD8 as described above and in Section 2.3. Mice containing less than 10% CD4+ PBLs were considered to be CD74-/-. 2.5 Lymphocyte isolation from spleen, lymph nodes and thymus Spleen, thymus and mesenteric lymph nodes (MLN) were harvested and mashed through a wire mesh to disrupt the connective tissue and create a single-cell suspension. Red blood cells were lysed with MRCRB buffer (0.15 M NH4Cl, 0.01M Tris base pH 7.2) for 2-3 minutes at room temperature. The reaction was quenched with excess RPMI completed media. The resultant single-cell suspension was used in further assays as outlined. 2.6 Adoptive T cell transfer of CFSE labeled cells OVA-specific T cell proliferation was assessed using CFSE- labeled OT-I cells. Spleen and MLN from OT-I transgenic mice were made into single-cell suspensions as described in Section 2.5. CD8+ T cells were isolated using the appropriate EasySep 30 Mouse selection kits from Stem Cell Technologies (Vancouver, Canada). Afterwards, cells were labeled with 1.5 µM CellTrace CFSE (Invitrogen) for 10 minutes at 37°C in 0.1% BSA/PBS as indicated by manufacturer. Labeled cells were washed twice with RPMI completed media and resuspended in PBS. CFSE labeled CD8+ T cells were then injected intravenously (i.v.) at 5-10 X 106 cells/mouse in 200 µL of PBS. 2.7 Induction of oral priming When indicated, mice received a 2% Grade VI Ovalbumin (OVA; Sigma-Aldrich) solution dissolved in drinking water (20 mg/mL) that was replaced every 48 hours for 5 consecutive days. Spleens and MLN were harvested on day 6 and lymphocytes were analyzed by flow cytometry. Alternatively, specified doses of Grade VI OVA (Sigma- Aldrich) dissolved in PBS were administered to mice intragastrically. Mice receiving adoptively transferred T cells were administered OVA intragastrically at 18 h after i.v. injection. 2.8 CFSE proliferation assay and tetramer staining Spleens and MLN from mice receiving the adoptively transferred CFSE labeled T cells were harvested 48 hours after oral administration of Grade VI OVA (Sigma- Aldrich). Single-cell suspensions were stained and analyzed by flow cytometry (as in Section 2.3). To identify the CFSE labeled OVA257-264-specific CD8+ T cells, single-cell suspension from MLN and spleen were stained with CD8α (53-6.7) antibody, H- 2KbOVA257-264 tetramer (iTag MHC Tetramer, Beckman Coulter) and when required CD44 (IM7) antibody. Data was acquired by flow cytometry using BD™ LSR II Flow Cytometer and CFSE dilution of tetramer positive cells was analyzed using FlowJo software (TreeStar). 2.9 Intracellular staining for Foxp3+ Isolated lymphocytes (1 x 106) were incubated with corresponding cell surface marker antibodies for 30 minutes at 40C and washed twice with PBS/2% FCS. For intracellular staining of Foxp3, the Foxp3-Staining Buffer Set (Fixation/Permeabilization and 31 Permeabilization buffers; eBioscience) was used according to the manufacturer's protocol. Briefly, after surface marker staining, cells were washed with PBS/2% FCS. The Fixation/Permeabilization buffer was then added and cells were incubated for 40 minutes in the dark. Cells were washed once with 1X Permeabilization buffer. Alexa647- conjugated Foxp3 antibody (eBio FJK-16s) diluted in 1X Permeabilization buffer was added to cells and incubated for 40 minutes on ice, protected from light. After washing twice with 1X Permeabliziation buffer, cells were resuspended in PBS/2% FCS. Data acquisition was performed on an LSRII machine with FACSDiVa software or FACSCalibur with CellQuest software (BD Biosciences). Data were analyzed with FlowJo software (TreeStar). 2.10 Intracellular IFN-γ staining IFN-γ production was assayed by flow cytometry. Lymphocytes from spleen and MLN were isolated as described in Section 2.5. Lymphocytes were incubated for 5 hours in a 96-well flat bottom plate in 0.2 mL of RPMI completed medium with 1 µL/mL Golgi Stop (BD Biosciences) to block cytokine secretion. Cells were either stimulated with 1 µM of the H-2Kb-restricted peptide OVA257–264 (SIINFEKL), through the TCR with anti- CD3ε antibody or left unstimulated in media alone. For anti-TCR stimulations, lymphocytes were incubated in wells that were precoated with 10 µg/mL anti-CD3ε antibody (145-2C11) in PBS. Following the incubation period, the lymphocytes were fixed and permeablized using Foxp3-Staining Buffer Set then stained with CD8 and IFN- γ antibody (XMG1.2; BD Biosciences). 2.11 Generation of bone marrow derived dendritic cells To generate immature bone marrow derived dendritic cells (bmDCs), the femurs and tibia bones were harvested from the appropriate mice strains. The bone marrow was flushed out with PBS and washed with PBS twice. The bone marrow was cultured in RPMI complemented media with 1% X63-Ag8-plasmacytoma-derived GM-CSF for indicated number of days. 32 2.12 Transfection of bone marrow derived dendritic cells BmDCs were transfected with constructs containing full-length murine CD74 (p31 isoform; FL) and CD74 lacking part of its cytosolic domain, amino acids 2–17 (Δ2–17) in the pBabe vector. The CD74 constructs in pBabe vector were a kind gift from Idit Shachar (Departments of Immunology, The Weizmann Institute of Science, Israel). Immature BmDCs were harvested at day 6 of culture as described above. Cells (3 × 106) were resuspended in 100 µL of Amaxa buffer (Amaxa ® Mouse Dendritic Cell Nucleofector® Kit) and mixed with 32 µg of the appropriate DNA construct. The 100 µL of cell suspension was transferred to a 2.0 mm electroporation cuvette, and electroporated with an Amaxa Nucleofector apparatus following the directions of the manufacturer. Transfected cells were transferred to a 48-well plate (1 × 106 cells/well) in RPMI completed media containing GM-CSF. One day after electroporation, cells were incubated with 20 mg/mL OVA, 1µM OVA257- 264 (SIINFEKL) peptide or media alone for 8 hours. OT-I CFSE labeled cells were added (1 × 106 cells/well) to the wells and CFSE dilution was assayed 3 days later by flow cytometry. 2.13 Cross presentation assay Cultured bmDCs were harvested on day 8 or 10 of culture. Cells were transferred to a 48-well plate (1 × 106 cells/well) and incubated with 20 mg/mL OVA, 1µM OVA257- 264 (SIINFEKL) peptide or media alone. Following an overnight incubation, cells were stained with antibodies against CD11c, H2Kb and antibody that detects H-2Kb/ OVA257- 264 complexes and then analyzed by flow cytometry. Alternatively, after an 8 hour incubation, with either OVA (20 mg/mL ), OVA257- 264 (SIINFEKL) peptide or media alone OT-I CFSE labeled T cells were added (1 × 106 cells/well) and CFSE dilution was assayed 3 days later by flow cytometry. 2.14 Statistical analysis All experiments were performed at least three times in triplicate. Statistical comparisons of data between treatment groups or transgenic mice and wild-type controls were performed using Student's t-test for unpaired values. All statistical analyses were 33 performed using Graphpad Prism software (version 5, La Jolla, CA). p-values less than 0.05 were considered significant. Values are expressed as mean ± standard deviation (SD). 34 Chapter 3. CD8+ T Cell Responses Towards Oral Antigens 3.1 Rationale The intestine is in constant contact with foreign antigens, the majority being harmless food derived antigens and commensal bacteria. Maintaining intestinal homeostasis is important and a disruption in this balance can lead to unnecessary inflammation in the intestine causing IBD, which includes ulcerative colitis and Crohn’s disease [24]. The cause of these various disruptions of intestine homeostasis varies from patient to patient and the exact mechanisms still need to be elucidated. However, it has been proposed that cytotoxic CD8+ T cells contribute to the pathogenesis of IBD [8, 88]. Oral administered antigens are normally seen as harmless by the immune system. These antigens gain access to the intestinal lymphoid tissue and LP where APCs, such as DCs process and present the antigens to T cells [78, 80]. This induces a type of immune tolerance known as oral tolerance[73]. The processing and presentation of oral antigens to naïve CD4+ T cells induces them to differentiate into CD4+ T cells with regulatory function (CD4+ T reg cells) [34, 35]; this is believed to be an important mechanism involved in oral tolerance. CD4+ Treg cells are also implicated in having a protective role in IBD by suppressing inflammatory responses in the gut [8]. Contrary to the response of CD4+ T cells towards oral antigens, there is not much known about the response of CD8+ T cell that are induced to the same oral antigens. It would be beneficial to better understand CD8+ T cell responses that are induced towards food derived antigens to prevent unwanted induction of cytotoxic CD8+ T cells in the intestine involved in intestine inflammation. The objective of this chapter is to characterize the CD8+ T cell responses that are induced towards food-derived antigens. These studies will help to compare between CD8+ and CD4+ T cell immune responses that are generated towards the model antigen, OVA. The working hypothesis of this chapter is that CD8 T cells differentiate in response to oral antigen. CD8+ T cell responses will be analyzed by flow cytometry measuring cell proliferation and expression of specific T cell activation markers in mice receiving oral OVA. Analysis of CD8+ T cell responses towards oral antigens will help to better understand their role in oral tolerance. It is hoped these interactions will shed light on 35 pathways used by the intestinal immune system to prevent CD8+ T cell mediated immunity towards oral antigens useful for treating IBD. 3.2 Results 3.2.1 CD4 Foxp3+ T cells are generated in response to orally administered OVA. Previous studies have shown that oral OVA induced CD4+ Foxp3+ from Foxp3- precursor cells [34]. Induction of antigen specific CD4+ Foxp3+ is considered part of the mechanism involved in oral tolerance[73]. It was therefore of interest to investigate if CD8+ T cells are also induced to express regulatory marker Foxp3 in response to oral OVA. However, before assessing induction of CD8+ Foxp3+ it was confirmed that oral delivery of OVA was being done effectively. For this purpose, OT-II mice, which express a transgenic TCR that recognizes the OVA peptide SIINFEKL in context of MHC-II IAb, were given OVA in their drinking water for 5 days (Fig 3.1A). As expected and previously shown [35], expression of T cell regulatory markers Foxp3 and CD25 in CD4+ T cells in the spleen and mesenteric lymph nodes (MLN) was higher in mice receiving OVA (Fig 3.1). On average mice receiving oral OVA had a 2-fold increase in the frequency of Foxp3+ CD4+ T cells in the MLN and a 1.5-fold increase in frequency in the spleen when compared to the controls (p<0.001; Figure 3.1B). Frequency of CD25 expression in CD4+ T cells was also statistically higher in mice receiving OVA (4.8% vs 3%; p < 0.01 Figure 3.1C). This result agrees with previous studies and confirms that orally administered OVA induces up-regulation of Treg markers Foxp3 and CD25 in antigen specific CD4+ Treg cells [34, 35]. The same administration of oral OVA was therefore used to further examine possible induction of Foxp3+ in CD8+ T cell in response to oral antigen. 36 37 Figure 3.1 OT-II mice have a higher frequency of Foxp3+ CD4+ T cells after oral administration of OVA protein. (A) A schematic representation of the protocol is outlined. OT-II mice were given OVA in the drinking water for 5 days. On day 6, T reg cell frequency was assessed in themesenteric lymph nodes (MLN) and spleen (Sp). (B) Flow cytometry analysis of Foxp3 expression in CD4+T cells of OT-II mice. Left: representative dot plot where values represent the percentage of Foxp3+ cells from total CD4+ cells. Right: the mean percentage of Foxp3+ cells of total CD4+ cells from either MLN or Sp as indicated (*<0.001). (C) Flow cytometry analysis of CD25 expression in CD4+ T cells in OT-II mice. Left: representative dot plot, value represents percentage of CD25+ cells of total CD4+ cells. Right: the mean percentage of CD25+ cells of total CD4+ cells from either MLN or Sp as indicated (**p<0.01). All dot plots are gated on CD4+ T cells. Graphs in all panels show mean ± SD (n=3). Data shown are representative of two independent experiments with similar results. 38 3.2.2 Orally administered OVA does not induce Foxp3 expression in OT-I cells. CD8+ Foxp3+ T cells in the MLNs have been characterized previously and shown to be activated in response to endogenously expressed antigen in the gut and gut associated lymph nodes [21]. Since oral antigen can induce expression of the transcription factor Foxp3 in CD4+ T cells [34, 35] (Fig 3.1) it was of interest to examine if administration of oral antigen induces Foxp3 expression in CD8+ T cells. In order to address this question, OT-I mice, which express a transgenic TCR that recognizes an OVA peptide in context of MHC-I (H-2Kb), were given OVA in their drinking water for 5 days (Figure 3.2A). This method of delivery provides a constant source of antigen that will induce stimulation in CD8+ T cells in the gut associated MLN. Following administration of OVA for 5 days, lymphocytes from spleen and MLN were analyzed by flow cytometry. The lymphocyte population of the MLNs contained Foxp3+ cells (Figure 3.2B); however the CD8+ T cells in OT-I mice receiving OVA or PBS were not positive for Foxp3+. The Foxp3+ population was shown to be comprised of CD4+ T cells (Figure 3.2B). Unlike CD4+ T cells in OT-II mice, it can be concluded that CD8+ T cells in OT-I mice do not express regulatory T cell marker Foxp3 after a 5-day administration of OVA in the drinking water. 