Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of regulatory T cells induced by toxic shock syndrome toxin-1 and their potential application… Li, Haowei 2006

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

Item Metadata

Download

Media
831-ubc_2007-267509.pdf [ 12.97MB ]
Metadata
JSON: 831-1.0100485.json
JSON-LD: 831-1.0100485-ld.json
RDF/XML (Pretty): 831-1.0100485-rdf.xml
RDF/JSON: 831-1.0100485-rdf.json
Turtle: 831-1.0100485-turtle.txt
N-Triples: 831-1.0100485-rdf-ntriples.txt
Original Record: 831-1.0100485-source.json
Full Text
831-1.0100485-fulltext.txt
Citation
831-1.0100485.ris

Full Text

C H A R A C T E R I Z A T I O N S H O C K  S Y N D R O M E  O F R E G U L A T O R Y T C E L L S I N D U C E D  T O X I N - 1 A N D T H E I R  A C U T E  P O T E N T I A L  G R A F T - V E R S U S - H O S T  B Y  D I S E A S E  by  H A O W E I  B . S c ,  A  T H E S I S T H E  JiNan  S U B M I T T E D  L I  I N P A R T I A L  R E Q U I R E M E N T S  D O C T O R  1997  University,  F U L L F I L L M E N T O F  F O R T H E D E G R E E  O F  O F  P H I L O S O P H Y  in  T H E  F A C L U L T I E S  O F G R A D U A T E  (Experimental  T H E  U N I V E R S I T Y  M a r c h  S T U D I E S  Medicine)  O F B R I T I S H  2007  © H a o w e i L i , 2007  T O X I C  A P P L I C A T I O N  C O L U M B I A  I N  Abstract Repeated administration o f a bacterial superantigen, toxic shock syndrome toxin-1 (TSST-1), has been shown to induce C D 4 + regulatory T cells in mice. However, it is not clear i f the characteristics o f T S S T - l - i n d u c e d Tregs differ from those o f Tregs induced by other bacterial superantigens,  such as staphylococcal  enterotoxin A ( S E A ) . The mechanisms o f T S S T - l - i n d u c e d Tregs are also not well characterized. Because in experimental settings Tregs can be used to effectively treat multiple diseases, the potential application o f T S S T - l - i n d u c e d Tregs also needs to be determined. In this study, a side-by-side comparative study o f Tregs induced by TSST-1 and S E A was conducted. Results showed that T S S T - l - i n d u c e d Tregs were different from those induced by S E A in suppressive function, cytokine profile and proliferative ability. Remarkably, T S S T - l - i n d u c e d Tregs were more potent than SEA-induced Tregs in suppressive activities and proliferative ability in vitro. The possible mechanism o f T S S T - l - i n d u c e d Tregs was then investigated. T S S T - l - i n d u c e d Tregs did not induce death o f target cells, inhibit the activation o f target cells, or cause their target cells to acquire regulatory functions.  Supernatants  from T S S T - l - i n d u c e d Tregs were not suppressive and blockade o f IL-10 by a monoclonal antibody did not reverse the suppression. In contrast, cell contact with target cells was required. In addition, T S S T - l - i n d u c e d Tregs were able to compete with their target cells for IL-2. Finally, the potential therapeutic application o f T S S T - l - i n d u c e d Tregs was examined by determining their ability to control acute graft-versus-host  disease  ii  ( a G V H D ) i n a murine model. Data showed that TS ST-1-induced Tregs were able to mediate bystander suppression o f a G V H D following re-activation upon administration of TSST-1 post transplant. The Tregs inhibited the production o f proinflammatory cytokines triggered by alloresponses, but neither affected the engraftment or expansion of donor T cells nor enhanced the elimination o f host A P C s . Blockade o f IL-10 by in vivo administration o f a monoclonal antibody did not reverse the suppression. Finally, although TSST-1-induced Tregs attenuated a G V H D , graft-versus-tumor ( G V T ) effects were preserved, suggesting that these cells differentially regulate a G V H D versus G V T effects.  iii  Table of contents Abstract  ii  Table o f contents  iv  List o f tables  viii  List o f figures  ix  List o f abbreviations  xi  Acknowledgements  xiii  Chapter 1. Introduction  1  1.1. Regulatory T cells (Tregs) 1.1.1. B r i e f history, definition and categories o f Tregs 1.1.2. Phenotypes o f Tregs 1.1.2.1 .Activation-induced proliferation and apoptosis 1.1.2.2. C e l l markers 1.1.2.3. Cytokine profiles 1.1.3. Generation o f Tregs 1.1.3.1. Role o f the thymus 1.1.3.2. Role o f antigens 1.1.3.3. Role o f cytokines 1.1.3.4. Role o f costimulation 1.1.3.5. Role o f other cell types.. 1.1.4. Mechanisms o f immunomodulation b y Tregs 1.1.4.1. Targets o f Tregs 1.1.4.2. Role o f cytokines 1.1.4.3. Cytotoxic mechanisms 1.1.4.4.Inhibition o f activation by Tregs 1.1.4.5.Infectious tolerance 1.1.4.6. Cytokine competition 1.1.4.7.In vivo actions 1.1.5. Role o f Tregs i n disease pathogenesis and potential applications 1.1.5.1. Autoimmune diseases 1.1.5.2. Transplantation 1.1.5.3. Infections and malignant tumors 1.2. Superantigens 1.2.1. Definition and categories o f superantigens 1.2.2. Biological effects o f superantigens 1.2.3. Staphylococcal superantigens 1.2.3.1 .TSST-1  1 :  1 4 4 8 10 12 12 13 15 19 20 22 22 24 26 28 29 31 32 33 34 36 44 46 46 47 48 49  iv  1.2.3.2.SEA and other staphylococcal enterotoxins 1.2.4. Role o f bacterial superantigens in diseases 1.3. Connections o f bacterial superantigens and Tregs  50 51 54  Chapter 2. Hypothesis and specific aims  57  Chapter 3. Methods and materials  59  3.1. M i c e , TSST-1 preparation and TSST-1 treatment  59  3.2. C e l l isolation 3.3. C e l l transfer 3.4. In vitro culture 3.5. C F S E staining and proliferation assay 3.6. Assay o f cell death 3.7. Assay o f cytokine production 3.8. Lethal shock induction 3.9. Antibodies and flow cytometry 3.10. V(3 analysis 3.11. Reisolation o f naive splenocytes co-cultured with T S S T - 1 induced Tregs and secondary co-culture 3.12. Culture o f naive splenocytes with TSST-1-induced Tregs conditioned medium 3.13. Culture o f TSST-1-primed splenocytes with naive splenocyte conditioned medium 3.14. Neutralization o f IL-10 3.15. Tumor cells 3.16. M i x e d lymphocyte reaction 3.17. a G V H D induction and monitoring 3.18. Assay o f endotoxin 3.19. Histological studies 3.20. Statistical analysis -.  59 60 60 61 62 62 62 63 64 66  Chapter 4. Generation and characterization o f TSST-1 and S E A induced Tregs  66 67 67 68 68 69 69 70 .70 71  4.1. Introduction 71 4.2. Experimental design 73 4.3. Results 74 4.3.1. Generation o f superantigen-induced Tregs i n C 5 7 B L / 6 mice ... 74 4.3.2. Cytokine production profile o f TSST-1 and S E A 76 induced Tregs 4.3.3. Proliferation profile o f TSST-1 and SEA-induced-Tregs 81 4.3.4. Other phenotypic properties o f TSST-1 and S E A 84 induced Tregs 4.3.5. TSST-1-induced Tregs had broader suppressive 89 activities than SEA-induced Tregs  v  4.3.6. VP usage by TSST-1 and SEA-induced Tregs 93 -Cross-reactive suppressive function o f TSST-1-induced Tregs was not be due to shared VP families o f T cells reactive to both TSST-1 and S E A 4.4. Discussion 98 Chapter 5. Possible Mechanism o f suppressive properties o f TSST-1 induced Tregs  108  5.1. Introduction 108 5.2. Experimental design 109 5.3. Results Ill 5.3.1. TSST-1-induced Tregs did not enhance cell death Ill o f the target cells 5.3.2. TSST-1-induced Tregs enhanced the activation o f Ill target cells 5.3.3. TSST-1-induced Tregs did not render their target 115 cells to acquire regulatory functions 5.3.4. Supernatant o f stimulated TSST-1-induced Tregs 117 did not mediate suppressive activities 5.3.5. Suppressive activities o f TSST-1-induced Tregs 117 were cell contact-dependent 5.3.6. Blockade o f IL-10 with an anti-IL-lOR monoclonal 120 antibody did not reverse the suppressive functions o f T S S T - l - i n d u c e d Tregs 5.3.7. Intracellular cytokines expression by naive splenocytes 122 was not suppressed in the presence o f T S S T - l - i n d u c e d Tregs 5.3.8. T S S T - l - i n d u c e d Tregs competed for IL-2 with 126 naive splenocytes 5.3.9. Addition o f exogenous IL-2 did not reverse the 131 suppression o f T S S T - l - i n d u c e d Tregs 5.4. Discussion 133 Chapter 6. Potential application o f T S S T - l - i n d u c e d Tregs i n the control acutegraft-vs-host disease ( a G V H D )  138  6.1. Introduction 138 6.2. Experimental design 138 6.3. Results 140 6.3.1. Development and optimization o f an a G V H D animal model.. 140 6.3.2. Activation o f T S S T - l - i n d u c e d Tregs led to suppression 144 of a G V H D 6.3.3. T S S T - l - i n d u c e d C D 4 + Tregs mediated suppression o f 148 aGVHD 6.3.4. Activation o f T S S T - l - i n d u c e d Tregs did not affect 150 donor T cell expansion or enhance host antigen presenting cell  vi  elimination in vivo 6.3.5. Activation o f TSST-1-induced Tregs modulated cytokine production and serum endotoxin level 6.3.6. IL-10 blockade with IL-10 receptor monoclonal antibody did not reverse the suppression by TSST-1-induced Tregs 6.3.7. Control o f a G V H D by activation o f T S S T - 1 induced Tregs did not block graft-vs-tumor ( G V T ) response 6.4. Discussion Chapter 7. General discussion, conclusions, and future directions 7.1. General discussion 7.2. Conclusions 7.3. Future directions Bibliography  152 157 159 163 171 171 174 175 179  List of tables Table 4.1. Expression o f Foxp3, C T L A - 4 , G I T R or C D 2 5 on C D 4 +  88  T cells o f mice treated with T S S T - 1 , S E A or P B S Table 4.2. VP analysis o f superantigen-induced Tregs Table 6.1. Summary o f cause o f death i n each group o f mice  96 162  vni  List of figures Figure 1.1. Summary o f Tregs categories  5  Figure 1.2. Diagram o f binding o f superantigens to their ligands  52  Figure 4.1. Experimental design o f studies described in Chapter 4  73  Figure 4.2. Initial studies to determine the suppressive activities of superantigen-induced Tregs in vitro  75  Figure 4.3. Cytokine levels in supernatant o f TSST-1 or S E A -  77  induced Tregs following reactivation with TSST-1 or S E A Figure 4.4. Expression o f cytokines by TSST-1 or SEA-induced Tregs  78  Figure 4.5. Proliferation response o f superantigen-induced Tregs  82  Figure 4.6. C e l l death o f superantigen-induced Tregs in vitro  85  Figure 4.7. Phenotype o f superantigen-induced Tregs  88  Figure 4.8. Suppressive function o f superantigen-induced Tregs  90  in vitro and in vivo Figure 4.9. VP analysis o f superantigen-induced Tregs  95  Figure 4.10. Assay o f V p l 5 expression by R T - P C R  97  Figure 5.1. Experimental design o f studies described in Chapter 5  110  Figure 5.2. Effect o f T S S T - l - i n d u c e d Tregs on the cell death o f target cells Figure 5.3. Effect o f T S S T - l - i n d u c e d Tregs on the activation o f target cells  112  Figure 5.4. T S S T - l - i n d u c e d Tregs did not render the target cells to acquire regulatory functions  116  Figure 5.5. The role o f soluble factors in the suppressive function  118  113  o f T S S T - l - i n d u c e d Tregs Figure 5.6. C e l l contact-dependency o f T S S T - l - i n d u c e d Tregs  119  Figure 5.7. Effect o f addition o f anti-IL-10 receptor antibody  121  ix  on the suppressive function o f T S S T - l - i n d u c e d Tregs Figure 5.8. Expression o f IL-2 by naive splenocytes in the presence  123  of T S S T - l - i n d u c e d Tregs Figure 5.9. Competition o f IL-2 by T S S T - l - i n d u c e d Tregs  127  Figure 5.10. Effect o f anti-CD25 antibody on the uptake o f IL-2 by T S S T - l - i n d u c e d Tregs  129  Figure 5.11. Expression o f C D 2 5 on T S S T - l - i n d u c e d Tregs i n the co-culture with naive splenocytes  130  Figure 5.12. Effect o f addition o f exogenous IL-2 on the  132  suppression mediated by T S S T - l - i n d u c e d Tregs Figure 6.1. Experimental design o f studies described i n Chapter 6  139  Figure 6.2. a G V H D i n B 6—•unconditioned B 6 D 2 F 1 mouse model  142  Figure 6.3. a G V H D in B6—»cyclophosphamide-conditioned  143  B 6 D 2 F 1 mouse model Figure 6.4. Histology o f target organs o f mice undergoing a G V H D  143  Figure 6.5. Activation o f TSST-l-induced Tregs attenuated a G V H D  146  Figure 6.6. Activation o f TSST-l-induced C D 4 + Tregs mediated suppression o f a G V H D Figure 6.7. Effect o f activation o f T S S T - l - i n d u c e d Tregs on donor T cells and host A P C s  149  Figure 6.8. Effect o f activated TSST-l-induced Tregs on cytokine production in vivo and in vitro and serum endotoxin levels  154  Figure 6.9. Role o f IL-10 in the suppression mediated by  158  151  T S S T - l - i n d u c e d Tregs on a G V H D Figure 6.10. Effect o f activated TSST-l-induced Tregs on G V T response... 