39 Figure 3.2 The Foxp3+ CD8+ T cell population in OT-I mice does not increase after oral administration of OVA protein. (A) A Schematic representation of the protocol is outlined. OT-I mice were given OVA in drinking water for 5 days. On day 6, Treg cell frequency was assessed in mesenteric lymph nodes (MLN). (B) Flow cytometry analysis of Foxp3 expression in lymphocytes of OT-I mice. Plots were gated on the total lymphocytes and values in plots represent percentage of Foxp3+ CD8+ cells or Foxp3+ CD4+ cells of total MLN lymphocyte population. Data shown are representative of three independent experiments with similar results. 40 3.2.3 OT-I cells express activation markers after orally administered OVA. DCs acquire and present oral antigen in the context of MHC-II to CD4+ T cells leading to the differentiation of Foxp3+ CD4+ T cells from naïve Foxp3- CD4+ T cell precursors. However, it is unclear as to what type of response is generated when oral antigens are presented to CD8+ T cells in the context of MHC-I. Therefore, CD8+ T cells were analyzed for the expression of specific surface activation markers following the administration of oral antigen. For this purpose OT-I mice received a single dose of OVA via oral gavage and after 72 hours CD8+ T cells from MLN and spleen were analyzed by flow cytometry (Figure 3.3A). OVA administered by this dosing protocol has previously been shown to be processed and presented by DCs to CD4+ and CD8+ T cells found in the MLN [71, 78, 80]. Here, a higher percentage of CD8+ T cells from OT-I mice receiving a single dose of OVA expressed the CD44+ activation marker on their cell surface compared to the control mice receiving oral PBS (Figure 3.3B). CD44 surface expression is upregulated in the earliest phases of clonal expansion and is the most commonly used marker for distinguishing activated T cells from their naïve precursors [89]. On average, 70% of the total CD8+ T cells in the MLN of OT-I mice expressed CD44 on the cell surface when receiving OVA, compared to 5% in the control (p<0.01; Figure 3.3C). In the spleen on average, 75% of the total CD8+ T cells in the MLN of OT-I mice expressed CD44 on the cell surface when receiving OVA, compared to 10% in the control (p<0.001; Figure 3.3C). Contrary to CD44 expression, surface CD62L levels are known to be downregulated in effector cells [12]. CD62L, a homing receptor expressed on naïve T cells, is required for T cells to gain entry into lymph nodes and is downregulated upon activation to allow T cells migrate to peripheral tissue. Treatment with oral OVA generated a larger population of CD8+ T cells with downregulated CD62L, also referred to as (CD62Llo; Figure 3.3D). On average, 15% of the total CD8+ T cells in the MLN of OT-I mice downregulated the surface expression of CD62L when receiving OVA while only 2% of total CD8+ T cells in the control were CD62Llo (p<0.05; Figure 3.3E). For lymphocytes in the spleen on average, 20% of the total CD8+ T cells in the MLN of OT-I mice downregulated the surface expression of CD62L when receiving OVA while only 6% of 41 total CD8+ T cells in the control were CD62Llo. Differences in CD62L downregulation for CD8+ T cells in the spleen were not statistically significant between mice receiving OVA and the controls (p=0.06; Figure 3.3E). Additionally, the program death 1 (PD-1) and CD45RB were also examined. PD-1 and CD45RB are surface markers also upregulated in effector T cells and therefore used to assess T cell activation [90, 91]. PD- 1 and CD45RB surface levels were increased on CD8+ T cells of mice given oral OVA as compared to PBS controls (Figure 3.3F). Altogether these results demonstrate that CD8+ T cells encountered the oral antigen and became activated as shown by increased surface expression of CD44, PD-1 and CD45RB and downregulation of surface CD62L. 42 43 Figure 3.3 Activation markers are increased on surface of OT-I cells following oral administration of OVA. (A) A Schematic representation of protocol is outlined. OT-I mice received 40mg of OVA or PBS (control) by oral gavage. 3 days later single cells were isolated from the spleen (Sp) and mesenteric lymph nodes (MLN) and analyzed by flow cytometry. (B) Flow cytometry analysis of CD44 expression in CD8+ T cells of OT-I mice. Values in dot plots indicate the percentage of CD44lo and CD44hi population of total CD8+ T cells. (C) Mean percentage of CD44hi and CD44lo of total CD8+T cells in OT-I mice. (D) Flow cytometry analysis of CD62L expression in CD8+ T cells of OT-I mice. Values in dot plots indicate the percentage of CD62Llo and CD62Lhi populations of total CD8+ T cells. (E) Mean percentage of CD62Lhi and CD62Llo of total CD8+ T cells in OT-I mice. All flow cytometric dot plots are gated on CD8+ T cells. (F) Representative histograms show expression of activation markers CD44, CD62L, CD45RB and PD-1in CD8+T cells of OT-I mice. Graphs in all panels show mean ± SD (n=3). Statistical comparisons were performed using the Student’s t test; *,p < 0.05; **,p < 0.01; ***,p < 0.001. Data shown are representative of two independent experiments with similar results. 44 3.2.4 OT-I cells are activated to secrete IFN-γ in response to oral OVA. The analysis of surface markers on CD8+ T cells encountering oral antigens suggested that the T cells take on an effector phenotype. Therefore, IFN-γ production by CD8+ T cells from OT-I mice fed OVA was assessed. To this end, lymphocytes from the spleen and MLN from OT-I mice receiving OVA were stimulated in-vitro with MHC class I- restricted peptide epitope OVA(257–264), also referred to as SIINFEKL, and monitored for intracellular IFN- γ by flow cytometry. CD8+ T cells in the MLN of mice receiving oral OVA had a statistically significant higher production of IFN-γ compared to the control (Figure 3.4A and 3.4B). On average, 9% of the total CD8+ T cells in the MLN of OT-I mice were positive for IFN-γ production when stimulated with SIINFEKL, compared to 0.8% in the control (p<0.05; Figure 3.4B). Contrary to the CD8+ T cells in the MLN, CD8+ T cells from the spleen produce similar levels of IFN- γ as the PBS control when stimulated with the peptide (Figure 3.4C and 3.4D). Overall the CD8+ T cells from the MLN, previously shown to express markers of activation in response to oral OVA, also produced IFN- γ when re-stimulated in vitro. 45 Figure 3.4 CD8+ T cells secrete IFN-γ in response to oral administration of OVA protein. Three days after 40mg administration of oral OVA, OT-I mice were sacrificed and T cells were assessed for IFN-γ production. Lymphocytes were isolated from spleen and mesenteric lymph nodes (MLN) and were stimulated with the MHC class I OVA 257-264 restricted peptide (SIINFEKL), α-CD3 antibody or left untreated and assayed for IFN- γ production. (A) Flow cytometry analysis of IFN-γ expression in CD8+ T cells of MLNs. (B) Mean percentage of IFN-γ expressing cells in MLNs. (C) Flow cytometry analysis of IFN-γ production in CD8+ T cells in the spleen. (D) A graph depicting mean percentage of IFN-γ expressing cells in the spleen. All dot plots are gated on CD8+ T cells, and values represent the percentage of IFN-γ secreting cells. Graphs show mean ± SDs (n=3). *p=0.05; **p<0.01; ***p<0.001. Data shown are representative of two similar independent experiments. 46 3.2.5 OT-I cells do not show increased cytolytic function in response to oral OVA. To evaluate whether the increase in IFN-γ production CD8+ T cells in MLN of mice receiving oral OVA is associated with cytolytic activity a 51Cr release assay was preformed. It has been shown previously that oral infection with OVA expressing Listeria monocytogenes generated CD8+ T cells capable of killing target cells at 9 days post infection [63]. Therefore, the cytolytic function of the CD8+ T cells in MLN of OT-I mice was analyzed 8 days after administration of oral OVA (Figure 3.5A). CD8+ T cells were incubated with 51Cr-labeled RMA-S target cells that were either untreated or pulsed with OVA peptide. 51Cr release was measured and compared between MLN cells from OT-I mice receiving OVA or PBS alone (Figure 3.5B). No difference in cytolytic activity was observed between cells from OT-I mice receiving OVA or PBS. This result shows oral administration of OVA induces CD8+ T cells in the MLN of OT-I mice to produce IFN- γ but not to generate cytolytic function. 47 Figure 3.5 No increased CTL responses of OT-I cells after oral OVA. (A) A schematic diagram of the timeline of the 51Cr-release assays performed ex vivo to measure specific killing of target cells. OT-I mice were given 40mg of OVA or PBS by oral gavage and 8 days later lymphocytes from mesenteric lymph nodes (MLN) were isolated and lytic activity was measured. (B) 51Cr-release assays to measure specific killing of target cells. The graph shows mean ± SD. 48 3.2.6 Adoptively transferred OT-I cells display activation markers in response to oral OVA The peripheral lymphocyte population of OT-I mice has extensive skewing towards the CD8+ T cell population [92]. To eliminate the possibility that responses in OT-I mice do not represent a physiologically relevant scenario, CD8+ T cell responses towards oral antigens were analyzed in a wild type C57Bl/6 mouse strain. To this end, CFSE-labeled OVA-specific CD8+ T cells were transferred to C57Bl/6 mice by intravenous injection. Afterwards mice received OVA or PBS by oral gavage. The lymphocytes from MLN and spleen were isolated and analyzed by flow cytometry (Figure 3.6A). Gating on tetramer+ CD8+ and CFSE+ cells allowed specific analysis of the CFSE OT-I cells that had been injected prior to oral treatment. Given the fact that OVA specific T cells showed activation in OT-I mice after oral OVA, adoptively transferred OT-I cells were also examined for expression of activation markers. Similar to what was seen in OT-I mice, a higher proportion of the adoptively transferred OT-I cells from the spleen and MLN of mice receiving oral OVA were CD44hi (spleen: 30% vs 13, MLN: 42% vs 7% Figure 3.6B,C). Moreover, a higher percentage of the adoptively transferred OT-I cells from mice receiving oral OVA expressed PD-1 (13% vs 1.5%; p<0.001; Figure 3.6D). Again, this observation corresponds to that observed with OT-I mice that received oral OVA. Taken together, these data demonstrate that antigen specific cells when found either in OT-I mice or C57Bl/6 mice, upregulate T cell effector markers in response to 40mg of oral OVA. 49 50 Figure 3.6 Oral administration of OVA induces expression of activation markers in OT-I T cells. A) A schematic representation of the protocol is outlined. CFSE-labeled OT-I cells were injected i.v. into C57BL/6 mice. 18 hours after the transfer, the mice were fed 40 mg OVA in PBS or PBS alone by oral gavage. Two days later, single cell suspensions of spleen (Sp) and mesenteric lymph nodes (MLNs) were analyzed by flow cytometry. Cells were stained with CD8+ and H2Kb-OVA specific tetramer for purposes of gating and analysis. (B) Flow cytometry analysis of CD44 expression of donor OT-I cells. Values in contour plots indicate the percentage of CD44hi and CD44lo cells. (C) Mean percentage of CD44hi and CD44lo cells of donor OT-I cells. (D). Flow cytometry analysis of PD-1 expression of donor OT-I cells. Values in contour plots show percentage of PD-1 high cells. (E) Mean percentage of PD-1hi cells of donor OT-I cells. All flowcytometric contour plots are gated on (CFSE+ Tetramer+ CD8+ T cells donor OT-I cells). Percentages shown are of total OT-I cells. Graphs in all panels show mean ± SDs (n=3). *p=0.05; **p<0.01; ***p<0.001. Data shown are representative of two similar independent experiments 51 3.2.7 Oral administration of OVA protein induces proliferation of OT-I cells. CFSE-labeled T cells adoptively transferred into C57Bl/6 mice receiving OVA were analyzed for proliferation (Figure 3.7A). During cell division of CFSE-labeled cells, the dye is distributed equally between the daughter cells and reduction of fluorescent intensity can be used as an indicator of cell proliferation (Figure 3.7A). In accordance to previous studies [71, 80, 93], analysis of CFSE dilution showed that adoptively OT-I cells proliferated in response to oral OVA (Figure 3.7B). OT-I cells transferred to mice that did not receive OVA did not proliferate. Cell proliferation was also greater in cells found in the MLN than in the spleen with 91.6% of cells in the MLN proliferating compared to 54% found in the spleen. 52 Figure 3.7 Oral administration of OVA induces proliferation of OT-I cells. (A) A schematic diagram showing the gating strategy used to analyze OT-I donor cells transferred into C57Bl/6 mice. CFSE dilution histograms represent the CD8+ Tetramer+ CFSE+ cells. (B) Flow cytometry histograms depicting proliferation of donor OT-I cells in the MLN or spleen of C57BL/6 mice that received 40mg of OVA or PBS (control) by oral gavage. Values indicate the percentage of cells that have proliferated. Data shown are representative of three independent experiments. 53 As CD8+ T cells in MLN of OT-I mice that received oral OVA had higher production of IFN-γ, adoptively transferred OT-I cells were also examined for IFN-γ production. Lymphocytes isolated from the MLN were stimulated with the SIINFEKL peptide, a MHC class I-restricted epitope for OVA(257–264), and intracellular IFN-γ production was analyzed by gating on the CFSE+ cells (Figure 3.8). Results showed that CD8+ T cells from mice receiving OVA stimulated with SIINFEKL had a trend towards higher IFN-γ production; however, this was not statistically significant (p=0.2, n=3). OT-I cells adoptively transferred to C57Bl/6 mice showed proliferation and showed trend towards higher production of IFN-γ in response to oral OVA. 54 Figure 3.8 IFN-γ expression in OT-I cells transferred into C57Bl/6 mice is not significantly different after oral administration of OVA. IFN-γ expression of transferred OT-I T cells in MLN and spleen of C57BL/6 mice that received 40mg of OVA or PBS (control) by oral gavage 18 h after transfer. IFN-γ expression of OT-I T cells was measured 48 h after antigen administration by flow cytometry. (A) Contour plot showing percentage of CD8+ IFN-γ+ cells after stimulation with MHC class I OVA 257-264 restricted peptide (SIINFEKL), anti-CD3 antibody or no peptide. Flow cytometric contour plots were gated on CFSE+ H2Kb-OVA specific tetramer+CD8+T cells. Numbers in contour plots indicate the percentage of IFN-γ secreting cells. (B) A graph depicting mean percentage of IFN-γ secreting cells of total OT-I donor cells. Graphs show mean ± SDs (n=3). *p=0.05; **p<0.01; ***p<0.001. 55 3.2.8 OT-I cells in C57Bl/6 mice are not induced to express Foxp3 in response to oral OVA Our observations reveal that adoptively transferred OT-I cells proliferate in response to oral antigen (Figure 3.7). It has been previously demonstrated that acquisition and presentation of oral OVA to CD4+ T cells by DCs in the MLN induced proliferation and differentiation of CD4+ Foxp3+ Treg cells from naïve precursors [34, 35]. It was therefore of interest to assess the possible induction of regulatory T cell markers, Foxp3 and CD25, in CD8+ T cells in response to oral OVA. For this purpose, cells from the MLN of C57Bl/6 mice receiving an adoptive T cell transfer and oral OVA were assessed for Foxp3 expression by flow cytometry. Naïve C57Bl/6 mice have a small endogenous CD8+ Foxp3+ cell population (Figure 3.9A). This population of CD8+ Foxp3+ T cells was shown to be at least 20-fold lower than the CD4+ Foxp3+ population suggesting that a great number of OT-I cells need to be analyzed for possible Foxp3+ expression. Therefore, all lymphocytes collected from MLN were stained accordingly and analyzed by flow cytometry. The adoptively transferred OT-I cells did not show expression of regulatory T cell marker Foxp3 and CD25 in mice receiving oral OVA (Figure 3.9C). Similar to what has been reported for CD4+ OVA specific T cell [34, 35], OT-I cells also proliferate in response to oral antigen (Figure 3.9B). However, unlike CD4+ OVA specific T cells CD8+ OT-I cells were not induced to express regulatory markers Foxp3 and CD25 (Figure 3.9C and Figure 3.9D). 56 Figure 3.9 OT-I T cells do not express Foxp3 in C57Bl/6 mice before or after oral administration of OVA protein. (A) Flow cytometry analysis of Foxp3+ expression in CD4+ and CD8+ T cells in MLNs of C57BL/6 mice. Values indicated the percentage of either Foxp3+ CD4 or Foxp3+ CD8 cells (B) CFSE dilution representing proliferation of donor OT-I cells in MLN and spleen of C57BL/6 mice that received OVA or PBS after transfer. Representative histograms of CFSE dilution were gated on CFSE+ H2Kb-OVA specific tetramer+ CD8+ T cells and numbers indicate percentage of cells that proliferated. (C,D) Flow cytometry analysis of Foxp3 and CD25 expression of donor OT-I cells. Values in contour plots indicate percentage of CFSE+ and CFSE- cells expressing (C) Foxp3 or (D) CD25. FACS plot are gated on CD8+ T cells. 57 3.3 Discussion 3.3.1 CD8+ T cell responses towards oral antigens In the current study, we investigated the CD8+ T cell response following oral delivery of OVA by examining activation and proliferation of OVA-specific OT-I CD8+ T cells in two models. In the first model, endogenous CD8+ T cells in OT-I mice were analyzed after receiving oral OVA. In the second model, OT-I cells were adoptively transferred to syngeneic C57Bl/6 mice then OVA was delivered by oral gavage. Using these two models, it was demonstrated that CD8+ T cells were activated and proliferated in response to the orally delivered antigen OVA. These responding cells did not have cytolytic activity but were induced to produce IFN- γ. Oral delivery of antigen clearly lead to activation of CD8+ T cells primarily in the MLN (Figure 3.6 and Figure 3.7). In this study, responding CD8+ T cells upregulated the surface expression of several activation markers including CD44, CD62L and PD-1 in response to administration of oral OVA. Furthermore, delivery of oral antigen induced expansion of CD8+ T cells, as seen by the dilution of CFSE labeled cells, when cells were transferred to C57Bl/6 mice (Figure 3.7). Activated and dividing cells were found in both the MLN and spleen; however, the efficiency was much higher in MLN. These results are consistent with previous findings showing DCs acquire exogenous oral antigens and travel to the MLN where they encounter and activate T cells presumably via cross presentation [78, 80, 93]. The work in this thesis along with previous studies demonstrates that presentation of oral antigen to T cells is primarily confined to sights of the intestinal immune system such as Peyer’s patches and MLN [78, 80, 93]. Oral tolerance, a mechanism of maintaining unresponsiveness and tolerance to oral antigens, is likely regulated by DCs found in the intestine and MLN that induce differentiation of Foxp3+ CD4+ regulatory T cells from naïve CD4+ Foxp3- precursors [34, 35, 73]. The work of this thesis agrees with the previous findings and shows that oral antigen can induce the accumulation of antigen specific Foxp3+ CD4+ T cells (Figure 3.1). However, in contrast to CD4+ T cells, CD8+ T cells were not induced to express the Treg marker Foxp3 in response to oral antigen (Figure 3.2 and Figure 3.8). This indicates 58 that responses induced towards oral antigens are different between CD4+ and CD8+ T cells and antigen-specific CD8+ T cells have other yet-to-be identified mechanisms to avoid inducing harmful responses towards benign oral antigens. Oral tolerance refers to the immune suppression of cellular and humoral responses when an antigen is fed prior to antigen priming [73]. Various studies previously reviewed by Faria et al. [74], examined the cellular immune response after administered of an antigen, such as OVA or insulin, via the oral route. Orally administered insulin in these mouse models suppressed or delayed induction of diabetes. Furthermore, orally administered OVA did not induce cells with cytolytic function after mice were primed with OVA-loaded syngenic splenocytes [71]. In contrast, other studies showed orally administered OVA did generate OVA specific cytotoxic T cells, which can induce diabetes in a mouse model expressing OVA in the pancreas [80, 83]. When examining the effects of oral OVA on OT-I mice, in this study cells isolated from mice receiving oral OVA did not show increased cytolytic activity when compared to cells from OT-I mice receiving PBS (Figure 3.5). These results are in agreement with the studies suggesting that oral antigens induce immune tolerance. Differences in oral delivery, T cell priming, and nature of the antigen may account for the contrasting results. In the process of T cell priming during the course of an infection, cells first become activated then produce and secrete cytokines such as the pro-inflammatory cytokine, IFN- γ. However, in cases of immune dysregulation including instances where intestinal homeostasis is disrupted, production of IFN-γ occurs in the absence of an invading pathogen. Specifically, in cases of IBD, IFN-γ production is increased in T cells in the intestine of patients [94]. Results within this thesis showed OT-I cells were activated in response to oral OVA (Section 3.2.3 and 3.25) and produced IFN-γ (Figure 3.4 and Figure 3.9). IFN-γ production induced by oral OVA could suggest that a possible unwanted immune response towards the oral antigen is being generated. However, the results presented here did not clarify whether the IFN-γ production is resulting in intestinal inflammation. Additionally, the generation of an IFN-γ-mediated immune response towards oral antigen that disrupts intestinal homeostasis is not consistent with the concept of oral tolerance. Recently, a new subset of CD8+ T cells with suppressive function and expressing surface latency associated peptide (LAP) were shown to produce 59 IFN-γ [20, 95]. These CD8+ LAP+ T cells with regulatory function have been described in the context of the experimental autoimmune encephalomyelitis (EAE) mouse model. The CD8+ LAP+ T cells were shown to have an IFN-γ dependent suppressive effect in the EAE model [95]. IFN- γ signaling was not required for the CD8+ LAP+ T cells to exhibit suppression nor did IFN- γ directly act on the responder cells. The mechanism by which IFN- γ production contributes to the suppressive effect of CD8+ LAP+ T cells still remains to be elucidated. Nevertheless, it can be speculated that the observed IFN-γ production by CD8+ T cells in response to OVA might have a regulatory function similar to the CD8+ LAP+ T cells with protective role described in context of the EAE model. Understanding the precise role of CD8+ T cells in intestine homeostasis and their response to oral antigens will help in finding treatments towards CD8+ T cell mediated gut inflammation and tissue destruction leading to IBD. 60 Chapter 4. CD8+ T Cell Responses Towards Oral Antigens in Cross Presentation Deficient Mouse Models 4.1 Rationale Cross presentation is a key process by which DCs can acquire, process and present exogenously-derived antigens to CD8+ T cells. As demonstrated in Chapter 3, DCs in the intestine acquire oral antigens and travel to the MLN to prime CD8+ T inducing cell proliferation and activation. Cross presentation is likely an important component of this process. The vacuolar model of cross presentation proposes MHC-I molecules are directed to an endocytic compartment to be loaded with exogneous derived peptides; although, the exact mechanism remains to be elucidated. In support of the vacuolar model the Jefferies lab has proposed two non-exclusive mechanisms by which MHC-I molecules are trafficked and directed to compartments containing exogenous acquired antigens. In the first pathway trafficking relies on a tyrosine motif in the cytoplasmic domain of MHC-I molecules [40, 62]. In the second pathway it is proposed chaperone protein CD74 directs MHC-I molecule from the ER to the proper endocytic compartment for acquisition of exogenously derived peptides [49, 50]. Ultimately cross presentation is essential for loading of exogenous derived peptides onto MHC-I molecules for induction of immunity or tolerance. The cytoplasmic region of MHC-I molecules contains important trafficking motifs necessary for ER export, endocytosis from the cell surface, recycling and targeting for degradation (reviewed in [46, 51]). It also plays a functionally significant role in determining what antigens will be sampled by dictating which cellular compartments the MHC-I molecules will access. Studies from the Jefferies lab have shown that a cytoplasmic tyrosine on MHC-I is responsible for directing surface MHC-I molecules to endolysosomal compartments [40, 62]. Transgenic mice expressing MHC-I (H-2Kb) containing a single substitution of phenylalanine for the conserved cytoplasmic tyrosine (ΔYKb) were shown to have impaired cytolytic T lymphocyte immunity against viruses and intracellular bacteria [40, 63]. In dendritic cells (DCs)-derived from the ΔYKbC3H mouse, few MHC-I molecules entered the endolysosomal compartments [40]. Furthermore, when the ΔYKbC3H DCs were incubated with soluble OVA, minimal H- 61 2Kb-OVA complexes were formed [40]. It was concluded that this conserved tyrosine motif is required for proper trafficking of MHC-I from the plasma membrane to the cross priming compartment to acquire exogenous antigens and induce CTL responses [40, 62, 63]. The ΔYKbC3H mouse has proven to be an important model to assess the contribution of cross presentation in various infectious settings. Chaperone protein, CD74, is known to associate with MHC-II in the ER and direct it through the secretory pathway using a cytoplasmic dileucine motif. Recently, an interaction between MHC-I and CD74 has also been demonstrated. CD74 was shown to direct MHC-I from the ER to an endolysosomal compartment, presumably the cross priming compartment [49, 50]. DCs-derived from CD74-deficient mice were shown to have reduced MHC-I loaded endolysosomal compartments and subsequently have impaired cross priming ability [49, 50]. Therefore, the CD74-deficient mouse was identified as an additional cross presentation deficient model. The cytoplasmic tyrosine on MHC-I and chaperone protein CD74, have both shown to be required for efficient cross presentation by DCs [40, 49, 50]. Therefore, the in vivo requirement of cytoplasmic tyrosine on MHC-I and chaperone protein CD74 for cross presentation of oral antigens was assessed in transgenic mouse strains, ΔYKb and CD74-/-. Analysis of CD8+ T cell responses towards oral antigens in the corresponding mouse models will provide insight on the role of cytoplasmic tyrosine on MHC-I and CD74 in intestinal homeostasis. 