160 Figure 6.11. Histological study o f tumor infiltration o f the liver undergoing tumor challenge  161  x  List of abbreviations aGVHD  Acute graft-vs-host disease  AHSCT  Allogeneic hematopoietic stem cell transplantation  APCs  Antigen presenting cells  BMT  Bone marrow transplantation  CD40L  C D 4 0 ligand  CFSE  Carboxy fluoroscein succinimidyl ester  CTLA-4  Cytotoxic T lymphocyte antigen 4  EAE  Experimental autoimmune encephalomyelitis  ELISA  Enzyme linked immunosorbant assay  FBS  Fetal bovine serum  FCM  F l o w cytometry  FITC  Fluorescein isothiocyanate  G-CSF  Granulocyte colony stimulation factor  GITR  Glucocorticoid-induced tumor necrosis factor receptor family-related gene  GVT  Graft-vs-tumor  HSCT  Hematopoietic stem cell transplantation  IFN-y  Interferon-gamma  IL-2  Interleukin 2  IL-4  Interleukin 4  IL-10  Interleukin 10  MHC  Major histocompatibility complex  NOD  Non-obese diabetic  OVA  Ovalbumin  PBMC  Peripheral blood mononuclear cells  PBS  Phosphate buffered saline  PE  Phycoerythrin  PTSAgs  Pyrogenic toxin superantigens  RT-PCR  Reverse transcription polymase chain reaction  SCLD  Severe combined inrmunodeficient  SEA  Staphylococcal enterotoxin A  SEB  Staphylococcal enterotoxin B  XI  sc  Splenocytes  TGF-P  Transforming growth factor-beta  TNF-a  Tumor Necrosis Factor-alpha  Tregs  Regulatory T cells  T r l cells  T regulatory cells type 1  TSS  Toxic shock syndrome  TSST-1  Toxic shock syndrome toxin-1  7-AAD  7-aminoactinomycin D  Acknowledgements I would first like to express m y foremost gratitude to m y supervisor, Dr. Anthony W . Chow, for his guidance and support during this three-year doctoral training which I am sure is the cornerstone for m y career and life in North America. I would also like to thank all m y supervisory committee members, Dr. Paul K e o w n , Dr. K i r k Schultz and Dr. Megan Levings, for kindly attending m y committee meetings despite their busy schedules and providing me extensive guidance and advice throughout this doctoral program. It is also a must to thank all the members o f our laboratory who have always selflessly given me a hand whenever I needed. I would also acknowledge Dr. L i s a X u for her excellent instructions and help with the flow cytometry analysis. A n d finally, I wish to sincerely thank m y family for their stalwart love and support for m y decision to seek a career in science. They are always the source o f m y courage in face o f difficulties.  xiii  Chapter 1. Introduction 1.1.  Regulatory T cells The main role o f the immune system o f the mammal is to protect the body from  myriads o f potentially harmful microbial pathogens. To this end, the immune system generates millions o f T and B cells that recognize these microbial antigens by randomly arranging the genes o f receptors that recognize these antigens. However, this random generation o f T and B cells w i l l inevitably result i n T and B cell clones that also recognize the antigens expressed by cells around the body. Therefore, the immune system has also to develop means to ensure the discrimination between the self and non-self antigens, inhibiting autoimmune responses while allowing effective immune responses against the non-self antigens. Over the past century, extensive research on immune tolerance has revealed several mechanisms by which the immune system establishes  and maintains unresponsiveness  to self antigens,  including physical  elimination (clonal deletion) or functional inactivation o f self-reactive lymphocytes (clonal anergy), and regulation by subsets o f T cells, also called dominant tolerance (Fehr and Sykes, 2004; Kurtz et al., 2004b). The later has been one o f the foci i n immunology research in the past decade.  1.1.1. B r i e f history, definition and categories o f regulatory T cells Gershon et al first discovered the regulation mediated by T cells about thirty years ago (Cantor, 2004). Following their findings, studies by many groups in different systems also pointed to the presence o f subsets o f T cells having the ability to suppress immune responses (Damle and Engleman, 1983; Jiang et al., 1992; K u m a r and Sercarz,  1  1993; Q i n et al., 1993; Taylor et al., 1983; Tsumfuji et al., 1983). Despite the evident experimental findings, the concept o f regulation by T cells was still met with skepticism in the 80's mainly because at that time no specific marker for the T cells with regulatory functions has been found, making these cells rather elusive (Moller G , 1988). In 1995, after more than a decade's effort, Sakaguchi et al identified a subset o f C D 4 + T cells i n the periphery expressing C D 2 5 as a subset o f naturally occurring regulatory T cells (Tregs, also called natural Tregs). Removal o f this subset o f Tregs by thymectomy 3 days after the mice were born induces autoimmune diseases i n multiple organs which can be prevented by transfer o f C D 4 + C D 2 5 + Tregs (Sakaguchi et al., 1995). This paper thus solved one o f the mysteries in this area. M o r e importantly, the discovery o f a subset o f Tregs with a specific marker finally established  the  phenomenon o f T cell-mediated immune regulation. Following these seminal studies, many groups also focused on the existence o f antigen-induced Tregs. In contrast to the naturally occurring Tregs which are generated during the development o f T cells i n the thymus, it was found that Tregs can also be generated after antigen encounter i n the periphery (Vigouroux et al., 2004). These Tregs were characterized by their ability to control immune responses i n an antigenspecific manner or v i a bystander inhibition pathway (Chen et al., 1994; Groux et al., 1997; Jonuleit et al., 2000; M c G u i r k et al., 2002). Collectively, these Tregs were called antigen-induced Tregs or adaptive Tregs, to be distinguished from the naturally occurring Tregs (Bluestone and Abbas, 2003). In the past decade, many studies have shown that Tregs, both naturally occurring and antigen-induced, play a pivotal role i n the control o f immune responses to multiple antigens,  including autoantigens,  2  alloantigens and microbial antigens. Research on Tregs not only provides insight into tolerance induction and maintenance by the immune system, but also holds great promise in the development o f novel therapies  for various conditions such as  autoimmune diseases, organ transplantation and malignant tumors. The current consensus on the definition o f Tregs is that this term refers to all subsets o f T cells possessing the ability to suppress immune responses (Zou, 2005). While the CD4+CD25+ natural Tregs have been found to suppress immune responses triggered b y any antigen, some Tregs may not always be suppressive. A case i n point is the N K T cells, which were found to act as Tregs i n some situations, such as in acute graft-vs-host disease (Lan et al., 2001), but were pathogenic i n other systems (Zeng et al., 2003). Therefore, when the name o f Tregs is used, one must be aware o f the situation where this definition applies. Based on the lineage markers that define T cell subsets, Tregs have been found in C D 4 + (Honey et al., 1999; Sakaguchi, 2005), C D 8 + (Chang et al., 2002; Colovai et al., 2001; Jiang et al., 2003; Jiang et al., 2001), C D 4 - C D 8 - (Zhang et a l , 2000) and N K 1 . 1 + a p T C R + T cell subsets (Sharif et al., 2002; Sharif et a l , 2001). CD4+CD25+ naturally occurring Tregs refer to a subset o f C D 4 + T cells that constitutively express C D 2 5 and depend on the thymus for their generation. They have been found to play an important role in controlling immune responses triggered b y multiple types o f antigens, including autoantigens, alloantigens and microbial antigens (Sakaguchi et al., 2006). In contrast, antigen-induced CD4+ Tregs have been found i n multiple systems after antigen priming (Barrat et al., 2002; Graca et a l , 2000; Jonuleit et al., 2000; Ploix et al., 1999). In many cases, natural Tregs were not depleted before antigen priming, thus it is  3  difficult to make conclusions related to the origin o f the antigen-induced C D 4 + Tregs. They might be derived from the natural Tregs, C D 4 + C D 2 5 - conventional T cells, or both (Kingsley et al., 2002; Zhang et al., 2001). However, i n some reports, antigen priming followed the depletion o f natural Tregs, and in this case, antigen-induced C D 4 + Tregs were found to arise mainly from the conventional C D 4 + T cell compartment (Karim et a l , 2004). A brief summary o f different types o f Tregs can be found i n Figure 1.1.  1.1.2. Phenotypes o f Tregs In addition to the suppressive activities o f Tregs, they have other unique characteristics that can be used to differentiate them from naive T cells. These include activation-induced proliferation and apoptosis, cell markers and cytokine profile.  1.1.2.1. Activation-induced proliferation and apoptosis It has been well established that activation o f antigen specific T cells leads to the proliferation o f these T cells. Antigen-experienced memory T cells demonstrate enhanced proliferation i n response to antigen stimulation. However, most o f the studies show that both natural Tregs and antigen-induced Tregs exhibit an anergic state in response to antigen stimulation in terms o f their proliferation response in vitro. Groux et al showed that the low degree o f proliferation o f T r l cells i n response to cognate antigen stimulation was due to the high production o f IL-10 and lack o f IL-2 by these cells. IL-10 has been shown to inhibit the proliferation o f T cells, while IL-2 is a well known growth factor that supports T cell proliferation. Neutralization o f IL-10 and  4  addition o f IL-2 both could restore the proliferation o f T r l cells (Groux et al., 1997). Similar to T r l cells, Tregs induced in other systems generally show a decreased proliferation in response to antigen stimulation (Jonuleit et a l , 2000; Levings et al., 2005). However, in some situations, antigen-induced Tregs might also exhibit an enhanced proliferation profile (Kemper et al., 2003), even though the mechanism for this response is unclear.  ^CD4+: CD25+Foxp3+CTLA-4+GITR+: The socalled naturally occurring Tregs. Generation dependent on the thymus. Actions dependent on cell contact i n vitro.  r  CD4+-  T r l cells: Generation and actions  are both dependent on IL-10. Can be induced in vitro and i n vivo.  