4.2 Results 4.2.1 ΔYKb breeding strategy The ΔYKbC3H mouse has been previously described to have an impaired ability to generate CD8+ T cell responses to viral, intracellular bacterial and cell-associated antigens [40, 62, 63]. In DCs of these mice, the MHC-I molecule (H-2Kb) molecule containing a single substitution of phenylalanine for the conserved cytoplasmic tyrosine (ΔYKb) was shown to be deficient in trafficking from the plasma membrane to the cross priming compartment for acquisition of exogenous antigens and cross presentation [40]. The ΔYKbC3H mouse and its control strain WTKbC3H (together referred to as KbTg +C3H), express a knocked in transgenic H-2Kb gene (KbTg) on a genetic background of a 62 C3H mouse. One caveat of this model is that the endogenous C3H-specific MHC class I genes (H-2Kk and H-2Dk) are expressed as well as the knocked-in KbTg, either the mutated version, ΔYKb, or the wild type version, WTKb. These endogenous MHC-I molecules are able to cross present exogenously-derived peptides for induction of CD8+ T cell immunity or tolerance. Therefore, the endogenously expressed H-2Kk and H-2Dk molecules in the ΔYKbC3H mouse strain may compensate for the KbTg molecule and impede assessment of its role in induction of CD8+ tolerance. For full assessment of the role of KbTg in the induction of CD8+ T cell tolerance, a mouse that solely expresses KbTg was generated. To this end, the KbTg+ C3H mice were bred onto a MHC-I deficient background as described in breeding schematic in Figure 4.1A. The MHC-I null mice have disrupted H- 2Kb and H-2Db genes (thus are double H-2KbDb knockout mice) resulting in a lack of MHC-I molecule expression on the cell surface [87]. Therefore, the H-2KbDb knockout mice, within this context, are referred to as H-2K-/-D-/-. The end result of these crosses is a mouse strain that only expresses the knocked in KbTg on an endogenously MHC-I null genetic background (red in Figure 4.1A). This unique mouse strain will be referred to as ΔYKb and WTKb (and collectively referred to as KbTg+) whereas the parental strain is referred to as ΔYKbC3H and WTKbC3H (collectively as KbTg+ C3H). All progeny resulting from the crosses were genotyped as described in Section 4.1.2. The initial crossing between KbTgC3H with H-2K-/-D-/- mice gave rise to the F1 generation that were heterozygous for respective genes (Figure 4.1B). There was 50% probability for the progeny to inherit the KbTg+. The progeny of interest from the initial crossing had the genotype of KbTg+ H-2Kk/-Dk/- (blue in Figure 4.1B), which were selected and then backcrossed with H-2K-/-D-/- mice (Figure 4.1C). In these crosses, the probability of obtaining progeny with the desired Kb Tg+ genotype on the MHC-I null background was 12.5%. The resulting Kb Tg+ mice were subsequently backcrossed to the H-2KbDb double knockout mouse strain for ten generations (Figure 4.1D). 63 64 Figure 4.1. ΔYKb and WTKb breeding strategy schematic The schematic of the breeding strategy to generate Kb Tg+ mice on the MHC-I null genetic background is shown. (A) The overall strategy to generate Kb Tg+ (shown in red) from the Kb Tg+C3H mice crossed with the H-2KbDb double knockout strain. Kb Tg+ represents the transgene ΔYKb or WTKb respectively. (B) A schematic of the initial cross of Kb Tg+C3H mice with MHC-I null mice (H-2KbDb double knockout) resulting in F1 heterozygous mice. Mice with genotype of interest that were selected for subsequent breeding are labeled in blue. (C) F1 hybrids positive (in blue) for the Kb transgene were back crossed with H-2KbDb double knockouts. This cross resulted in 12.5% probability of getting progeny with the desired Kb Tg+ genotype depicted in red. (D) Kb Tg+ genotype mice were back crossed to the H2KbDb doulbe knockouts 10 times, where there is a 50% probability of obtaining progeny positive for the Kb Tg+ genotype that was used for subsequent back crossing. 65 4.2.2 Genotyping and phenotyping of the Kb Tg+ mice Genotyping of the Kb Tg+ related progeny resulting from crosses described in Section 4.1.1 was first performed by PCR as described in Section 2.2. A 1.8kb and 1.6kb long PCR generated DNA fragments indicated the presence of the inserted DNA sequences at the H-2K and 2D locus, respectively (Figure 4.2B and Figure 4.2C). Therefore, the presence of 1.8kb and 1.6kb long fragments confirmed if the progeny had inherited disrupted H-2Kb or Db genes from the MHC-I null parental strain. This genotyping strategy however did not discriminate between progeny with the H-2Kk/-Dk/- and H-2K-/- and D-/- genotype. Additional PCR using primers specific for H-2Kk and Dk genes generated a 0.48kb fragment indicating the presence of wild type H-2Kk and Dk MHC-I genes (Figure 4.2A). Even though homologous H-2Kb or Db genes, as found in C57Bl/6 and H-2KbDb double knockout mice, have a high sequence similarity to H-2Kk and Dk genes, primers used were shown to be specific for H-2Kk and Dk genes (Figure 4.2D). Primers successfully generated PCR fragments exclusively from genomic DNA of C3H mice. Thus PCR using H-2Kk and Dk specific primers allowed the discrimination between animals with the genotype of H-2Kk/-and Dk/- from H-2K-/- and D-/- obtained from the crosses. 66 Figure 4.2. MHC-I genes for genotyping by PCR (A) Schematic representation of intron and exon structure of the gene encoding murine MHC-I. The boxes represent 8 exons. PCR generated 0.48Kb diagnostic fragment to assess presence of wild type H-2Kk and H-2Dk genes (B) Schematic representation of disrupted H2Kb gene. Second and third exons were replaced by the HPRT minigene; the 1.8-kb PCR amplified diagnostic fragment. (C) Schematic representation of disrupted H2Db gene. First three exons were replaced by the whole pGNA vector; the 1.6-kb PCR amplified diagnostic fragment. (D) Agarose gels of PCR generated H-2Kk and H-2Dk 0.48Kb long diagnostic fragments. PCR was performed on genomic DNA of C3H, C57Bl/6 and H-2KbDb double KO (KO) mice using H-2Kk specific (right) and H-2Dk specific (left) primers. 67 Peripheral blood leukocytes (PBLs) from the genotyped KbTg+ mice, lacking endogenous H-2Kk and Dk, were subsequently assessed for surface expression of specific MHC-I molecules by flow cytometry (Figure 4.3). The mouse strains C3H and C57Bl/6 (serving as controls), known to express H-2Kk, 2Dk and Kb respectively, were confirmed to express the respective MHC-I variants reflective of their respective strains. As described previously [40, 86], results showed that ΔYKbC3H mice expressed high levels of H-2Kb, whereas H-2Kk, Dk are expressed at reduced levels compared to C3H mice. The KbTg+ mice, both ΔYKb and WTKb genotypes, expressed high levels of H-2Kb and lacked of H-2Kk and Dk expression. Thus, confirming the desired genotype and phenotype of the KbTg+ mice. 68 Figure 4.3. KbTg+ mice express H-2Kb and lack H-2Kk and Dk surface expression. Flow cytometry analysis of MHC-I (H-2Kk, 2Dk and Kb) expression on PBLs from KbTg+ mice. Representative histograms show expression of MHC-I molecules from transgenic mice obtained from the crosses ΔYKb H2D-/- and control WTKb H2D-/- and control mice C3H, C57BL/6 and ΔYC3H. 69 4.2.3 Phenotype of CD74-/- mice The CD74 molecule (invariant chain) has recently been shown to play a role in DC cross presentation [50]. Therefore, CD74-/- mice were obtained to use as a second cross presentation defective mouse model to study the role of cross presentation in cross tolerance. First, the CD74-/- mice were phenotyped to confirm the deletion of the CD74 protein. The chaperone CD74 is known to associate with MHC-II molecules and affect surface expression of MHC-II on thymic epithelial cells that mediate T cell selection in the thymus [46]. Lower MHC-II expression on the cell surface of thymic epithelial cells impairs positive selection of CD4+ T cells and results in a skewed ratio towards CD8+ T cells [96]. Thus, analysis of peripheral CD4+ and CD8+ T cell populations may be used to confirm the deletion of the CD74 protein. In the following, CD4+ and CD8+ T cell populations from spleen and PBLs were analyzed to confirm the phenotype of CD74- deficient mice. Flow cytometry analysis of CD4 and CD8 expression on lymphocytes confirmed a skewed ratio towards CD8+ T cells (Figure 4.4A and 4.4B). CD74-/- mice contained an increased frequency of CD8+ T cells compared to wild type C57Bl/6 mice (>20% vs 15%). On the other hand, CD74-/- mice had less than 10% CD4+ T cells in the blood and spleen whereas C57Bl/6 mice had more than 17% CD4+ T cells. Overall results showed CD74-/- mice contained an increased frequency of CD8+ T cells and reduced frequency of CD4+ T cells when compared to C57Bl/6 mice. Given this skewed T cell ratio, CD74-/- mice were phenotyped by analysis of CD4+ and CD8+ T cell numbers in PBLs. Mice having less than 10% of CD4+ T cells were considered to be CD74-/- and used for further experiments. 70 Figure 4.4. Skewed CD8+ T to CD4+ T cell ratios in CD74-/- mice. (A,B) Representative flow cytometry dot plots of CD4+ and CD8+ T cells of CD74-/- mice as compared to wild type C57Bl/6 mice (WT). Surface expression of CD4 and CD8 in (A) PBLs and (B) splenocytes. The numbers in plots represent the percentage of CD4+ and CD8+ cell populations from the total number of lymphocytes. Mice containing less than 10% of CD4+ PBLs were considered to be CD74-/-. 71 4.2.4 Requirement of CD74 in bone marrow derived dendritic cells for cross presentation Studies from the Jefferies Lab have suggested that the chaperone protein CD74 binds to MHC-I and directs it from the ER to endolysosomal compartments for acquisition of exogenous derived peptides and cross presentation [50]. To confirm these results, the requirements of CD74 and CD74 cytoplasmic domain for cross presentation were assessed. To this end, immature DCs were derived from the bone marrow of either CD74-/- or C57Bl/6 mice. The DCs were collected on day 6 to obtain a >80% pure population of immature DCs (Figure 4.5A). The immature DCs expressed high levels of MHC-II, MHC-I and CD11c but low levels of costimulatory molecules CD40 and CD86. Immature CD74-/- bmDCs were then transfected with plasmids containing either full length CD74 (FL) or CD74 lacking the cytosolic trafficking domain (Δ2-17), incubated with OVA and analyzed for ability to induce CFSE labeled OT-I cell proliferation. CD74-/- DCs showed a decreased ability to induce OT-I proliferation compared to the wild type DCs (100% vs 11%; Figure 4.5) consistent with previous findings [50]. CD74-/- DCs reconstituted with FL CD74 induced similar OT-I proliferation levels as wild-type DCs (100% vs 80% Figure 4.6). On the other hand, when CD74-/- DCs were reconstituted with CD74 lacking the cytoplasmic domain cross priming ability was not restored (100% vs 18%). This demonstrated that the presence of CD74 contributed to MHC-I acquisition of exogenously-derived antigens. Specifically, the cytoplasmic domain of CD74 is required for efficient cross presentation. Overall, these results suggest that the sorting signals in the cytoplasmic domain of CD74 were necessary to direct bound MHC-I molecules to the compartment containing exogenous derived OVA peptides for cross presentation. 72 Figure 4.5. CD74 is required for efficient DC cross presentation. (A) Histograms show surface marker expression of immature DC phenotype collected on day 6 of culture. Dark line shows staining for specific surface molecules; filled grey graphs show unstained control. (B) Representative dot plots showing proliferation of CFSE labeled OT-I cells incubated with bmDCs. CD74-/- bmDCs were reconstituted with full length (FL) CD74 or truncated (Δ2-17) CD74 lacking the endolysosomal trafficking motif, incubated with soluble OVA and assessed for the ability to induce OT-I cell proliferation. Percentages represent the proportion of proliferating OT-I cells normalized to C57Bl/6 controls. (B) Reprinted by permission from Macmillan Publishers Ltd: [Nature Immunology] ([49]) copyright (2012). 73 4.2.5 The role of cross presentation in oral tolerance. The cytoplasmic tyrosine on MHC-I and chaperone protein CD74, have both shown to be required for efficient presentation by DCs [40, 50]. Therefore, the CD74-/- and ΔYKb mice mouse models were used to assess the requirement for cross presentation of oral antigens. Cross presentation efficiency was measured by analyzing induction of OT-I cell proliferation in ΔYKb and CD74-/- mice. To this end, CFSE labeled OT-I cells were adoptively transferred into the corresponding mouse strains and then OVA was administered by oral gavage. Next, isolated lymphocytes from MLN and spleen were analyzed by flow cytometry. As previously seen in Chapter 3, results show OT-I cell proliferation levels in wild type mice were higher in the MLN than in the spleen (89% vs 43%; Figure 4.6A). Administration of a high dose (40mg) of OVA induced comparable OT-I proliferation in MLN of ΔYKb and its wild type equivalent WTKb (87% vs 92% Figure 4.6A). However, upon administration of a low dose (4mg) of OVA by oral gavage, ΔYKb mice showed reduced OT-I cell proliferation in the MLN as compared to its wild type equivalent WTKb (63% vs 82%; p<0.05; Figure 4.6B). Similarly, administration of a low dose of OVA induced on average lower OT-I cell proliferation in CD74-/- mice compared to wild type C57Bl/6 mice (23% vs 38%; Figure 4.7A and Figure 4.7B). Despite a strong trend, differences were not statistically significant (p=0.15; n=3). Taken together, the tyrosine on H-2Kb is required for efficient cross presentation and induction of CD8+ T cell proliferation to orally administered antigen. 