Th3 cells: Induced  by oral tolerance. Actions dependent  on T G F - B .  Other CD4+ Tregs: Induced  by multiple measures and antigens with various mechanisms of actions.  Tregs  Qa-1 restricted: Induced by Qa-1-presented antigens. Exert functions v i a cytotoxic pathway. CD8+Other CD8+ Tregs: Induced  by multiple measures and antigens with various mechanisms o f actions.  CD4-CD8-TCR+: The so-called double-negative  Tregs. Induced by alloantigens. Exert functions v i a cytotoxic pathway.  N K T cells: Generation dependent on C D 1 .  Actions dependent on IL-4.  Figure 1.1. Summary o f Tregs categories.  5  The ex vivo isolated natural Tregs proliferate poorly i n response to anti-CD3 stimulation in vitro as compared to C D 2 5 - conventional T cells. Addition o f exogenous IL-2 could restore their proliferative ability in vitro (Levings et al., 2001; Takahashi et al., 1998; Thornton and Shevach, 1998). This was also the basis o f expanding natural Tregs for therapeutic purposes. In contrast, the proliferation o f natural Tregs in vivo was somewhat different. When ex vivo purified natural Tregs were transferred into lymphopenic hosts, such as S C I D mice, these cells were able to undergo normal homeostatic proliferation and still maintained their suppressive activities (Gavin et al., 2002). Several groups further showed that natural Tregs could proliferate in normal mice and that this proliferation was dependent on the recognition o f the cognate antigen o f natural Tregs, demonstrating that the proliferation o f natural Tregs also played an important role in their response to foreign antigens and i n maintaining their peripheral homeostasis (Cozzo et al., 2003; Fisson et al., 2003; Walker et al., 2003). The proliferative state o f Tregs has important implications for the development o f effective immunotherapy. One critical issue in the clinical settings is the feasibility to obtain a sufficient number o f Tregs for therapeutic purposes. In humans, only about 1% of C D 4 + C D 2 5 + T cells actually possess suppressive activities (Baecher-Allan et al., 2001). Thus, it is extremely difficult to obtain a sufficient number o f these cells for therapeutic application, where a very large number o f Tregs are needed, such as in hematopoietic stem cell transplantation. Thus laborious in vitro expansion must be performed to harvest sufficient numbers o f natural Tregs. Multiple studies have shown that the expansion o f natural Tregs without loss o f function was possible both for human and murine natural Tregs (Beyersdorf et al., 2006; Karakhanova et al., 2006;  6  Levings et al., 2001; Masteller et al., 2005). In contrast to the proliferation profile o f Tregs, the apoptotic profile o f Tregs has been less well characterized. Although both murine and human natural Tregs are anergic i n response to stimulation v i a the T cell receptor in vitro, their apoptotic state seems to be different. Studies by Papiernik et al showed that mouse natural Tregs were resistant to apoptosis as they could survive both virus-induced clonal deletion and Fasinduced apoptosis in vivo (Papiernik ef a l , 1998). However, different apoptotic properties o f human natural Tregs were found. Taams et al reported that human natural Tregs isolated from human peripheral blood were prone to apoptosis in vitro, probably due to their low expression o f the anti-apoptotic gene bcl-2. Exogenous cytokines, such as IL-2 or IFN-P, could prevent the apoptosis (Taams L S et al., 2001). Another report showed that human natural Tregs were more resistant to T C R induced apoptosis but more susceptible to C D 9 5 induced apoptosis as compared to conventional T cells (Fritzsching et al., 2005). These data suggest that i f in vitro expanded natural Tregs are to be used for immunotherapy, more studies are required to characterize their apoptotic properties and determine their survival in vivo, since their ability to persist in vivo could be a crucial determinant o f their therapeutic outcome. Even less information on the apoptotic properties o f antigen-induced Tregs is available. Since antigen-induced Tregs are generated following antigen encounter, these Tregs are antigen-experienced cells. Thus it may be reasonable to assume that their ability to survive is somewhat similar to memory T cells. In one study, Zheng et al did show that Tregs generated in vitro in the presence o f exogenous I L - 2 and TGF-P were able to survive in vivo upon transfer to recipients and were able to suppress  7  chronic graft-vs-host disease (Zheng et al., 2004). However, more studies are necessary to characterize the apoptotic properties o f antigen-induced Tregs i n other systems.  1.1.2.2. C e l l markers Most studies o f the surface markers o f Tregs were conducted for natural Tregs. The first identified marker for natural Tregs, i n addition to the lineage marker C D 4 , was C D 2 5 (Sakaguchi et a l , 1995). Although under steady state, C D 2 5 can be used to identify the Tregs population, this marker is not specific for Tregs due to several reasons. First, C D 2 5 is also expressed by activated non-Tregs T cells. Thus this marker can not be used to identify natural Tregs from other conventional T cells under antigen stimulation. Second, it has been shown i n human natural Tregs that only T cells expressing high level o f CD25 had suppressive activities while those having low to medium level o f C D 2 5 expression were not suppressive (Baecher-Allan et al., 2001). Third, T cell clones generated from CD4+CD25+ Tregs showed that not all CD25+ clones were suppressive (Levings et al,, 2002). Fourth, further separation o f the CD25+ population by other markers, such as C D 6 2 L identified that suppressive Tregs were i n the C D 6 2 L + or C D 6 2  h i g h  compartment (Ermann et al., 2005; Taylor et a l , 2004).  Finally, C D 2 5 - Tregs have been identified frequently (Graca et al., 2002; Oida et al., 2003; Ono et a l , 2006; Stephens and Mason, 2000; Unger et al., 2003; Uraushihara et al., 2003). Despite these drawbacks, C D 2 5 is still a valuable marker for natural Tregs. This marker was also functionally linked to the suppressive activities o f natural Tregs as the constitutive expression o f this IL-2 receptor (CD25) allow them to engage IL-2 available i n the microenvironment, and natural Tregs can compete for I L - 2 produced  8  by naive T cells stimulated by antigens (de la Rosa M et a l , 2004). Following the identification o f CD25 as a marker o f natural Tregs, C T L A - 4 (Read et al., 2000; Takahashi et al., 2000) and G I T R ( M c H u g h et al., 2002) were also found to be specifically expressed by natural Tregs. However, as for C D 2 5 , these markers were also expressed on antigen stimulated T cells, thus they are not specific for Tregs. A major advance in the field came when several groups published their findings that the Foxp3 transcription factor was preferentially expressed on natural Tregs. Importantly, Foxp3 was found to be a critical factor i n the differentiation o f natural Tregs, as mice lacking the functional Foxp3 due to a mutation were deficient i n natural Tregs and succumbed to autoimmune diseases which could be cured by transfer o f natural Tregs. Additionally, over-expression o f Foxp3 i n mouse naive T cells caused them to acquire regulatory functions, suggesting this factor is not only necessary but also sufficient for the development o f Tregs (Fontenot et al., 2003; H o r i et al., 2003; Khattri et al., 2003). Although these original findings created much excitement, later studies revealed several drawbacks o f following expression o f Foxp3 as an exclusive Treg marker. It was found that human natural Tregs also preferentially express Foxp3, but unlike in mouse cells, retroviral-mediated over-expression o f this protein in human T cells did not result in acquisition o f stable and potent regulatory function (Allan et a l , 2005). Moreover, since Foxp3 is a transcriptional factor and is not expressed on the cell surface, it can not serve as a marker for purifying natural Tregs for clinical application purposes. Recent studies have also identified other proteins that are preferentially expressed by natural Tregs (Baecher-Allan et a l , 2006; Bruder D et al., 2004; Seddiki et al., 2006), however, further studies are needed to confirm these  9  findings. Markers for antigen-induced Tregs are even scarcer. Antigen-induced Tregs often exhibit a stimulated phenotype with enhanced expression o f C D 2 5 or C T L A - 4 . One study did identify L A G - 3 as a marker for antigen-induced Tregs (Huang et al., 2004). More studies are required to confirm that L A G - 3 can serve as a marker for antigen-induced Tregs, and to identify more suitable markers for antigen-induced Tregs. In summary, there is no satisfactory or specific marker identifying Tregs i n general, especially for antigen-induced Tregs. Thus the functional studies are the most reliable means for determining the presence o f Tregs, especially for the antigeninduced Tregs.  1.1.2.3. Cytokine profile The cytokine production profile has been an important characteristic for identifying distinct subsets o f effector T cells as exemplified i n the T h l and Th2 cells (Mosmann et al., 1986).. The most classic study which showed a distinct profile for antigen-induced Tregs was the first report o f T r l cells. In this study, the authors found that T r l cells, generated by stimulation o f T cells i n the presence o f exogenous IL-10, demonstrated a cytokine profile distinct from those o f T h l and Th2 cells. These cells did not produce T h l cytokines, such as IL-2 and IFN-y, or typical Th2 cytokines, such as IL-4. Yet they produced high levels o f IL-10 and medium levels o f TGF-p\ and in humans low levels o f LFN-y. Moreover, their cytokine profile was closely associated with their suppressive functions as IL-10 was necessary for the suppression mediated by T r l cells (Groux et al., 1997). Other grups also reported the existence o f Tregs that  10  mainly produced T G F - p (Fukaura et al., 1996; Ochi et a l , 2006; Yoshida et al., 2000; Zheng et a l , 2002). In contrast, the cytokine profile o f natural Tregs in vitro is characterized by a general lack o f cytokines, although there is some evidence that they may produce T G F - p (Levings et al., 2002) and/or IL-10 (Barthlott et al., 2005) in vitro. It must be noted that in contrast to T r l cells, natural Tregs exert suppression in vitro via  a contact-dependent  and cytokine-independent  manner.  Thus  although  the  production o f these immunosuppressive cytokines was detected, it does not necessarily imply that these cytokines play a role in the suppression mediated b y natural Tregs as in that mediated by antigen-induced Tregs. Generally speaking, the cytokine profile o f Tregs is characterized by decreased production o f proinflammatory T h l cytokines such as I L - 2 and IFN-y and increased production o f anti-inflammatory cytokines, such as IL-10 and T G F - p . However, more recent studies have shown that antigen-induced Tregs in different systems do not have a cytokine profile as typical as that o f T r l cells. Jonuleit et al showed that Tregs induced v i a repeated stimulation by allogeneic immature dendritic cells did increase the production o f IL-10. However, the suppression o f these Tregs was cell contact dependent (Jonuleit et al., 2000). Several recent studies showed that antigen-induced Tregs actually produced IFN-y (Hong et al., 2005; Kemper et al., 2003; Riemekasten et al., 2004; Sawitzki et al., 2005; Stock et al., 2004). Some types o f Tregs such as the C D 4 - C D 8 - Tregs, might not even have a distinct cytokine profile, and be capable o f producing many proinflammatory cytokines, including I L - 2 , IFN-y and T N F - a , yet nevertheless exert immunosuppressive effects leading to long term tolerance o f the allogeneic graft (Zhang et al., 2000) . Therefore, the presence o f Tregs cannot be  11  determined only by the cytokine profile o f the T cells, and a functional assay is still the best gold standard method to detect the presence o f Tregs.  