74 75 Figure 4.6. ΔYKb mice have impaired CD8+ T cell proliferation and activation after oral administration of OVA protein. (A,B) Proliferation of transferred OT-I T cells in MLN and spleen of ΔYKb and WTKb mice that received graded doses of OVA or PBS (control) by oral gavage 18 h after transfer. Proliferation of OT-I T cells was assessed 48 h after antigen administration by CFSE dilution. Gated on CFSE+H-2Kb/OVA tetramer+ CD8+ cells (A) Mice were administered 40mg of OVA after T cell transfer. Left, representative histograms of CFSE dilution are shown and numbers indicate percentage of cells that proliferated. Right, mean proliferation of OT-I cells. (B) Mice were administered 4mg of OVA after T cell transfer. Left, representative histograms of CFSE dilution are shown and numbers indicate percentage of cells that proliferated. Right, mean proliferation of OT-I cells. (C) Flow cytometry analysis of CD44 expression on donor OT-I cells from MLNs of mice receiving 4mg oral OVA. Left, representative dot plots of CD44 expression of total donor OT-I cells. Values in plots indicate the percentage of CD44 low and CD44 high cells among OT-I donor cells. Right, mean percentage of CD44 high, CD44 low cells of donor OT-I cells. Graphs in all panels show mean ± SDs (n=3). Statistical comparisons were performed using the Student’s t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001. 76 Figure 4.7. CD74-/- mice exhibit a trend towards reduced proliferation and activation of OT-I cells after oral administration of OVA protein. Proliferation and activation of donor OT-I cells after oral administration of OVA. (A,B)Proliferation of transferred OT-I T cells (gated on CFSE+ on H-2Kb/OVA tetramer+ CD8+ T) in MLN and spleen ΔYKb and wild type control WTKb mice that received 4mg OVA or PBS (control) by oral gavage 18 h after transfer. Proliferation of OT-I T cells was assessed 48 h after antigen administration by CFSE dilution. (A) Representative histograms of CFSE dilution are shown and numbers indicate percentage of cells that proliferated. (B) Mean proliferation of OT-I cells. (C) Flow cytometry analysis of CD44 expression of donor OT-I cells (gated on CFSE+ on H-2Kb/OVA tetramer+ CD8+ T cells) from MLNs. Values in plots indicate CD44 low and CD44 high percent cells. (D) Mean percentage of CD44 high and CD44 low cells of donor OT-I cells. Graphs in all panels show mean ± SDs (n=3). Statistical comparisons were performed using the Student’s t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001. 77 In Chapter 3, it was shown that oral OVA induced CD8+ T cell activation as measured by expression of activation markers on the cell surface. Therefore, cell surface expression of CD44 was used to measure activation efficiency of OT-I cells in ΔYKb and CD74-/- in response to oral OVA. As before, CFSE labeled OT-I cells were transferred to ΔYKb and CD74-/- mouse strains and analyzed by flow cytometry 48 hours following oral gavage with 4mg OVA. In ΔYKb mice, OT-I cells showed a lower frequency of activated cells as indicated by lower numbers of CD44hi cells compared to WTKb control mice (74% vs 87%; p<0.05; Figure 4.6C). Upon antigen presentation and before clonal expansion T cells upregulate surface CD44 expression [89]. Therefore lower frequency of activated cells observed in ΔYKb is consistent with lower OT-I proliferation in MLN of ΔYKb compared to WTKb (Figure 4.6B). The donor OT-I donor population in CD74-/- mice showed a trend towards a reduced frequency of CD44hi cells and increased frequency of CD44lo cells compared to the wild type C57Bl/6 control mice (Figure 4.7C). Although on average frequency of OT-I CD44hi cells in CD74-/- mice was 35% compared to a 45% in C57Bl/6 control mice (Figure 4.7D), differences were not statistically significant (p=0.2). Taken together, cross presentation of oral antigen plays a key part in activating the CD8+ T cell response described earlier. Furthermore, the pathway mediated by the cytoplasmic tyrosine of MHC I plays a larger role than the one dependent on CD74. 4.2.6 Cross presentation motifs found in cytoplasmic domains of MHC-I molecules are conserved in human MHC-I alleles As the cytoplasmic tail of MHC class I played an important role in cross presentation of oral antigens, it was important to assess the conservation of this motif in humans. Human MHC-I genes, the HLA genes, are grouped into six classes called HLA-A, -B, -C, -E, -F and –G. In the genomic sequences of HLA, exon 5,6,7 and 8 encode amino acids that form the cytoplasmic region (Figure 4.8) [97]. The cytoplasmic region contains trafficking motifs necessary for ER export, endocytosis from the cell surface, recycling and targeting for degradation (Table 1.2) [51]. Here, the cytoplasmic domains of HLA molecules were analyzed. The first amino acid sequence alignment allowed a comparison 78 of the cytoplasmic domains amongst the six HLA genes (Figure 4.8). A second set of amino acid alignments allowed for the examination of allelic variations in the transmembrane and cytoplasmic domain (Figure 4.9). The first amino acid sequence alignment comparing the cytoplasmic domain of HLA- A, -B, -C, -E, -F and –G showed that some of the important trafficking motifs are conserved across all the six HLA genes (Figure 4.9). In particular, HLA-A, -B, -E and -F –G contain a tyrosine based endocytosis motif, YXXØ, where X represents any amino acid and Ø represents any bulky, hydrophobic amino acid. The alignment reveals that HLA-C and HLA-G lack the tyrosine based YXXØ endocytosis motif; however, HLA-C contains a dileucine based endocytosis motif, DXSLI. Seeing that cytoplasmic tail of MHC-I molecules contains important trafficking motifs for cross presentation, the next step was to assess human MHC class I alleles for possible amino acid polymorphism in the cytoplasmic region, with a particular focus on the endocytosis motifs YXXØ and DXSLI. 79 Figure 4.8. Multiple sequence alignment of transmembrane and cytoplasmic domain of HLA molecules Alignment was done using ClustalX software and GenBank accession numbers AJ249241 (HLA-A), AJ420238 (HLA-B), AJ420242 (HLA-C), M20022 (HLA-E), X17093 (HLA-F), and X17273 (HLA-G). The colored amino acids represent conserved trafficking motifs (See Table 1.2). 80 Sequences of HLA alleles were obtained from sequence alignments available on the IMGT/HLA database, aligned on 14 January 2011 by Steven GE Marsh, Anthony Nolan Research Institute [98]. The IMGT/HLA Database is part of the international ImMunoGeneTics project (IMGT) and provides the official sequences of the human HLA alleles. Only alleles that contained the full amino acid sequence of the cytoplasmic region (Exon 5-8) were chosen for analysis. The database showed that all alleles for HLA-E, -F and –G have completely conserved cytoplasmic region and will therefore not be further discussed. For HLA-A, -B and –C, the database provided a total of 155, 246 and 145 alleles respectively (Figure 4.10). In HLA-A there were 12 different combinations of polymorphic amino acid residues in the domains of interest. HLA-B displayed fewer polymorphism because out of the 246 fully sequenced alleles available on the database there were only 6 different types of polymorphic combinations, one containing an insertion. A closer look at the different cytoplasmic regions of all the HLA-A and –B alleles showed that the endocytosis YXXØ motif was conserved in all of them. HLA-A alleles showed variation in the amino acid found next to the tyrosine, which was either threonine (T) or serine (S). However, a functional motif allows the amino acid following the tyrosine (Y) to be any amino acid. Overall the analysis showed that alleles for HLA-A and –B have a conserved YXXØ motif (Figure 4.2A and Figure 4.2B). HLA-C gene showed the highest number of polymorphic combinations at the cytoplasmic region. Of the 145 alleles available for the analysis there were 16 different variations of amino acid sequences in the cytoplasmic region. Three alleles of HLA-C contained an insertion that resulted in the addition of 6 amino acids. The allele HLA- C*04:09N, is designated with the N of Null allele because it is not expressed and as shown by the alignment contains a C-terminal proline instead of alanine. Nevertheless, the DXSLI endocytosis motif was found to be conserved in all the HLA-C alleles analyzed (Figure 4.2C). Moreover, two alleles, HLA-C*08:22 and C*15:29, contain a cysteine (C) substitution for a tyrosine (Y). This change in amino acids allows the formation of the YXXØ motif in these two HLA-C alleles. This means that alleles HLA- C*08:22 and C*15:29 contain two endocytosis motifs, YXXØ and DSXLI. Nevertheless, the DSXLI endocytosis motif itself is conserved in all HLA-C alleles. 81 Altogether HLA-A, -B and C alleles do not display polymorphism at the endocytosis motifs, YXXØ and DXSLI respectively. Conservation of endocytosis motifs on MHC-I molecules signifies the evolutionary importance of the intracellular trafficking motifs. 82 83 Figure 4.9. Polymorphism in transmembrane and cytoplasmic domains of HLA-A, - B and –C. (A,B and C) Sequence alignment of amino acids in exons 5 through 8 of HLA-A alleles. Alleles with sequenced exons 5 through 8 were chosen for polymorphism analysis in the cytoplasmic domains. One representative of each polymorphic combination is shown. Frequency was calculated as follows (number of times a polymorphic combination was present/ total alleles analyzed). A missing sequenced amino acid in a particular allele is represented with a “*”. An additional amino acid not found in the other alleles is represented with a “.”. The colored amino acids represent conserved trafficking motifs (See Table 1.2). Amino acids that varied amongst alleles are shown in grey boxes. (A) Amino acid variation in HLA-A alleles. 164 sequences of HLA-A alleles were available. One representative of each variation is shown. (B) Amino acid variation in HLA-B alleles. For HLA-B, 242 sequences were used for the analysis and one representative of each combination is shown. (C) Amino acid variation in HLA-C alleles. For HLA-C, 146 sequences were used for the analysis and one representative of each combination is shown. 84 4.3 Discussion 4.3.1 Endocytosis motifs on MHC-I cytoplasmic tails of HLA alleles The exact mechanisms and pathways that direct MHC-I intracellular trafficking and cross presentation still remain to be elucidated. Nevertheless, the cytoplasmic domain of MHC-I molecules contains important signals for proper trafficking and cross presentation. The Jefferies Lab has proposed two non-mutually exclusive pathways by which MHC-I molecules are trafficked to compartments for acquisition of exogenous derived peptides. One pathway involves the requirement of a cytoplasmic tyrosine-based internalization motif on MHC-I. The motif is critical for endocytosis and directing of MHC-I molecules from the plasma membrane to the appropriate intracellular compartment [40, 62, 63]. The second pathway requires CD74, which binds to a fraction of MHC-I molecules in the ER and directs them to the endolysosomal compartment for loading of exogenously-derived antigens [49, 50]. The cytoplasmic domain of CD74, which contains two leucine based sorting motifs [43], was shown in this study to be required for cross presentation, presumably by directing bound MHC-I to endolysosomal compartments. Mutations in trafficking motifs or deficiency of CD74 could lead to aberrant trafficking of MHC-I and ultimately affect antigen presentation. Hence, this chapter investigated the importance of these MHC-I trafficking mechanisms in cross presentation of oral antigens. Previously in this thesis, it was shown that a high dose of orally administered OVA induced OT-I cell proliferation and CD44 surface expression. The latter indicating that CD8+ T cells are being activated and proliferate after encountering the antigen. Other studies indicate that CD8+ T cell activation and proliferation observed in this study is due to DCs residing in the intestine taking up the antigen and travelling to the MLN for cross presentation to T cells [71, 73, 78, 80]. Specifically, the CD8α- CD11b+ DCs were found to be the main DC subset in the MLN to cross present oral antigens and induce CD8+ T cell proliferation [71]. This indicates that cross presentation of oral antigens by DCs is necessary for inducing CD8+ T cell activation and proliferation. The MHC-I cytoplasmic tyrosine and the cytoplasmic domain of chaperone protein CD74 have been shown to 85 play a role in the pathway of cross presentation [49, 50, 86]. Therefore, two mouse models, ΔYKb and CD74-/-, were used to study the in vivo contribution of CD74 and the cytoplasmic tyrosine on MHC-I to cross presentation of oral antigens and subsequent activation of CD8+ T cells in the MLN. Assessment of transferred OT-I CFSE-labeled cells revealed that a low dose of OVA resulted in more than 