1.1.3. Generation o f Tregs While natural Tregs depend on the thymus for their development, antigeninduced Tregs rely on antigen encounter in a specific environment for their generation. Multiple factors may be involved, and these are described below.  1.1.3.1. Role o f the thymus The thymus is the primary lymphoid organ that mediates the maturation o f T cells by positive and negative selection. The discovery o f natural Tregs revealed its third function: generation o f a subset o f Tregs (Seddon and Mason, 2000). The importance o f the thymus in the generation o f natural Tregs was revealed by the absence o f these Tregs in mice that received thymectomy on day 3 after birth. Natural Tregs were absent i n these mice resulting in the occurrence o f autoimmune diseases i n multiple organs. Transfer o f C D 4 + T cells from normal mice could prevent the diseases (Shevach, 2000). The process by which these Tregs are generated i n the thymus is not clear. Jordan et al used a transgenic mouse model to investigate the developmental process o f natural Tregs i n the thymus. The major finding was that thymocytes developing into natural Tregs required their T cell receptor ( T C R ) to have high affinity for autoantigen, but not too high to induce negative selection, because thymocytes i n T C R transgenic mice that have low affinity for autoantigen failed to develop into natural Tregs (Jordan et al., 2001). Romagnoli et al recently showed that bone marrow  12  derived antigen presenting cells (APCs), rather than thymic epithelial cells, were responsible for inducing selection o f natural Tregs (Romagnoli et al., 2005). Definitely, more studies are required to elucidate the developmental process o f natural Tregs i n the thymus. In addition to the role o f thymus in the generation o f natural Tregs, recent studies provided evidence for the generation o f CD4+CD25+Foxp3+ natural Tregs in the periphery, suggesting that the generation o f natural Tregs involved both thymusdependent and thymus-independent processes (Akbar et al., 2003). While the role o f the thymus in generating natural Tregs is clearly defined, the role o f the thymus in the generating antigen-induced Tregs is far less important than that o f natural Tregs, as it has been clearly shown that antigen-induced Tregs could develop i n athymic mice (Apostolou and von Boehmer, 2004). However, a recent paper by Deng et al showed that Tregs controlling transplant tolerance induced by a n t i - C D 4 5 R B was dependent upon the thymus for their generation, suggesting that the thymus may also play an important role in generating antigen-induced Tregs i n some systems (Deng et al., 2006).  1.1.3.2. Role o f antigens Antigens play a more crucial role in the development o f antigen-induced Tregs after their interactions with T cell subsets in the periphery. Several parameters affect the generation o f antigen-induced Tregs, including the structure and dose o f the antigen, and the method o f antigen administration. Studies have shown that modified antigens, such as altered peptide ligands, preferentially induced Tregs. Chen et al found that peptides with  low T cell stimulatory ability induce long term tolerance to a  transplantation antigen v i a generation o f Tregs (Chen et al., 2004). Some bacterial  13  antigens have an innate ability to induce Tregs, which may be an important mechanism for pathogenic organisms to evade the immune attack by the host (Lavelle et al., 2003; Marshall et al., 2003; M c G u i r k et al., 2002). The dose o f antigen is also important in regulating the generation o f Tregs. The most remarkable example has been shown in oral tolerance, where low doses o f antigen administered orally induce Tregs, while high doses o f the antigen actually induced deletion o f the antigen-specific T cells (Weiner, 2001). The way o f introducing antigens to the immune system is also critical for the generation o f Tregs. Firstly, the site o f antigen encounter may determine the generation o f Tregs vs. effector T cells. The most prominent example is mucosal tolerance. While subcutaneous immunization o f antigen with adjuvant activated antigen-specific T cells to become T h l cells, oral administration o f the same antigen induced antigen-specific T cells to become Tregs (Faria and Weiner, 2005). Secondly, repeated injection or chronic exposure to antigens was shown to be effective i n inducing Tregs. For example, use o f a pump to chronically release antigen i n mice i n which natural Tregs had been previously depleted by thymectomy and anti-CD25 resulted in the generation o f readily detectable Tregs (Apostolou and von Boehmer, 2004). M a i l e et al found that repeated stimulation o f T cells with alloantigens could induce C D 8 + Tregs (Maile R et al., 2006). The mechanisms by which antigens with altered structure or following chronic exposure induced Tregs are currently unclear. Garca et al proposed that the induction of Tregs by these two approaches is due to the delivery o f a suboptimal activation signal to antigen specific T cells, which drove them to become Tregs (Graca et al., 2005).  14  1.1.3.3. Role o f cytokines Cytokines are important players in deciding the T cell fate. Multiple cytokines have been shown to play important roles i n the generation o f both natural Tregs and antigen-induced  Tregs,  including interleukin-2  (IL-2),  interleukin-10  (IL-10),  transforming growth factor-(3 (TGF-P), and granulocyte-colony stimulating factor ( G CSF). IL-2 has long been regarded as a crucial cytokine for generating effective T cell response. Thus, it was rather surprising when it was found that disruption o f the IL-2 signaling pathway, by knockout o f IL-2 or I L - 2 receptor, resulted in autoimmune diseases o f multiple organs i n these knockout mice due to overreactive T cells to autoantigens (Sadlack B et al., 1995; Sadlack et al., 1993; Willerford et al., 1995). These findings demonstrated  that IL-2 is not only a growth factor for antigen  stimulated T cells, but may also be implicated i n the maintenance o f tolerance. However, it was not clear how IL-2 did this until the publication o f several papers demonstrating that IL-2 was a crucial factor i n the generation and maintenance o f natural Tregs (Furtado et a l , 2002; M a l e k et al., 2002; W o l f M et al., 2001). These authors showed that disrupting IL-2 signaling pathway specifically led to the failure o f development o f natural Tregs. The absence o f natural Tregs gave rise to the autoimmune diseases o f multiple organs, which also occurred i n mice lacking natural Tregs due to thymectomy early after they were born. Transfer o f natural Tregs could cure the diseases in these mice. They also showed that I L - 2 was not only required for T cell progenitor to develop into natural Tregs, but also required i n the periphery where IL-2 maintained the homeostasis o f natural Tregs. However, the role o f IL-2 in the  15  generation o f antigen-induced Tregs has not been w e l l studied, and remains unclear at this time. Ever since IL-10 was discovered, its role in suppressing immune responses has been well-defined (Moore et al., 2001). It was the classic paper by Groux et al that further demonstrated its important role i n the generation o f antigen-induced Tregs (Groux et al., 1997). These investigators stimulated T cells from transgenic mice exprssing T C R that is specific for O V A in the presence o f exogenous IL-10 in vitro, and then generated T cell clones also in the presence o f exogenous IL-10. They then found that the generated T cell clones could shut down the production o f T h l and Th2 cytokines but produced large amounts o f IL-10 and low level o f T G F - p i n response to cognate antigen stimulation in vitro. These T cells suppressed the response o f naive T cells to O V A in vitro v i a the production o f soluble IL-10. These cells could also mediate suppression o f colitis in vivo when they were activated. Neutralization o f I L 10 in vivo could reverse their suppression, further confirming their effects observed in vitro. More importantly, stimulation o f human T cells i n the presence o f exogenous I L 10 could also induce Tregs similar to mouse Tregs. This paper for the first time demonstrated the generation o f an effector T cell subset distinct from T h l and Th2 cells, and established the relationship between immunosuppressive cytokines and the generation o f Tregs, showing that a single cytokine i n addition to antigen stimulation was able to induce Tregs. The authors called Tregs generated i n this manner T regulatory cells type 1, or T r l cells. The induction o f T r l cells is also a typical example of antigen-induced Tregs. It also suggests that during immune responses endogenous IL-10 is able to drive the generation o f Tregs which i n turn w i l l regulate the immune  16  responses. The generation o f T r l cells has been confirmed in other systems (Akasaki et al., 2004; D i n g et al., 2006; Lundqvist et al., 2005). M o r e importantly, Massey et al showed that intranasal administration o f antigens were not able to induce Tregs in I L 10 knockout mice as compared to w i l d type mice, further demonstrating the important role o f IL-10 in the generation o f antigen-induced Tregs (Massey et al., 2002). TGF-P is another major immunosuppressive cytokine. Its role in maintaining immune tolerance can also be revealed in TGF-P knockout mice which show autoimmune diseases i n multiple organs (Shull et al., 1992). Studies showed that inducing Tregs is one o f the ways by which TGF-P maintains immune tolerance. Chen et al and Park et al both demonstrated that stimulation o f mouse T cells depleted o f the natural Tregs population (CD4+CD25-) could induce them to become Tregs in the presence o f TGF-P (Chen et al., 2003a; Park et al., 2004). Additionally, studies by Horwitz et al further expanded this finding in human T cells (Zheng et al., 2002). Thus it is clear that like IL-10, TGF-P has the ability to drive the generation o f antigeninduced Tregs. Moreover, it has been shown that IL-10 and T G F - P have synergistic effects i n inducing Tregs. Blazar's group studied the effects o f IL-10 and TGF-P in inducing tolerance to alloantigens. In mixed lymphocyte reactions, they  added  exogenous IL-10, TGF-P or both to the cultures. Then they reisolated the donor T cells to test their response to alloantigens. They found that while IL-10 or TGF-P alone decreased responses to alloantigens by donor T cells, as shown by a reduced ability to mediate graft-vs-host disease in vivo, donor T cells stimulated by alloantigens in the presence o f both IL-10 and TGF-P in fact lost the ability to mediate  graft-vs-host  disease. Moreover, they became Tregs as they suppressed acute graft-vs-host disease  17  ( a G V H D ) mediated by naive T cells. Additionally, Tregs generated i n the presence o f both IL-10 and TGF-P have more potent suppressive functions than those generated in IL-10 or TGF-P alone  (Chen et al., 2003b; Zeller et al., 1999). O n the other hand,  recent studies have also demonstrated that TGF-P is critical for maintaining the suppressive function and homeostasis o f natural Tregs i n the periphery (Huber et al., 2004; Marie et a l , 2005). G - C S F is a cytokine that directs the differentiation o f neutrophils. It is used to mobilize hematopoietic stem cells into the peripheral blood which can be collected for hematopoietic stem cell transplantation (Rutella et al., 2005). Although a G - C S F mobilized graft contains about 10 times more mature T cells as compared to a graft collected from the bone marrow, it does not significantly increase the incidence o f a G V H D , suggesting that this cytokine itself may have immune modulating effects. One study clearly demonstrated that the G - C S F mobilized graft contained dendritic cells that preferentially induced a Th2 response (Arpinati et al., 2000), suggesting that Th2 cells in response to alloantigen might inhibit a T h l response. A more recent study further showed that a structurally modified G - C S F could induce a unique subset o f antigen presenting cells (APCs) in vivo. Co-transplant o f this subset o f A P C s into an allogeneic host could suppress acute graft-vs-host disease v i a induction o f IL-10producing Tregs (Morris et a l , 2004). Moreover, it has been demonstrated that G - C S F can also induce the generation o f human T r l cells i n response to alloantigen stimulation (Rutella et al., 2002).  18  1.1.3.4. Role o f costimulation Effective T cell activation requires two signals, one delivered by the T cell receptor engaging its ligand, the MHC-peptide complex, and the other b y costimulatory signals which can be delivered by multiple pathways, including C D 2 8 - C D 8 0 / 8 6 or C D 4 0 / C D 4 0 L . The importance o f costimulation in T cell responses can be shown by the development o f T cell tolerance following costimulation blockade. Disruption o f the costimulation pathways has been shown to effectively induce transplantation tolerance and prevent autoimmune diseases. However, its role i n generating natural Tregs was only recognized a few years ago. Salomon et al crossed the N O D mice with C D 2 8 knockout mice, and surprisingly these mice showed a more dramatic decrease o f CD4+CD25+ natural Tregs accompanied by accelerated disease progression. Thus, for the first time it was shown that the generation or maintenance o f natural Tregs required an intact costimulation system (Salomon et al., 2000). Kumanogoh et al later showed that C D 4 0 and C D 4 0 L interaction was also required for the generation o f natural Tregs (Kumanogoh et a l , 2001). It was thought that the costimulation might not be required for the development o f natural Tregs i n the thymus, but was essential for the homeostasis o f natural Tregs in the periphery (Liang et al., 2005). Although it is clear that costimulation blockade could induce tolerance, the mechanism might not only be the inhibition o f T cell activation. Sykes's group showed that induction o f tolerance by blockade o f C D 4 0 and C D 4 0 L interaction mainly involved clonal deletion (Kurtz et al., 2004a).  However, it was frequently seen that  blockade o f costimulation actually led to the development o f dominant regulation mediated by Tregs. Waaga et al induced tolerance to M H C - m i s m a t c h kidney graft by  19  using C L T A - 4 I g and they isolated alloreactive T cell clones from these tolerant rats. A s compared to the control rats which rejected the graft, T cell clones from the tolerant rats showed a Th2 phenotype and were able to suppress the responses o f naive T cells to alloantigens (Waaga et al., 2001). Blazar's group had similar finding studying the role o f C D 4 0 and C D 4 0 L costimulation in a G V H D . They stimulated the donor T cells with alloantigens in vitro in the presence o f anti-CD40L antibody to block this pathway. The stimulated T cells not only showed dramatically decreased ability to mediate a G V H D , but also became Tregs as they suppressed a G V H D when cotransferred with na'ive T cells (Blazar et al., 1998; Taylor et al., 2002a). However, the mechanisms o f Tregs induction by costimulation blockade are not clear currently. In contrast to the induction o f Tregs v i a costimulation blockade, activation o f certain costimulation pathways could lead to the induction o f Tregs (Akbari et al., 2002; Masuyama et a l , 2002; Watanabe et a l , 2006).  1.1.3.5. Role o f other cell types A t the cellular level, the outcome o f T cell fate depends on their interactions with other cell types. For the generation o f Tregs, especially antigen-induced Tregs, other cell types have great impact. T cells require A P C s to be activated and their ultimate development is dependent very much on their interaction with A P C s . Dendritic cells are the only type of professional A P C s that are able to prime naive T cells. This unique ability enables them to play an important role in determining the cell fate o f T cells with which they are interacting. Dendritic cells are a heterogeneous population which consists o f  20  different subsets. Based on the activation status, dendritic cells can be divided into mature and immature dendritic cells. While mature dendritic cells are characterized by high expression o f M H C and costimulation molecules and ability to induce the development o f effector T cells, immature dendritic cells express low levels o f M H C and costimulation molecules and the ability to induce tolerance o f T cells (Banchereau et al., 2000). The induction o f T cell tolerance by immature dendritic cells includes T cell anergy and generation o f Tregs. Several studies have shown that priming o f nai've T cells with immature dendritic cells led to the generation o f Tregs (Charbonnier et al., 2006; Dhodapkar and Steinman, 2002; Gilliet and L i u , 2002; Jonuleit et al., 2000; Mahnke et al., 2003). Thus immature dendritic cells have also been called tolerogenic dendritic cells. Immunosuppressive cytokines i n some cases are the critical mediator for the effects o f dendritic cells on the generation o f Tregs. Steinbrink et al showed that IL-10-treated dendritic cells had an immature dendritic cell phenotype and induced Tregs (Steinbrink et al., 2002). Dendritic cells residing in some anatomical sites preferentially show ability to induce Tregs. For example, dendritic cells isolated from the respiratory tract produce high levels o f IL-10 i n response to antigen stimulation and induce responding T cells to become Tregs. The induction o f Tregs by dendritic cells at such sites is an important mechanism by which tolerance to harmless antigens is maintained (Akbari et al., 2001). Although the ability o f immature dendritic cells to induce Tregs has been attributed to their low expression o f M H C and costimulation molecules and secretion o f immunosuppressive cytokines, other mechanisms may also exist.. In addition to mediating direct suppression on their target cells, natural Tregs  21  may function to allow other tolerogenic measures to induce Tregs. Blazar et al developed an in vitro system using anti-CD40L antibody to induce tolerance o f alloantigens. The donor T cells became Tregs when stimulated by alloantigens in the presence o f anti-CD40L antibodies as these donor T cells could suppress a G V H D when co-transferred with naive T cells. However, i f natural Tregs were depleted from the donor T cell population before alloantigen stimulation, blockade o f C D 4 0 / C D 4 0 L did not lead to generation o f Tregs (Taylor et al., 2002a). However, this effect o f natural Tregs might not apply to other system for Tregs induction. In the absence o f natural Tregs, T G F - p \ IL-10 or repeated stimulation by antigens are still able to effectively induce Tregs. For example, Levings et al showed that human allogeneic specific T r l cells were still able to be generated i n the absence o f natural Tregs, while neutralization o f IL-10 led to the failure to generate these cells (Levings et al., 2005). Other than natural Tregs, N K T cells and mast cells have also been found to be essential for the generation o f antigen-induced Tregs in some systems ( K i m et al., 2006; L u et a l , 2006).  1.1.4. Mechanisms o f immunomodulation by Tregs 1.1.4.1. Targets o f Tregs Studies revealed that natural Tregs could exert their suppression over a wide range o f target cells. The most studied target is the C D 4 + T cell. Thornton et al developed the in vitro suppression system where natural Tregs and C D 2 5 - conventional C D 4 + T cells were co-cultured in the presence  o f anti-CD3 stimulation. The  proliferation o f these T cells was used as the readout o f suppression. While natural  22  Tregs showed very poor proliferation in response to anti-CD3 stimulation, they suppressed the proliferation o f conventional C D 4 + T cells (Thornton and Shevach, 1998). In vivo suppression by natural Tregs on the C D 4 + T cells was demonstrated in several disease models where transfer o f conventional C D 4 + T cells caused diseases and co-transfer o f natural Tregs attenuated or prevented the disease (Annacker et a l , 2001; Mottet et al., 2003; Suri-Payer et al., 1998). Later on Piccirillo et al first found in mice that natural Tregs could also directly suppress the functions o f C D 8 + T cells, including the activation o f and IFN-y production by C D 8 + T cells (Piccirillo and Shevach, 2001). Kursar et al demonstrated a role for natural Tregs in vivo i n an infection model. These studies implied that natural Tregs could act like a two edged-sword. O n the one hand, they might regulate the C D 8 + T cell response to limit their harmful effects during immune responses. However, on the other hand, their presence might also blunt the favorable response mediated by C D 8 + T cells, such as tumor immunity (Kursar et al., 2002). Subsequent studies i n a tumor model did prove the latter. Antony et al demonstrated that natural Tregs suppressed the tumor immunity mediated by C D 8 + T cells (Antony et al., 2005). Studies have also demonstrated that human natural Tregs can suppress C D 8 + T cell functions similarly (Robertson et al., 2006). A P C s , especially dendritic cells, play an important role in initiating and regulating immune responses. Thus it is not surprising that Tregs could exert their effects on this component. Cederbom et al first showed that mouse natural Tregs downregulated the costimulatory molecules on A P C s (Cederbom L et al., 2000). Prevention of maturation o f dendritic cell by human natural Tregs was also demonstrated. Similar  23  to suppression o f T cells, the effects o f natural Tregs on A P C s were also cell-contact dependent (Misra et al., 2004). Houot et al recently demonstrated that natural Tregs were only able to suppress  the activation o f myeloid dendritic cells, but  not  plasmacytoid dendritic cells (Houot et al., 2006). In addition, natural Tregs have also been found to regulate the functions o f B cells (Janssens et al., 2003; L i m et al., 2004; L i m et al., 2005; Seo et al., 2002; Zhao et al., 2006). In vitro studies showed that they mainly exert their effects on B cells v i a cytotoxic mechanisms. Moreover, they were also able to inhibit the functions o f N K and N K T cells (Azuma et al., 2003; Ghiringhelli et al., 2005; Smyth et al., 2006; Trzonkowski et al., 2006).  