10% reduction of OT-I cell proliferation and activation in ΔYKb mice compared to the WTKb control. The tyrosine on H-2Kb is necessary for efficient cross presentation and induction of CD8+ T cell responses to orally administered antigen. Previous studies suggest the nature and the abundance of the antigens found in the intestine influence the mechanism of the cross presentation pathway [99]. For example, oral infection of ΔYKb C3H mice with a recombinant A (Listeria- OVA) resulted in reduced cross presentation [63]. The authors suggested DCs were processing apoptotic bodies of infected cells in endolysosomal compartments and the ΔYKb molecules could not be trafficked to the compartments containing the pathogen derived peptides resulting in deficient cross presentation [63]. Contrary to Listeria-OVA, orally administered OVA protein, normally perceived as a harmless food antigens, could trigger different uptake mechanisms and cross presentation pathways. OVA protein has been shown to be processed through a variety of routes [99]. Processing of exogenous OVA for cross presentation is dependent on the proteasome, the TAP transporter, endosomal trafficking, recycling of cell surface molecules and lysosomal acidification [99, 100]. In contrast, OVA coupled to virus-like particles are processed in lysosomal compartments, where OVA-derived peptides are loaded onto MHC-I molecules recycled from the cell surface [99, 100]. Unlike soluble OVA protein, processing of OVA coupled to virus-like particles did not require the proteasome or the TAP transporter. This means the nature of the antigen must dictate the processing pathway for cross presentation. Taking all this into account, results presented in this study suggest a high abundance of harmless OVA in the intestine may result in a high uptake of OVA by DCs, which could be mainly translocated to the cytosol to be processed predominately via the protoesome and TAP dependent pathways. Therefore, ΔYKb and WTKb molecules would have equal access to OVA- derived peptides and oral OVA induced similar levels of OT-I T cell proliferation and activation in ΔYKb and WTKb mouse strains. On the other hand at lower doses, the 86 majority of OVA might be processed in lysosomal compartments via a TAP and proteasome independent pathway. A fraction of MHC-I is trafficked from the plasma membrane to OVA containing endolysosomal compartments and interrupting this trafficking route could therefore result in reduced cross presentation. During uptake and processing of low doses of OVA, few ΔYKb molecules would have access to OVA peptides in the endolysosomal compartments as shown by a 19% reduction in CD8+ T cell proliferation. Although not statistically significat, CD74-/- mice had a trend towards a decreased ability to activate CD8+ T cells to proliferate in response to a low dose of OVA compared to the control mice. This could indicate that DCs in the intestine of CD74-/- mice have reduced ability to acquire, process and cross present OVA to induce CD8+ T cell proliferation. This trend is supported by previous studies showing the requirement of CD74 for cross presentation and the induction of immunity [49, 50]. Nevertheless, the requirements for CD74-mediated cross presentation have yet to be defined and there may be varying requirements for CD74 mediated cross presentation depending on the particular DC subset. Cross presentation efficiency towards exogenously-derived peptides has shown to differ amongst DC subsets [34, 46, 71]. Moreover, DCs derived from the intestine display differential surface markers, cytokine secretion patterns and promote higher differentiation of Tregs when compared to DCs derived from the spleen [26, 35, 101]. Previous studies exploring the role of CD74 in cross presentation were performed using bone marrow-derived DCs as well as spleen-derived DCs [49, 50]. On the other hand, oral OVA is processed by intestinal DC subsets which acquire the antigen and travel to the MLN for cross presentation to CD8+ T cells. Therefore, these differences in dependence on CD74-mediated cross presentation might be due to differences between DCs subsets found in the intestine and spleen. Mutations in the tyrosine trafficking motifs were shown to affect antigen presentation of oral antigens. Therefore, assessment of human MHC-I alleles for possible polymorphism in the cytoplasmic domain with special focus of the endocytosis motif, a requirement for cross presentation, was carried out. The tyrosine based YXXØ internalization motif, where X represents any amino acid and Ø represents any bulky and hydrophobic amino acid, is found in various transmembrane proteins like MHC-I [40] 87 and well characterized transferrin receptor [102]. This tyrosine based YXXØ motif is involved in endocytosis from the plasma membrane [61]. Substituting the tyrosine residue with a cysteine in the YXXØ motif of the human transferrin receptor decreased the rates of endocytosis resulting in a 60% inhibition in iron accumulation [60]. Similarly, replacing tyrosine for phenylalanine in the YXXØ motif of human and murine MHC-I molecules resulted in impaired endocytosis and trafficking from the cell surface to endocytic compartments and ultimately cross presentation of exogenous derived peptides [40, 55, 62]. The cytoplasmic tyrosine of the YXXØ motif on MHC-I is conserved across species [62] and is found on the majority of HLA molecules. In all the corresponding HLA alleles analyzed in this study, it was further revealed that the YXXØ motif is conserved reflecting the evolutionary importance of the motif. Unlike HLA-A and –B, HLA-C lacks the cytoplasmic tyrosine in the YXXØ motif but instead relies on a dileucine motif, DSXLI, for endocytosis [56]. The dileucine motif consists of two adjacent cytoplasmic leucine or isoleucine residues and requires a serine phosphorylation to initiate internalization and eventual translocation to the lysosomal compartments [56]. Analysis of HLA-C alleles showed that the DSXLI endocytosis motif was also conserved. It was still surprising to observe that for two HLA-C alleles (HLA- C*08:22 and C*15:29) a cysteine (C) was replaced with a tyrosine (Y), resulting in the formation of a tyrosine based YXXØ motif in these two alleles. Therefore, HLA-C*08:22 and C*15:29 alleles, were shown to contain two endocytosis motifs, YXXØ and DSXLI. The tyrosine of the YXXØ motif on murine (H-2Kb) or human (HLA-B27) has been shown to be important for directing molecules from the cell surface to endocytic compartments [40, 55, 62]. In particular, trafficking of MHC-I molecules to these endocytic compartments containing extracelluar derived antigens is crucial for cross presentation. It would be intriguing to determine how intracellular trafficking of HLA-C molecules, which lack the cytoplasmic tyrosine in the YXXØ motif, compares to MHC-I molecules HLA-A and –B which contain YXXØ motif. Additionally it would be interesting to investigate if HLA-C alleles HLA-C*08:22 and C*15:29, which contain both endocytosis motifs, have altered intracellular trafficking patterns. Another change in the cytoplasmic domain of HLA-C was seen in allele HLA- C*04:09N. This allele was shown to contain a C-terminal proline instead of alanine. This 88 allele was previously reported not to be expressed at the cell surface and is therefore designated with the N of Null allele [103]. The C-terminal alanine in HLA-C is known to facilitate ER export [58] and it is possible that in HLA-C*04:09N, the substitution of the C-terminal amino acid alanine for a proline could affect ER export and consequently cell surface expression. Overall analysis of HLA alleles showed endocytosis motifs YXXØ and DSXLI were conserved in HLA-A, -B and –C respectively. Endocytosis is a key mechanism that regulates surface expression and conservation of endocytosis motifs such as YXXØ and DSXLI throughout evolution is of critical importance for maintenance of HLA function. Cross presentation pathways for induction of immunity and tolerance is an active field of research with many unanswered questions. This study explored the requirement of CD74 and the MHC-I cytoplasmic tyrosine for cross presentation of orally acquired antigens. The presented results suggest that the processing of antigens found in the intestine for efficient cross presentation depends on several factors including the nature of the antigen, the amount of antigen available and the specific cross presentation mechanism available to the responding DC subset. Further studies of cross presentation on specific DC subsets and using additional antigen models, will shed light on the exact mechanisms and responses that are induced by antigens acquired orally. 89 Chapter 5. Conclusions and Future Directions 5.1 General conclusions 5.1.1 CD8+ T cell responses towards oral antigen The results presented in Chapter 3 of this thesis describe the phenotype of CD8+T cells in response to oral antigens. It is known that antigens delivered via the oral route induce a type of antigen specific tolerance referred to as oral tolerance. This means that challenging the immune system with a foreign antigen does not elicit an immune response if the foreign antigen was given by the oral route before the challenge. The mechanism that leads to an antigen specific immune suppression is not clearly defined. Different immune cells are thought to be involved in maintaining intestinal homeostasis and preventing unwanted immune responses towards harmless oral antigens. However, it is not clear how CD8+ T cells preserve unresponsiveness to oral antigens. The current studies attempted to better understand how CD8+ T cell respond to oral antigens. The use of the transgenic mouse strain OT-I, allowed the analysis of CD8+ T cells responses after oral delivery of the OVA. OT-I mice express a transgenic TCR that recognizes an OVA peptide in context of MHC-I H-2Kb. It was shown that CD8+ T cells encountered the oral antigen and were activated, as shown by changes of surface expression of T cell activation markers. Moreover, CD8+ T cells proliferated and secreted higher amounts of the inflammatory cytokine, IFN-γ. Consistent with the concept of unresponsiveness to oral antigens, even though CD8+ T cells showed activation in response to oral OVA these cells did not display increased cytolytic function when compared to the control. 90 5.1.2 Cross presentation Oral antigens need to be taken up and cross presented by APCs in the intestine in order to induce CD8+ T cell activation, proliferation and cytokine production. This role has been specifically attributed to the DCs residing in the intestine that take up the antigen and travel to the MLN and cross present it to T cells [71, 73, 78, 80]. It has been previously reported that the protein chaperone CD74 and the tyrosine based endocytosis motif YXXØ on MHC-I play important roles in the cross presentation pathway [40, 50, 86]. The cytoplasmic tyrosine of the YXXØ motif on MHC-I molecules is important for endocytosis. The tyrosine is needed for directing MHC-I molecules from the plasma membrane to endolysosomal compartments for loading of exogenous derived peptides [40, 55]. The critical role of this endocytosis motif is reflected by the conservation of the motif across species [40] and all corresponding human MHC-I alleles, as shown in this study. The cytoplasmic domain of CD74 was also shown here to be important for cross presentation, suggesting a role in directing MHC-I molecules to the proper intracellular compartment. The in vivo requirement for inducing CD8+ T cell proliferation and activation towards oral antigen by the endocytosis motif on MHC-I and the chaperon protein CD74 was investigated. Using two mouse models, ΔYKb and CD74-/-, MHC-I cross priming was measured in response to oral OVA. CD8+ T cell responses towards a low concentration of antigen concentration were statistically reduced in ΔYKb mouse strain. However, CD74-/- mice trended towards decreased ability (although not statistically significant) to generate CD8+ T cell proliferation and activation towards oral antigen. The mechanistic details of CD74 mediated cross presentation have just started to be explored. There exist the possibility of varying requirements for CD74 mediated cross presentation depending on the DC subset. The results presented here suggest that DCs subsets found in the intestine processing harmless oral antigens require cytoplasmic tyrosine on MHC-I for cross presentation and activation of CD8+ T cells. Conversely, there is the possibility for the chaperone protein CD74 to be involved in cross presentation of oral OVA in order to induce activation and proliferation of CD8+ T cells. 91 In summary, oral antigens are taken up and presented by intestinal DCs to T cells in the MLN lymph nodes. It was shown here that cross presentation of oral antigens to CD8+ T cells induced activation, proliferation and cytokine production. The cytoplasmic tyrosine on MHC-I was required to cross present oral antigen for efficient CD8+ T cell activation and proliferation. These results shed new light on the phenotype of CD8+ T cells that is induced in response to oral antigens. However, it is still unknown how these CD8+ T cells remain unresponsive to the oral antigen for proper maintenance of intestinal homeostasis. A more detailed assessment of CD8+ T cell activity might reveal how these cells avoid inducing an unwanted immune response towards harmless oral antigens and thereby contribute to intestinal homeostasis and oral tolerance. 