1.1.4.2. Role o f cytokines Immunosuppressive cytokines, mainly IL-10 and T G F - p \ are crucial in the suppression mediated by antigen-induced Tregs. IL-10 has been found to act as the main soluble factor that mediated antigen-induced Tregs. T r l cells, generated i n the presence o f exogenous IL-10, relied on releasing IL-10 for their actions. Thus in vitro, they were still suppressive even when they were separated from the target cells by a permeable membrane (Groux et al., 1997; M c G u i r k et al., 2002). IL-10 could act on multiple cell types, including T cells and A P C s . It could down-regulate the expression of M H C and costimulatory molecules by A P C s , and suppress the production o f proinflammatory cytokines (Moore et al., 2001). Thus, by releasing this cytokine, I L 10-producing Tregs were able to affect multiple targets, leading to down-regulation o f immune responses.  Moreover, IL-10-treated A P C s  were  able to induce  Tregs  themselves, thus amplifying the effect o f IL-10 i n addition to its direct suppressive  24  effects (Steinbrink et al., 2002). TGF-P is another important immunosuppressive cytokine implicated in the suppressive function o f Tregs. Chen et al first found that oral tolerance was associated with the generation o f TGF-P-producing T cells with regulatory function. A t that time, the concept o f Tregs was still not well established, and they called this cell type Th3 cells to distinguish them from the T h l and Th2 cells that had been described several years  before  (Chen et al., 1994). Other groups  using different  systems  also  demonstrated a critical role played by TGF-P-producing Tregs (Fukaura et al., 1996; Ochi et a l , 2006; Yoshida et al., 2000; Zheng et al., 2002). T G F - P was also able to induce naive T cells to develop into Tregs (Chen et al., 2003a; Park et al., 2004), or to confer dendritic cells with  tolerogenic ability which i n turn induced the T cells  interacting with them to develop into Tregs (Kosiewicz et al., 2004). Interleukin-4 (IL-4) is a cytokine that drives the Th2 response and attenuates the T h l response (Cher and Mosmann, 1987; Krenger et al., 1995; Mosmann and Sad, 1996). IL-4-dependent suppression is prominent in N K T cell-mediated inhibition. C o transfer o f donor bone marrow derived N K T cells effectively inhibited a G V H D i n an IL-4-dependent manner, as N K T cells from IL-4 knockout mice were not protective (Zeng et al., 1999). Activation o f host N K T cells after bone marrow transplantation or by conditioning the host with multiple doses o f total lymphoid irradiation and antibody-mediated T cell depletion to enrich the host N K T cells, suppressed a G V H D , and this effect was also IL-4-dependent (Hashimoto et al., 2005; L a n et al., 2001; L a n et al., 2003). The mechanism for this IL-4-dependent suppression was thought to be mediated by a skewed Th2 response, as donor T cells showed a Th2 cytokine profile.  25  In the case o f natural Tregs, the role o f cytokines i n their suppression is controversial. O n the one hand, the suppression mediated by natural Tregs i n co-culture systems in vitro was found to be cell-contact dependent and cytokine-independent (Thornton and  Shevach,  1998). This  seems to rule out the  role o f soluble  immunosuppressive factors. However, in vivo studies found that IL-10 was required for the suppression mediated by natural Tregs i n some systems. The studies conducted by Dieckmann et al and Jonuleit et al provided a possible answer for the observation. They found that natural Tregs were able to induce their target cells to develop into IL-10producing Tregs which i n turn exerted suppressive functions (Dieckmann et al., 2002; Jonuleit et al., 2002). A s to the role o f T G F - p , more conflicting results have been reported. Kullberg et al and Piccirillo et al showed by using Smad3 knockout mice that natural Tregs did not require TGF-P to suppress (Kullberg M C et al., 2006; Piccirillo et al., 2002). Levings et al generated natural Tregs clones and showed that they produced TGF-P as compared to T r l cell clones which produced IL-10 (Levings et al., 2002), and showed that T G F - p was not required for suppression. In contrast, other groups showed that T G F - p was definitely required for the suppression by natural Tregs (Fahlen et al., 2005; Green et al., 2003; Nakamura et al., 2004). The discrepancy in these results could probably be due to the different strains o f knockout mice used and/or the fact that the role o f TGF-P is different depending on the model system.  1.1.4.3. Cytotoxic mechanisms Apoptosis is an essential mechanism for immune tolerance and regulating immune responses. For example, apoptosis mediated by Fas and FasL interaction is  26  one o f the main ways to ensure tolerance to antigens i n immune privileged sites (Ferguson and Griffith,  2006). Activation induced cell death  is an  important  mechanism for controlling the expansion o f effector T cells i n the late phase o f the immune response (Schluns and Lefrancois, 2003; W o n g and Pamer, 2003). The importance o f the cytotoxic pathway i n immune regulation can also be shown in mice that have mutations in molecules mediating cytotoxic effects. For example, the Ipr mice have uncontrolled lymphoproliferative diseases due to the mutated Fas molecule (Adachi et al., 1996). The cytotoxic effects o f Tregs were first found i n antigen-induced Tregs. In the experimental allergic encephalitis ( E A E ) model, it was found that T cell immunization attenuated E A E , which was associated with the generation o f C D 8 + cytotoxic T cells restricted by Qa-1 molecule. These CD8+ cytotoxic T cells were able to recognize and k i l l the T cells activated by autoantigens which expressed Qa-1, thus attenuating E A E (Jiang et al., 1995; Jiang et al., 1998). Another example o f cytotoxicity mediated by Tregs was the C D 4 - C D 8 - double negative Tregs. Donor-specific transfusion-induced double negative Tregs conferred long term tolerance to M H C mismatched heart graft. Zhang et al showed that these cells were able to acquire MHC-peptide complex from the host cells and interacted with the pathogenic T cells v i a these molecules, triggering the cytotoxic process mediated by the double negative Tregs (Ford et al., 2002; Zhang et al., 2000). Cytotoxicity could also be mediated by soluble factors. Q i n et al recently identified granzyme B and perforin to be the soluble factors that mediated the suppressive effects o f a Treg clone. They also showed that the granzyme B induced apoptosis was perforin dependent (Qin et al., 2006).  27  The cytotoxic effects mediated by natural Tregs were recognized several years ago. Janssens et al and Zhao et al both found that mouse natural Tregs were able to induce the death o f B cells, which might be a major pathway to regulate B cell responses (Janssens et al., 2003; Zhao et al., 2006). Grossman et al showed that human antigen-induced Tregs expressed mainly granzyme A while natural Tregs mainly expressed granzyme B and were able to mediate cytotoxic effects v i a these molecules (Grossman et al., 2004). However, the in vivo significance o f cytotoxic effects mediated by natural Tregs was not established and more studies are needed to further investigate this issue.  1.1.4.4. Inhibition o f activation by Tregs Activation o f antigen-specific C D 4 + or C D 8 + T cells is a critical step in generating productive immune responses. Activation o f antigen-reactive T cells was associated with the expression o f activation markers, such as C D 2 5 and C D 6 9 . Several studies using in vitro systems showed that natural Tregs co-cultured with conventional C D 4 + T cells or C D 8 + T cells inhibited the expression o f C D 2 5 on target cells (Barthlott et al., 2005; de la Rosa M et al., 2004; George T C et al., 2003; Piccirillo et a l , 2002), indicating that the natural Tregs could directly suppress the activation o f target cells v i a T cell to T cell interactions. T w o studies showed that natural Tregs suppressed the activation o f C D 4 + conventional T cells by deprivation o f IL-2 produced by the activated target cells, but it was not known i f natural Tregs suppressed the activation o f C D 8 + T cells in the same way (Barthlott et al., 2005; de la Rosa M et al., 2004). In vivo studies showed that natural Tregs were able to inhibit the  28  upregulation o f C D 2 5 by naive T cells when they were co-transferred to lymphopenic recipient mice (Barthlott et a l , 2005). Trenado et al demonstrated similar effects in an a G V H D model. They isolated natural Tregs and then expanded them by stimulation with allogeneic A P C s i n the presence o f IL-2. These allogeneic specific natural Tregs were then co-transferred with naive T cells to irradiated allogeneic host to study their effects on naive T cells. They found that naive T cells co-transferred with ex vivo expanded natural Tregs showed a decreased expression o f activation markers C D 2 5 and C D 6 9 (Trenado et al., 2006), thus further confirming the results from in vitro studies. However, more studies are needed to determine i f this mechanism also applies to human Tregs.  1.1.4.5. Infectious tolerance The term "infectious tolerance" was first coined by Gershon to describe the regulation by T cells. Infectious tolerance was further demonstrated by Waldmann's group in a series o f studies using antibody to induce transplantation tolerance. They found that injection o f a non-depleting anti-CD4 antibody was able to induce transplantation tolerance which was due to the induction o f dominant regulation mediated by C D 4 + Tregs. Transfer o f C D 4 + Tregs generated b y anti-CD4 antibody to other naive hosts prevented rejection o f allogeneic graft. Importantly, C D 4 + T cells o f the naive hosts developed regulatory functions and were able to confer regulatory functions to other naive hosts. Thus the original C D 4 + Tregs induced by anti-CD4 antibody were able not only to directly suppress the responses o f target T cells interacting with them, but also to convey the ability to induce other T cells to become  29  Tregs to the target T cells. Thus, the term "infectious" was used to describe this property (Qin et al., 1993). Their studies established that non-depleting antibodies to C D 4 and C D 8 , and an antibody that blocks C D 4 0 / C D 4 0 L interactions, all were able to induce infectious tolerance (Cobbold and Waldmann, 1998; Graca et al., 2000). This system was the best characterized model where infectious tolerance mediated by antigen-induced Tregs could be generated. Recent studies also revealed that infectious tolerance was involved in the regulation mediated by Tregs induced v i a mucosal tolerance (Alpan et a l , 2004; Unger et al., 2003). A l p a n et al further showed a critical role that might be played by A P C s . Using a unique mouse model, they showed that oral tolerance-induced Tregs were able to induce other T cells to acquire a cytokine profile similar to Tregs. They demonstrated that A P C s were critical mediators o f this effect, as A P C s co-cultured with oral tolerance-induced Tregs were able to induce other T cells to become Tregs. IL-4 and IL-10 both played a role i n transferring the Tregs-inducing effects from oral tolerance-induced Tregs to A P C s . However, other unidentified molecules may also be involved. Subsequent  studies demonstrated  that infectious tolerance might be  one  important mechanism for the suppression by natural Tregs. Several studies found that human natural Tregs were able induce their target cells to become Tregs in vitro (Dieckrnann et a l , 2002; Dieckmann et al., 2005; Jonuleit et al., 2002). Dieckmann et al showed that human C D 4 + C D 2 5 - conventional T cells co-cultured with natural Tregs in the presence o f anti-CD3 stimulation not only became anergic and but also acquired suppressive functions mediated by the production o f IL-10. The Treg-induction ability of natural Tregs required cell-cell contact and was mediated by molecules expressed  30  after activation as paraformaledhyde-fixed ex vivo isolated natural Tregs did not have this effect. However, the authors did not identify the molecules responsible for this effect (Dieckmann et al., 2002). Using a similar system, Jonuleit et al reported similar results. However, these author did demonstrate that TGF-P was i n part responsible for the ability o f natural Tregs to induce their target cells to become Tregs (Jonuleit et al., 2002). Collectively, these studies suggest that natural Tregs might exert their suppression by inducing the target cells to become Tregs. O n the one hand, this could change the response o f naive T cells to antigens; on the other hand, these induced Tregs could suppress the responses o f naive T cells that had not been changed into Tregs, thus leading to the amplification o f suppressive functions o f natural Tregs.  1.1.4.6. Cytokine competition Cytokines are crucial for the development and regulation o f immune responses. It has been found over two decades ago that competition o f cytokines was an important way to regulate immune responses (Gunther J et al., 1982; Scheffold A et al., 2005). A previous study suggested that anergic T cells with regulatory functions were able to inhibit responses o f target cells by competition for I L - 2 . Recent studies showed that cytokine competition was one o f the main mechanisms by which natural Tregs exert their suppressive functions (Barthlott et al., 2005; de la Rosa M et al., 2004). de la Rosa et al, using an in vitro co-culture system, showed that human natural Tregs inhibited the activation o f conventional T cells and suppressed their proliferation by depriving them o f I L - 2 . Moreover, in the presence o f activation and I L - 2 , the suppressive functions o f natural Tregs were also enhanced (de la Rosa M et al., 2004). Barthlott et  31  al made similar observations using mouse natural Tregs i n an in vitro system. They further showed that cytokine competition also applied to the suppression by natural Tregs in vivo. Moreover, they observed that upon consuming the I L - 2 produced by target cells, IL-10 production by natural Tregs was increased, and i n their system, I L 10 was required for the suppression mediated by natural Tregs in vivo (Barthlott et al., 2005). Cytokine competition mediated by Tregs is o f great importance to tumor immunotherapy. Recently, it has been shown that adoptive transfer o f ex vivo expanded tumor infiltrating lymphocytes to non-lethally irradiated recipients induced potent anti-tumor effects and tumor regression (Dudley et al., 2002). In a mouse tumor model, Overwijk et al showed that administration o f IL-2 and transfer o f tumor specific T cells was also able to induce tumor regression (Overwijk et al., 2003). These studies suggested that the presence o f natural Tregs deprived effector T cells o f cytokines which are important for the generation o f effective tumor immunity. That was why removal o f natural Tregs by irradiation or addition o f exogenous I L - 2 to saturate the consumption by natural Tregs was able to boost the anti-tumor immunity (Gattinoni et al., 2006).  1.1.4.7. In vivo actions Most o f the studies concerning the mechanisms o f Tregs were performed using in vitro systems, and not much is known about the in vivo behavior o f Tregs in terms o f how they exert their effects on their targets. Recent studies have begun to focus on the in vivo behavior o f Tregs, mainly on that o f the natural Tregs. For the natural Tregs to  32  mediate their suppression in vivo, the first important issue is how these cells go to the sites where they are needed. Some recent studies demonstrated that natural Tregs might have a unique migration pattern made possible by their expression o f molecules facilitating their migration (Huehn and Hamann, 2005; Siegmund et al., 2005; W e i et al., 2006; Wysocki et al., 2005; Yurchenko et a l , 2006). While in vitro co-culture systems are able to reveal the outcome o f suppression, advances o f in vivo imaging techniques have made it possible to dissect the in vivo interactions o f natural Tregs with their target cells in more detail. Although studies using in vitro systems showed that natural Tregs are able to directly suppress the responses o f C D 4 + T cells, in vivo studies revealed that natural Tregs did not directly make contact with pathogenic C D 4 + T cells and suggested that their suppressive effects were mainly exerted v i a the interactions with dendritic cells (Tadokoro et al., 2006; Tang et al., 2006). However, natural Tregs could directly inhibit the functions o f C D 8 + cytotoxic T lymphocytes in vivo (Mempel et al., 2006). In addition, in vivo studies also suggested that antigen-induced Tregs might exert a suppressive effect on target cells in ways that differ from those o f natural Tregs (Tischner et al., 2006).  1.1.5. Role o f Tregs i n disease pathogenesis and their potential applications Research on Tregs has generated insightful understandings i n multiple diseases, and has also suggested novel therapeutic strategies for them. Three categories o f diseases are selected for discussion as they are the most representative examples in this regard.  33  1.1.5.1. Autoimmune diseases The discovery o f natural Tregs has been closely linked to the pathogenesis o f autoimmune diseases since it was the lack o f this subset o f Tregs that induced the generation o f autoimmune diseases i n animal models. Thus, i n addition to neagtive selection, the generation o f Tregs by the thymus is an important regulatory pathway used by the immune system to maintain tolerance to autoantigens i n the periphery (Itoh et al., 1999; Sakaguchi et al., 1995; Shevach, 2000). The presence o f a counterpart to mouse natural Tregs in humans is a further indication that this is an evolutionally conserved mechanism (Baecher-Allan et al., 2001). The studies o f murine natural Tregs suggested that the malfunction o f natural Tregs might lead to the loss o f tolerance to autoantigens, and i n turn cause autoimmune diseases, although this may not be the only factor in inducing autoimmune diseases ( M c H u g h and Shevach, 2002). Multiple recent studies in human subjects with autoimmune diseases support the importance o f Tregs in their pathogenesis as functional defects have been detected in the natural Tregs o f these patients (Bacchetta et a l , 2006; Balandina et al., 2005; Kriegel et al., 2004; Lawson et al., 2006; Lindley et al., 2005; Viglietta et al., 2004). Ehrenstein et al showed that effective treatment o f autoimmune disease was associated with the recovery o f suppressive functions o f natural Tregs, further demonstrating a link between the function o f natural Tregs and the pathogenesis o f autoimmune diseases (Ehrenstein et al., 2004). The studies o f natural Tregs i n autoimmune diseases not only shed lights on the pathogenesis o f autoimmune diseases, but also provided novel strategies for the treatment o f autoimmune diseases. In multiple diseases, transfer o f ex vivo purified  34  natural Tregs before disease induction was able to prevent autoimmune  diseases  (Kohm et al., 2002). For example, transfer o f natural Tregs to mice with established colitis could prevent disease progression and lead to cure o f the disease (Mottet et al., 2003). Although transfer o f natural Tregs is an effective therapy for autoimmune diseases experimentally, there are obstacles for this to be applied clinically. First, as mentioned above, since the functions o f the natural Tregs i n autoimmune diseases are defective, infusion o f autologous natural Tregs may not be effective in controlling the diseases. Second, the truly functional natural Tregs consist o f only about 1-2% o f the peripheral C D 4 + T cells. This small number o f cells can be o f any therapeutic value only when they are expanded in vitro, a labor-intensive task. Thus, in the face o f obstacles for using natural Tregs to treat autoimmune diseases, the use o f antigen-induced Tregs seems to be more attractive. Oral tolerance and T cell vaccination, which'are known to induce TGF-P-producing Th3 cells (Chen et al., 1994) and Qa-1-restricted C D 8 Tregs respectively, have been shown to control autoimmune diseases experimentally (Jiang et a l , 1998). These studies did lead to clinical trials with some success (Achiron et al., 2004; Correale et al., 2000). Although natural Tregs isolated from patients  with autoimmune  disease may not be o f  therapeutic value, Tregs generated against other model antigens from the whole T cell repertoire in vitro still can suppress autoimmune disease v i a bystander inhibition as long as these Tregs are activated in vivo by their cognate antigens (Groux et al., 1997; L i et al., 2006; Stohlman et a l , 1999). A typical example is the transfer o f T r l cells raised against O V A which suppressed colitis when the mice were fed O V A to activate these T r l cells in vivo (Groux et al., 1997). Therefore, use o f Tregs to control  35  autoimmune diseases has potential therapeutic value, although a lot o f work is still required.  1.1.5.2. Transplantation Solid organ transplantation and hematopoietic stem cell transplantation are used to treat multiple diseases. However, graft rejection and G V H D remain major obstacles. Although the use o f powerful immunosuppressive drugs has partially solved the problem, induction o f tolerance to transplantation has been regarded as a more desirable solution. Research in Tregs has generated great impact on the field o f transplantation in recent years. First, studies have shown that alloantigen-specific Tregs can be induced to mediate transplantation tolerance in experimental models, such as anti-CD4 antibody infusion (Qin et al., 1993), co-stimulation blockade (Taylor et al., 2002a) and transfer o f dendritic cells with regulatory functions (Morris et al., 2004). In patients  who  have  received  transplanted  organs  and  withdrawn  from  immunosuppressive drugs, a tolerant state was associated with the generation o f donorspecific Tregs (Waaga et al., 2001). These studies highlighted the importance o f Tregs in the control o f immune tolerance to alloantigens. Second, studies on the effects o f immunosuppressive drugs on Tregs revealed important clinically relevant findings. Certain immunosuppressive drugs may have differential effects on Tregs. Although cyclosporine A is the most used immunosuppressive drug, it works by interfering with the IL-2 pathway, which is also required by natural Tregs. Given the role o f natural Tregs i n transplantation, use o f cyclosporine A may not have favorable effects in some circumstances (Segundo et al., 2006; Zeiser et al., 2006). Thus, there is a need to  36  consider how to optimize the use o f this powerful drug. O n the other hand, it was also found that certain types o f immunosuppressive drug might preferentially preserve the functions