5.2 Future directions 5.2.1 CD8+ T cell responses towards oral antigen It is important for CD8+ T cells to be unresponsive to oral antigens to avoid the induction of an unwanted immune response in the intestine towards harmless oral antigens. In this study it was shown that upon presentation to an oral antigen CD8+ T cells are activated, proliferate and produce cytokines. What still remains to be elucidated is the exact mechanism(s) and pathways that maintain this unresponsive state and avoid inducing cell mediated immunity towards the oral antigens. Using a mouse model in which OVA is expressed by enterocytes in the intestine, transferred OT-I cells proliferated at first and were deleted weeks after the transfer [93]. This suggested that endogenously expressed OVA induced clonal deletion of the donor OT-I cells from the peripheral lymphocyte pool. To determine if clonal deletion is also induced in response towards OVA acquired by the oral route OT-I cells can be transferred to C57Bl/6 mice prior administration of oral OVA. The percentage of remaining transferred cells can be examined by flow cytometry at different time points. Decreasing numbers of transferred OT-I cells with time could indicate if cells are being deleted after oral delivery of OVA. Instead of being clonally deleted, an alternative option would be that CD8+T cells exhibit regulatory activity in response to oral OVA. A newly identified CD8+ T cell subset was shown to have suppressive activity dependent on IFN- γ production[95]. 92 These cells are identified by surface expression of latency-associated peptide (LAP). We previously showed that CD8+ T cells were activated and produced IFN-γ in response to oral OVA. Therefore, CD8+ T cells responding and proliferating in response to oral antigen could be assessed for LAP surface expression by flow cytometry. This would further indicate if CD8+ T cells have similarities with the previous newly identified regulatory CD8+ T cell subset. However, to have a better insight on function, T cell suppressive activity could bee assessed for suppressive activity regardless of LAP expression. T cell suppressive activity is measured by ability to suppress T effector cell proliferation. A possible regulatory activity of CD8+ T cells responding to oral OVA can be assessed by their ability to suppress T effector cell proliferation using a T cell suppression assay. These additional experiments could reveal the mechanism by which CD8+ T cells maintain unresponsiveness towards oral antigens and avoid mounting an unwanted immune response. 5.2.2 Cross presentation The tyrosine on the endocytosis motif YXXØ on MHC-I is one of several proposed molecular mechanisms for explaining cross presentation of exogenous derived antigens. It is interesting to observe that the tyrosine-based endocytosis motif YXXØ is found in 4 out of 6 human MHC-I molecules HLA-A, -B, -E and –F. HLA-C contains a di-leucine based endocytosis motif. This suggests that the HLA-C cytoplasmic domain may interact with different adaptor proteins than the rest of the HLA molecules and ultimately be targeted to different intracellular compartments. It would be interesting to compare intracellular trafficking between HLA-A, -B, -E and –F with HLA–C to establish specific role of each endocytosis motif, YXXØ and DXSLI, in directing molecules to specific intracellular compartments. Analysis of the endocytosis motifs on the HLA alleles showed that two HLA-C alleles, HLA-C*08:22 and C*15:29, have a cytoplasmic cysteine (C) substitution with tyrosine (Y) resulting in the creation of a tyrosine based YXXØ motif in the two alleles. It would be of interest to determine if HLA-C molecules with both endocytosis motifs, YXXØ and DSXLI, have altered intracellular trafficking patterns and if so how they compare and contrast to HLA-A, -B, -E and –F trafficking. Confocal microscopy could be used to determine intracellular the localization, trafficking and distribution of different HLA molecules. By staining with for endolysosomes and late 93 endosomes markers microscopy analysis could show to which intracellular compartment each HLA molecule colocalizes. Additionally, HLA molecules could be assessed for colocalization to intracellular compartments with internalized fluorescent OVA. This would indicate which HLA molecule has access to the compartment containing exogenously derived antigens. Examination and comparison of HLA-A, -B, -E and –F with HLA–C trafficking would provide better insight on the significance of the endocytosis motifs in dictating to which cellular compartments will be accessed by the different HLA molecules. It was previously shown that the lack of the cytoplasmic tyrosine on murine MHC-I resulted in reduced OT-I proliferation in response to oral antigen in ΔYKb mice. However it was not confirmed that this reduction in OT-I proliferation was due to decreased cross presentation efficiency by DCs and not any other cell type found in the intestine. The latter could be confirmed by isolating DCs found in the MLN and intestine of ΔYKb mouse strain and assessing for cross presentation of OVA in vitro. Moreover, DCs from OVA fed ΔYKb mice could be isolated and stained for presence of MHC-I/OVA complexes and analyzed by flow cytometry. It is expected that isolated DCs from the intestine and MLN from ΔYKb mice would have reduced levels of MHC-I/OVA complexes on the cell surface and show reduced cross presentation in vitro. These experiments would assure cross presentation was impaired due to reduced cross presentation in ΔYKb DCs. CD74 mediated cross presentation is another pathway proposed by which MHC-I molecules acquire exogenous derived peptides. Previous studies on CD74 deficient mice have shown that bmDCs as well as splenic DCs require CD74 for cross presentation and suggest a CD74 mediated cross presentation pathway to be taking place [50]. Cross presentation of orally derived antigens has been shown to be mediated by intestinal DC subsets [26, 71]. It was shown in this thesis that CD74 deficient mice had a trend towards lower cross presentation efficiency of orally administered antigens when compared to their wild type counterpart. However, the experiments preformed to assess cross presentation of orally derived antigen in CD74 deficient mice had some complications and have to be repeated. By doing so it will be confirmed if CD74 deficient mice have indeed lower ability to cross present oral antigens for generating CD8+ T cell activation 94 and proliferation. Nevertheless, CD74 mediated cross presentation might differ amongst DC subsets. Splenic, bone marrow derived and intestinal DCs from CD74-/- mice could be used to assessed for OVA cross presentation in vitro. Thereby, the CD74 requirement for cross presentation efficiency could be compared amongst DCs found in different anatomical locations. Moreover DCs from the spleen and the intestine could be sorted based on surface expression markers and then examined for cross presentation in vitro. These assays would help to further understand the requirement for CD74 mediated cross presentation in different DC subsets based on location and surface markers. 95 References  1.   Mowat,  A.M.,  Anatomical  basis  of  tolerance  and  immunity  to  intestinal   antigens.  Nat  Rev  Immunol,  2003.  3(4):  p.  331-­‐41.  2.   Hooper,  L.V.  and  A.J.  Macpherson,  Immune  adaptations  that  maintain   homeostasis  with  the  intestinal  microbiota.  Nat  Rev  Immunol,  2010.  10(3):  p.  159-­‐169.  3.   Izcue,  A.,  J.L.  Coombes,  and  F.  Powrie,  Regulatory  lymphocytes  and  intestinal   inflammation.  Annu  Rev  Immunol,  2009.  27:  p.  313-­‐38.  4.   Abreu,  M.T.,  Toll-­‐like  receptor  signalling  in  the  intestinal  epithelium:  how   bacterial  recognition  shapes  intestinal  function  (vol  10,  pg  131,  2010).  Nature  Reviews  Immunology,  2010.  10(3).  5.   McDole,  J.R.,  et  al.,  Goblet  cells  deliver  luminal  antigen  to  CD103+  dendritic   cells  in  the  small  intestine.  Nature,  2012.  483(7389):  p.  345-­‐349.  6.   Rescigno,  M.,  et  al.,  Dendritic  cells  express  tight  junction  proteins  and  penetrate   gut  epithelial  monolayers  to  sample  bacteria.  Nature  Immunology,  2001.  2(4):  p.  361-­‐7.  7.   Macpherson,  A.J.  and  K.  Smith,  Mesenteric  lymph  nodes  at  the  center  of   immune  anatomy.  J  Exp  Med,  2006.  203(3):  p.  497-­‐500.  8.   Westendorf,  A.M.,  et  al.,  T  cells,  dendritic  cells  and  epithelial  cells  in  intestinal   homeostasis.  Int  J  Med  Microbiol,  2010.  300(1):  p.  11-­‐8.  9.   van  Wijk,  F.  and  H.  Cheroutre,  Intestinal  T  cells:  facing  the  mucosal  immune   dilemma  with  synergy  and  diversity.  Semin  Immunol,  2009.  21(3):  p.  130-­‐8.  10.   Bridgeman,  J.S.,  et  al.,  Structural  and  biophysical  determinants  of  αβ  T-­‐cell   antigen  recognition.  Immunology,  2012.  135(1):  p.  9-­‐18.  11.   Kabelitz,  D.,  γδ  T-­‐cells:  cross-­‐talk  between  innate  and  adaptive  immunity.  Cellular  and  Molecular  Life  Sciences,  2011.  68(14):  p.  2331-­‐2333.  12.   Obar,  J.J.  and  L.  Lefrancois,  Early  events  governing  memory  CD8+  T-­‐cell   differentiation.  Int  Immunol,  2010.  22(8):  p.  619-­‐25.  13.   Maloy,  K.J.  and  F.  Powrie,  Intestinal  homeostasis  and  its  breakdown  in   inflammatory  bowel  disease.  Nature,  2011.  474(7351):  p.  298-­‐306.  14.   Campbell,  D.J.  and  M.A.  Koch,  Phenotypical  and  functional  specialization  of   FOXP3+  regulatory  T  cells.  Nat  Rev  Immunol,  2011.  11(2):  p.  119-­‐130.  15.   Ostanin,  D.V.,  et  al.,  T  cell  transfer  model  of  chronic  colitis:  concepts,   considerations,  and  tricks  of  the  trade.  Am  J  Physiol  Gastrointest  Liver  Physiol,  2009.  296(2):  p.  G135-­‐46.  16.   Nancey,  S.p.,  et  al.,  CD8+  Cytotoxic  T  Cells  Induce  Relapsing  Colitis  in  Normal   Mice.  Gastroenterology,  2006.  131(2):  p.  485-­‐496.  17.   Westendorf,  A.M.,  et  al.,  Autoimmune-­‐mediated  intestinal  inflammation-­‐ impact  and  regulation  of  antigen-­‐specific  CD8+  T  cells.  Gastroenterology,  2006.  131(2):  p.  510-­‐24.   96 18.   Westendorf,  A.M.,  et  al.,  Intestinal  epithelial  antigen  induces  CD4+  T  cells  with   regulatory  phenotype  in  a  transgenic  autoimmune  mouse  model.  Ann  N  Y  Acad  Sci,  2006.  1072:  p.  401-­‐6.  19.   Billerbeck,  E.  and  R.  Thimme,  CD8+  regulatory  T  cells  in  persistent  human  viral   infections.  Hum  Immunol,  2008.  69(11):  p.  771-­‐5.  20.   Tang,  X.L.,  T.R.  Smith,  and  V.  Kumar,  Specific  control  of  immunity  by  regulatory   CD8  T  cells.  Cell  Mol  Immunol,  2005.  2(1):  p.  11-­‐9.  21.   Fleissner,  D.,  et  al.,  Local  induction  of  immunosuppressive  CD8  T  cells  in  the   gut-­‐associated  lymphoid  tissues.  PLoS  One,  2010.  5(10):  p.  e15373.  22.   Menager-­‐Marcq,  I.,  et  al.,  CD8+CD28-­‐  regulatory  T  lymphocytes  prevent   experimental  inflammatory  bowel  disease  in  mice.  Gastroenterology,  2006.   131(6):  p.  1775-­‐85.  23.   Chaput,  N.,  et  al.,  Identification  of  CD8+CD25+Foxp3+  suppressive  T  cells  in   colorectal  cancer  tissue.  Gut,  2009.  58(4):  p.  520-­‐529.  24.   Xavier,  R.J.  and  D.K.  Podolsky,  Unravelling  the  pathogenesis  of  inflammatory   bowel  disease.  Nature,  2007.  448(7152):  p.  427-­‐34.  25.   Coombes,  J.L.  and  F.  Powrie,  Dendritic  cells  in  intestinal  immune  regulation.  Nat  Rev  Immunol,  2008.  8(6):  p.  435-­‐46.  26.   Scott,  C.L.,  A.M.  Aumeunier,  and  A.M.  Mowat,  Intestinal  CD103+  dendritic  cells:   master  regulators  of  tolerance?  Trends  Immunol,  2011.  32(9):  p.  412-­‐9.  27.   Shortman,  K.  and  S.H.  Naik,  Steady-­‐state  and  inflammatory  dendritic-­‐cell   development.  Nat  Rev  Immunol,  2007.  7(1):  p.  19-­‐30.  28.   Morelli,  A.E.  and  A.W.  Thomson,  Tolerogenic  dendritic  cells  and  the  quest  for   transplant  tolerance.  Nature  Reviews  Immunology,  2007.  7(8):  p.  610-­‐621.  29.   Kulkarni,  A.B.,  et  al.,  Transforming  growth  factor-­‐beta  1  null  mice.  An  animal   model  for  inflammatory  disorders.  Am  J  Pathol,  1995.  146(1):  p.  264-­‐75.  30.   Rennick,  D.,  N.  Davidson,  and  D.  Berg,  Interleukin-­‐10  gene  knock-­‐out  mice:  a   model  of  chronic  inflammation.  Clin  Immunol  Immunopathol,  1995.  76(3  Pt  2):  p.  S174-­‐8.  31.   Rimoldi,  M.,  et  al.,  Intestinal  immune  homeostasis  is  regulated  by  the  crosstalk   between  epithelial  cells  and  dendritic  cells.  Nature  Immunology,  2005.  6(5):  p.  507-­‐514.  32.   Zaph,  C.,  et  al.,  Epithelial-­‐cell-­‐intrinsic  IKK-­‐beta  expression  regulates  intestinal   immune  homeostasis.  Nature,  2007.  446(7135):  p.  552-­‐6.  33.   Hart,  A.L.,  et  al.,  Characteristics  of  Intestinal  Dendritic  Cells  in  Inflammatory   Bowel  Diseases.  Gastroenterology,  2005.  129(1):  p.  50-­‐65.  34.   Coombes,  J.L.,  et  al.,  A  functionally  specialized  population  of  mucosal  CD103+   DCs  induces  Foxp3+  regulatory  T  cells  via  a  TGF-­‐beta  and  retinoic  acid-­‐ dependent  mechanism.  J  Exp  Med,  2007.  204(8):  p.  1757-­‐64.  35.   Sun,  C.M.,  et  al.,  Small  intestine  lamina  propria  dendritic  cells  promote  de  novo   generation  of  Foxp3  T  reg  cells  via  retinoic  acid.  J  Exp  Med,  2007.  204(8):  p.  1775-­‐85.  36.   Yamazaki,  S.,  et  al.,  Dendritic  cells  are  specialized  accessory  cells  along  with   TGF-­‐  for  the  differentiation  of  Foxp3+  CD4+  regulatory  T  cells  from  peripheral   Foxp3  precursors.  Blood,  2007.  110(13):  p.  4293-­‐302.   97 37.   Matteoli,  G.,  et  al.,  Gut  CD103(+)  dendritic  cells  express  indoleamine  2,3-­‐ dioxygenase  which  influences  T  regulatory/T  effector  cell  balance  and  oral   tolerance  induction.  Gut,  2010.  59(5):  p.  595-­‐604.  38.   Grakoui,  A.,  et  al.,  The  immunological  synapse:  a  molecular  machine   controlling  T  cell  activation.  Science,  1999.  285(5425):  p.  221-­‐7.  39.   Neefjes,  J.,  et  al.,  Towards  a  systems  understanding  of  MHC  class  I  and  MHC   class  II  antigen  presentation.  Nat  Rev  Immunol,  2011.  11(12):  p.  823-­‐836.  40.   Lizee,  G.,  et  al.,  Control  of  dendritic  cell  cross-­‐presentation  by  the  major   histocompatibility  complex  class  I  cytoplasmic  domain.  Nature  Immunology,  2003.  4(11):  p.  1065-­‐73.  41.   Lefranc,  M.P.,  IMGT,  the  international  ImMunoGeneTics  database.  Nucleic  Acids  Res,  2001.  29(1):  p.  207-­‐9.  42.   Lizee,  G.,  G.  Basha,  and  W.A.  Jefferies,  Tails  of  wonder:  endocytic-­‐sorting  motifs   key  for  exogenous  antigen  presentation.  Trends  Immunol,  2005.  26(3):  p.  141-­‐9.  43.   Landsverk,  O.J.,  O.  Bakke,  and  T.F.  Gregers,  MHC  II  and  the  endocytic  pathway:   regulation  by  invariant  chain.  Scand  J  Immunol,  2009.  70(3):  p.  184-­‐93.  44.   Hsieh,  C.S.,  et  al.,  A  role  for  cathepsin  L  and  cathepsin  S  in  peptide  generation   for  MHC  class  II  presentation.  J  Immunol,  2002.  168(6):  p.  2618-­‐25.  45.   Villadangos,  J.A.a.P.S.,  Intrinsic  and  cooperative  antigen-­‐presenting  functions  of   dendritic-­‐cell  subsets  in  vivo.  Nat  Rev  Immunol,  2007.  7(7):  p.  543-­‐  55.  46.   Lin,  M.L.,  et  al.,  The  cell  biology  of  cross-­‐presentation  and  the  role  of  dendritic   cell  subsets.  Immunol  Cell  Biol,  2008.  86(4):  p.  353-­‐62.  47.   Touret,  N.,  et  al.,  Quantitative  and  Dynamic  Assessment  of  the  Contribution  of   the  ER  to  Phagosome  Formation.  Cell,  2005.  123(1):  p.  157-­‐170.  48.   Duan,  F.  and  P.K.  Srivastava,  An  invariant  road  to  cross-­‐presentation.  Nat  Immunol,  2012.  13(3):  p.  207-­‐208.  49.   Basha,  G.,  et  al.,  A  CD74-­‐dependent  MHC  class  I  endolysosomal  cross-­‐ presentation  pathway.  Nat  Immunol,  2012.  13(3):  p.  237-­‐245.  50.   Omilusik,  K.D.,  The  requirement  for  competent  antigen  presenting  dendritic   cells  and  poised  T  cells  for  immune  responses,  in  Microbiology  and   Immunology2011,  University  of  British  Columbia:  Vancouver.  p.  289.  51.   Donaldson,  J.G.  and  D.B.  Williams,  Intracellular  assembly  and  trafficking  of   MHC  class  I  molecules.  Traffic,  2009.  10(12):  p.  1745-­‐52.  52.   Coscoy,  L.  and  D.  Ganem,  Kaposi's  sarcoma-­‐associated  herpesvirus  encodes  two   proteins  that  block  cell  surface  display  of  MHC  class  I  chains  by  enhancing  their   endocytosis.  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America,  2000.  97(14):  p.  8051-­‐8056.  53.   Bartee,  E.,  et  al.,  Downregulation  of  major  histocompatibility  complex  class  I  by   human  ubiquitin  ligases  related  to  viral  immune  evasion  proteins.  J  Virol,  2004.   78(3):  p.  1109-­‐20.  54.   Boyle,  L.H.,  et  al.,  Selective  export  of  HLA-­‐F  by  its  cytoplasmic  tail.  J  Immunol,  2006.  176(11):  p.  6464-­‐72.  55.   Santos,  S.G.,  et  al.,  Lack  of  tyrosine  320  impairs  spontaneous  endocytosis  and   enhances  release  of  HLA-­‐B27  molecules.  J  Immunol,  2006.  176(5):  p.  2942-­‐9.   98 56.   Schaefer,  M.R.,  et  al.,  A  novel  trafficking  signal  within  the  HLA-­‐C  cytoplasmic   tail  allows  regulated  expression  upon  differentiation  of  macrophages.  J  Immunol,  2008.  180(12):  p.  7804-­‐17.  57.   Lynch,  S.,  et  al.,  Novel  MHC  class  I  structures  on  exosomes.  J  Immunol,  2009.   183(3):  p.  1884-­‐91.  58.   Cho,  S.,  et  al.,  Receptor-­‐Mediated  ER  Export  of  Human  MHC  Class  I  Molecules  Is   Regulated  by  the  C-­‐Terminal  Single  Amino  Acid.  Traffic,  2011.  12(1):  p.  42-­‐55.  59.   Belizaire,  R.  and  E.R.  Unanue,  Targeting  proteins  to  distinct  subcellular   compartments  reveals  unique  requirements  for  MHC  class  I  and  II  presentation.  Proc  Natl  Acad  Sci  U  S  A,  2009.  106(41):  p.  17463-­‐8.  60.   Ponka,  P.  and  C.N.  Lok,  The  transferrin  receptor:  role  in  health  and  disease.  International  Journal  of  Biochemistry  &  Cell  Biology,  1999.  31(10):  p.  1111-­‐1137.  61.   Traub,  L.M.,  Tickets  to  ride:  selecting  cargo  for  clathrin-­‐regulated   internalization.  Nat  Rev  Mol  Cell  Biol,  2009.  10(9):  p.  583-­‐96.  62.   Basha,  G.,  et  al.,  MHC  Class  I  Endosomal  and  Lysosomal  Trafficking  Coincides   with  Exogenous  Antigen  Loading  in  Dendritic  Cells.  Plos  One,  2008.  3(9).  63.   Reinicke,  A.T.,  et  al.,  Dendritic  cell  cross-­‐priming  is  essential  for  immune   responses  to  Listeria  monocytogenes.  PLoS  One,  2009.  4(10):  p.  e7210.  64.   Kruger,  T.,  et  al.,  Lessons  to  be  learned  from  primary  renal  cell  carcinomas:   novel  tumor  antigens  and  HLA  ligands  for  immunotherapy.  Cancer  Immunology  Immunotherapy,  2005.  54(9):  p.  826-­‐836.  65.   Bluestone,  J.A.,  Mechanisms  of  tolerance.  Immunol  Rev,  2011.  241(1):  p.  5-­‐19.  66.   Probst,  H.C.,  et  al.,  Resting  dendritic  cells  induce  peripheral  CD8(+)  T  cell   tolerance  through  PD-­‐1  and  CTLA-­‐4.  Nature  Immunology,  2005.  6(3):  p.  280-­‐286.  67.   Jin,  H.T.,  R.  Ahmed,  and  T.  Okazaki,  Role  of  PD-­‐1  in  regulating  T-­‐cell  immunity.  Curr  Top  Microbiol  Immunol,  2011.  350:  p.  17-­‐37.  68.   Wing,  K.,  T.  Yamaguchi,  and  S.  Sakaguchi,  Cell-­‐autonomous  and  -­‐non-­‐ autonomous  roles  of  CTLA-­‐4  in  immune  regulation.  Trends  Immunol,  2011.   32(9):  p.  428-­‐33.  69.   Luckashenak,  N.,  et  al.,  Constitutive  crosspresentation  of  tissue  antigens  by   dendritic  cells  controls  CD8(+)  T  cell  tolerance  in  vivo.  Immunity,  2008.  28(4):  p.  521-­‐532.  70.   Schildknecht,  A.,  et  al.,  Antigens  expressed  by  myelinating  glia  cells  induce   peripheral  cross-­‐tolerance  of  endogenous  CD8+  T  cells.  European  Journal  of  Immunology,  2009.  39(6):  p.  1505-­‐1515.  71.   Chung,  Y.,  et  al.,  A  CD8  alpha-­‐11b+  dendritic  cells  but  not  CD8  alpha+  dendritic   cells  mediate  cross-­‐tolerance  to  intestinal  antigen.  Faseb  Journal,  2005.  19(4):  p.  A949-­‐A949.  72.   Kurts,  C.,  et  al.,  Cross-­‐presentation  of  self  antigens  to  CD8(+)  T  cells:  the   balance  between  tolerance  and  autoimmunity.  Immunological  Tolerance,  1998.  215:  p.  172-­‐181.  73.   Weiner,  H.L.,  et  al.,  Oral  tolerance.  Immunol  Rev,  2011.  241(1):  p.  241-­‐59.  74.   Faria,  A.M.  and  H.L.  Weiner,  Oral  tolerance.  Immunol  Rev,  2005.  206:  p.  232-­‐59.   99 75.   Chen,  Y.,  J.  Inobe,  and  H.L.  Weiner,  Induction  of  oral  tolerance  to  myelin  basic   protein  in  CD8-­‐depleted  mice:  both  CD4+  and  CD8+  cells  mediate  active   suppression.  J  Immunol,  1995.  155(2):  p.  910-­‐6.  76.   Higgins,  P.J.  and  H.L.  Weiner,  Suppression  of  experimental  autoimmune   encephalomyelitis  by  oral  administration  of  myelin  basic  protein  and  its   fragments.  J  Immunol,  1988.  140(2):  p.  440-­‐5.  77.   Strobel,  S.  and  A.M.  Mowat,  Oral  tolerance  and  allergic  responses  to  food   proteins.  Current  Opinion  in  Allergy  and  Clinical  Immunology,  2006.  6(3):  p.  207-­‐213  10.1097/01.all.0000225162.98391.81.  78.   Worbs,  T.,  et  al.,  Oral  tolerance  originates  in  the  intestinal  immune  system  and   relies  on  antigen  carriage  by  dendritic  cells.  J  Exp  Med,  2006.  203(3):  p.  519-­‐27.  79.   Matteoli,  G.,  et  al.,  Gut  CD103+  dendritic  cells  express  indoleamine  2,3-­‐ dioxygenase  which  influences  T  regulatory/T  effector  cell  balance  and  oral   tolerance  induction.  Gut,  2010.  59(5):  p.  595-­‐604.  80.   Blanas,  E.,  et  al.,  A  bone  marrow-­‐derived  APC  in  the  gut-­‐associated  lymphoid   tissue  captures  oral  antigens  and  presents  them  to  both  CD4(+)  and  CD8(+)  T   cells.  Journal  of  Immunology,  2000.  164(6):  p.  2890-­‐2896.  81.   Zhang,  X.,  et  al.,  Activation  of  CD25(+)CD4(+)  regulatory  T  cells  by  oral  antigen   administration.  J  Immunol,  2001.  167(8):  p.  4245-­‐53.  82.   Blanas,  E.,  et  al.,  Induction  of  Autoimmune  Diabetes  by  Oral  Administration  of   Autoantigen.  Science,  1996.  274(5293):  p.  1707-­‐1709.  83.   Hanninen,  A.,  et  al.,  Mucosal  antigen  primes  diabetogenic  cytotoxic  T-­‐ lymphocytes  regardless  of  dose  or  delivery  route.  Diabetes,  2001.  50(4):  p.  771-­‐5.  84.   Ke,  Y.  and  J.A.  Kapp,  Oral  antigen  inhibits  priming  of  CD8+  CTL,  CD4+  T  cells,   and  antibody  responses  while  activating  CD8+  suppressor  T  cells.  J  Immunol,  1996.  156(3):  p.  916-­‐21.  85.   Ehirchiou,  D.,  et  al.,  CD11b  facilitates  the  development  of  peripheral  tolerance   by  suppressing  Th17  differentiation.  J  Exp  Med,  2007.  204(7):  p.  1519-­‐24.  86.   Lizee,  G.,  MHC  class  I  cytoplasmic  domain:  defining  a  role  for  conserved  amino   acid  residues  in  class  I  expression,  trafficking,  and  antigen  presentation,  in   Department  of  Zoology  and  the  Biotechnology  Laboratory2000,  UNIVERSITY  OF  BRITISH  COLUMBIA:  Vancouver.  p.  263.  87.   Perarnau,  B.,  et  al.,  Single  H2K(b),  H2D(b)  and  double  H2K(b)D(b)  knockout   mice:  peripheral  CD8(+)  T  cell  repertoire  and  anti-­‐lymphocytic   choriomeningitis  virus  cytolytic  responses.  European  Journal  of  Immunology,  1999.  29(4):  p.  1243-­‐1252.  88.   Nishimura,  T.,  et  al.,  IL-­‐6-­‐dependent  spontaneous  proliferation  is  required  for   the  induction  of  colitogenic  IL-­‐17-­‐producing  CD8(+)  T  cells.  Journal  of  Experimental  Medicine,  2008.  205(5):  p.  1019-­‐1027.  89.   Mitchell,  D.M.  and  M.A.  Williams,  An  activation  marker  finds  a  function.  Immunity,  2010.  32(1):  p.  9-­‐11.  90.   Carter,  L.L.,  et  al.,  PD-­‐1  :  PD-­‐L  inhibitory  pathway  affects  both  CD4(+)and   CD8(+)  T  cells  and  is  overcome  by  IL-­‐2.  European  Journal  of  Immunology,  2002.  32(3):  p.  634-­‐643.   100 91.   Ten  Hove,  T.,  et  al.,  Expression  of  CD45RB  functionally  distinguishes  intestinal   T  lymphocytes  in  inflammatory  bowel  disease.  J  Leukoc  Biol,  2004.  75(6):  p.  1010-­‐5.  92.   Clarke,  S.R.,  et  al.,  Characterization  of  the  ovalbumin-­‐specific  TCR  transgenic   line  OT-­‐I:  MHC  elements  for  positive  and  negative  selection.  Immunol  Cell  Biol,  2000.  78(2):  p.  110-­‐7.  93.   Lee,  J.W.,  et  al.,  Peripheral  antigen  display  by  lymph  node  stroma  promotes  T   cell  tolerance  to  intestinal  self.  Nat  Immunol,  2007.  8(2):  p.  181-­‐90.  94.   Cho,  J.H.,  The  genetics  and  immunopathogenesis  of  inflammatory  bowel   disease.  Nat  Rev  Immunol,  2008.  8(6):  p.  458-­‐66.  95.   Chen,  M.L.,  et  al.,  Novel  CD8+  Treg  suppress  EAE  by  TGF-­‐beta-­‐  and  IFN-­‐gamma-­‐ dependent  mechanisms.  Eur  J  Immunol,  2009.  39(12):  p.  3423-­‐35.  96.   Viville,  S.,  et  al.,  Mice  lacking  the  MHC  class  II-­‐associated  invariant  chain.  Cell,  1993.  72(4):  p.  635-­‐48.  97.   Lefranc,  M.P.,  et  al.,  IMGT,  the  international  ImMunoGeneTics  information   system.  Nucleic  Acids  Res,  2009.  37(Database  issue):  p.  D1006-­‐12.  98.   Robinson,  J.,  et  al.,  The  IMGT/HLA  database.  Nucleic  Acids  Res,  2011.   39(Database  issue):  p.  D1171-­‐6.  99.   Win,  S.J.,  et  al.,  Cross-­‐presentation  of  epitopes  on  virus-­‐like  particles  via  the   MHC  I  receptor  recycling  pathway.  Immunol  Cell  Biol,  2011.  89(6):  p.  681-­‐8.  100.   Burgdorf,  S.,  et  al.,  Spatial  and  mechanistic  separation  of  cross-­‐presentation   and  endogenous  antigen  presentation.  Nat  Immunol,  2008.  9(5):  p.  558-­‐66.  101.   Fink,  L.N.  and  H.  Frøkiær,  Dendritic  Cells  from  Peyer’s  Patches  and  Mesenteric   Lymph  Nodes  Differ  from  Spleen  Dendritic  Cells  in  their  Response  to   Commensal  Gut  Bacteria.  Scandinavian  Journal  of  Immunology,  2008.  68(3):  p.  270-­‐279.  102.   Collawn,  J.F.,  et  al.,  YTRF  is  the  conserved  internalization  signal  of  the   transferrin  receptor,  and  a  second  YTRF  signal  at  position  31-­‐34  enhances   endocytosis.  J  Biol  Chem,  1993.  268(29):  p.  21686-­‐92.  103.   Balas,  A.,  et  al.,  Elongation  of  the  cytoplasmic  domain,  due  to  a  point  deletion  at   exon  7,  results  in  an  HLA-­‐C  null  allele,  Cw*0409  N.  Tissue  Antigens,  2002.   59(2):  p.  95-­‐100.