o f Tregs, and these drugs should be further explored to maximize their effects (Battaglia et al., 2005; Coenen et al., 2006; Segundo et al., 2006; Zeiser et al., 2006). Third, the application o f Tregs in the control o f transplantation rejection may revolutionize the therapy for this disease. In addition to these use o f Tregs from the whole T cell repertoire without in vitro manipulation, the use o f natural Tregs also provides further opportunities for treatment. It has been shown that alloantigen-specific Tregs can be generated in vitro by stimulating isolated natural Tregs with alloantigens and further expanded using cytokines favoring their growth, such as I L - 2 . Transfer o f these alloantigen-specific natural Tregs could lead to permanent graft acceptance (Nishimura et al., 2004). More importantly, unlike the effects o f immunosuppressive drugs which are non-specific and may also blunt favorable immune responses, such as immune responses to infections and tumors, regulation by Tregs allows more favorable immune responses to occur. Several studies showed that transfer o f donor-derived natural Tregs to recipients suppressed a G V H D while allowing the (GVT)  effects  to be preserved  in murine  graft-vs-tumor  allogeneic hematopoietic  stem  cell  transplantation models. This w i l l definitely enable allogeneic hematopoietic stem cell transplantation to be more widely used to treat solid or hematopoietic malignancies (Edinger et al., 2003; Jones et al., 2003; Trenado et al., 2003). Furthermore, K a r i m et al showed that natural Tregs primed by a model antigen could mediate active suppression o f transplantation rejection o f skin grafts v i a bystander inhibition, providing another novel strategy for using Tregs to control transplantation rejection (Karim et al., 2005).  37  Hematopoietic stem cell transplantation ( H S C T ) was first used as a rescue following  high dose  chemotherapy  and  irradiation therapy  for  hematopoietic  malignancies since these treatments cause hematopoietic failure. However, after about three decades o f research, it was found that the curative effects o f H S C T were not due to the  ability to rescue  hematopoietic  failure. Experimentally, murine  studies  demonstrated that transfer o f allogeneic immune cells could eradicate leukemia cells (Truitt, 2004). In clinical settings, it was observed that H S C T between twins and removal o f T cells in the graft had higher recurrence rate than allogeneic H S C T ( A H S C T ) with T cell-replete graft, suggesting that the curative effects o f A H S C T was due to the immune responses directed against alloantigens mediated by donor T cells (Horowitz et al., 1990). This notion was finally confirmed when donor lymphocytes infused into recipients with recurrent leukemia following H S C T successfully induced disease regression (Kolb et al., 1990; K o l b et al., 1995). Currently A H S C T has been regarded  as the  most powerful and well-characterized immunotherapy  against  hematopoietic malignancies, whose anti-tumor effects have been termed graft-vs-tumor ( G V T ) effects (Fowler, 2006; K o l b et al., 2004) . Realizing its potency against hematopoietic malignancies, A H S C T has been successfully used to treat various solid tumors in recent years (Childs et al., 2000; Talmadge, 2003). However, the great benefits o f G V T effects have been seriously undermined by concurrent graft-vs-host disease ( G V H D ) (Kolb et al., 2004; Reddy and Ferrara, 2003). The occurrence o f G V H D requires the presence o f three conditions: 1) inability o f the host to reject the donor graft; 2) a donor graft with immune competent cells; and 3) mismatched transplantation antigens between the donor and the host. These  38  conditions are frequently seen in conditioned recipients o f A H S C T (Devetten and Vose, 2004). Thus, G V H D occurs most frequently i n the setting o f A H S C T . Based on the pathophysiology and clinical manifestation, there are two types graft-vs-host diseases, acute G V H D and chronic G V H D . While chronic G V H D happens generally 100 days after bone marrow transplantation and manifests  as an autoimmune disease-like  syndrome, acute G V H D happens within 100 days after transplantation and causes a very high mortality rate. a G V H D manifests as an inflammatory process with acute onset that affects multiple target organs, including the skin, liver, gut and lungs. Studies i n the past decade have gained insightful understandings o f this life-threatening disease. Before transplantation, the potent conditioning measures, such as total body irradiation and high dose chemotherapy, cause tissue damage, to the gut in particular, and lead to endotoxin translocation from the gut. After transfusion o f a donor graft, T cells i n the graft are activated by alloantigens presented by the host A P C s . This response can be greatly potentiated by the presence o f endotoxin released into the systemic circulation. Then, activated T cells i n turn develop into a T h l response and induce high levels o f proinflammatory cytokines, including I L - l p Y L F N - y and T N F - a , leading to the so-called "cytokine storm". This further results i n damage o f the gut that leads to more endotoxin translocation, forming a positive feedback amplification loop. The activated T cells w i l l then home to and infiltrate multiple target organs, inducing tissue damage that finally gives rise to failure o f the target organs (Devetten and Vose, 2004; Jaksch and Mattsson, 2005; Reddy and Ferrara, 2003). During the priming o f donor T cells, memory T cells w i l l be generated and are responsible for the persistence  39  of the disease (Zhang et al., 2005a; Zhang et al., 2005b). The use o f powerful immunosuppressants has been effective to some degree i n preventing or treating a G V H D . However, this treatment can also lead to serious problems such as disease recurrence, infection and toxicity. After it was demonstrated that T cells were responsible for a G V H D , T cell depletion v i a antibody was used for the prevention o f a G V H D , and has been highly successful. However, due to the loss o f T cells that mediate graft-vs-tumor and anti-infection effects, disease recurrence and serious infections make this treatment less useful (Goker et al., 2001). A s the conditioning-induced damage plays a critical role i n the pathogenesis o f a G V H D , conditioning measures o f reduced intensity have been developed to prevent a G V H D . However, a G V H D still remains the most serious complication despite some success (Slavin et al., 2002). The identification o f natural Tregs has prompted the development o f novel approaches against a G V H D . Several groups investigated the use o f donor derived natural Tregs to prevent a G V H D . They found that mouse natural Tregs not only showed low responses to alloantigen stimulation, but also suppressed the responses o f naive T cells to alloantigens in vitro. More importantly, they did not mediate a G V H D in vivo, but suppressed the ability o f naive T cells to mediate a G V H D when cotransferred with naive T cells to irradiated recipients (Cohen et al., 2002; Edinger et al., 2003; Hoffmann et a l , 2002; Jones et al., 2003; Taylor et al., 2001; Taylor et al., 2002b; Trenado et al., 2003; Trenado et al., 2006). In addition, several groups found that suppression o f a G V H D by natural Tregs did not affect the G V T responses, making it a potentially very attractive form o f therapy clinically (Edinger et al., 2003; Jones et al.,  40  2003; Trenado et a l , 2003). Based on the report that the suppressive functions o f natural Tregs were enhanced when these Tregs were activated (Thornton and Shevach, 1998), Taylor et al showed that natural Tregs stimulated and expanded in vitro still retained their suppressive activities on a G V H D in vivo (Taylor et al., 2002b). Trenado et al also developed an in vitro expansion system, where they stimulated natural Tregs with allogeneic A P C s in the presence o f exogenous I L - 2 to generate alloantigenspecific natural Tregs. They reported that this approach could expand natural Tregs by 100-fold in 30 days. These expanded alloantigen-specific natural Tregs had more superior suppressive activities over a G V H D  as compared to the natural Tregs  polyclonally expanded with anti-CD3 and anti-CD28 microbeads, and lead to long term survival o f the recipients. Furthermore, co-transplant o f these alloantigen-primed natural Tregs preserved G V T effects and led to accelerated immune reconstitution (Trenado et al., 2003; Trenado et al., 2006). These studies have important implications for the use o f natural Tregs clinically. For natural Tregs present in human peripheral blood, only 1% o f the CD4+CD25+ T cells are truly suppressive (Baecher-Allan et al., 2001). Based on the murine studies above, suppression o f a G V H D in vivo by natural Tregs was observed when the ratio of naive T cells to natural Tregs was 2:1 or 1:1, meaning that the number o f natural Tregs should be at least half o f that o f the naive T cells. This is a great challenge due to the scarcity o f the truly suppressive natural Tregs. However, the studies mentioned above did show the feasibility o f expanding natural Tregs in vitro to the desired number. Moreover, it has also been shown that human natural Tregs indeed can be expanded in large scale (Hoffmann et al., 2004). Currently, there is an ongoing clinical trial to  41  investigate i f expanded human natural Tregs can be used to suppress a G V H D in A H S C T (Lechler et al., 2005). Tregs generated in other systems were also shown to inhibit a G V H D i n murine studies. Blazar's group demonstrated that natural Tregs-depleted C D 2 5 - conventional T cells became hyporesponsive to alloantigens and had a greatly decreased ability to mediate a G V H D in vivo after they were stimulated with alloantigens i n the presence o f exogenous IL-10 and TGF-P in vitro. More importantly, these cells became Tregs as they suppressed a G V H D mediated by naive T cells when they were co-transferred with naive T cells (Chen et al., 2003b; Zeller et al., 1999). They also showed i n an in vitro mixed lymophocyte reaction system that blockade o f C D 4 0 and C D 4 0 L interactions could also lead to hyporesponsiveness to alloantigens b y donor T cells with suppressive activities on a G V H D in vivo (Taylor et al., 2002a). However, these studies did not lead to a clinical trial. One possible reason is the anergic state o f the generated Tregs which makes it extremely difficult to obtain sufficient number o f Tregs to be used clinically. Th2 cells have been known to suppress the responses mediated by T h l cells. Thus Th2 cells were also studied for their ability to suppress a G V H D which was regarded as a pathogenic T h l response. E x vivo generated Th2 cells were demonstrated to suppress a G V H D in vivo while not affecting the G V T responses (Foley et al., 2005). A clinical trial has been performed in a very small cohort o f patients, making it difficult to draw any conclusions at the present time and its efficacy still needs to be further tested. Furthermore, it was reported that C D 4 - C D 8 - T cells were also able to inhibit a G V H D (Young et al., 2003). However, due to the scarcity o f this cell type,  42  these cells may be o f very little clinical importance. In addition to the suppression mediated by conventional  NKl.rapT  cells,  N K l . l a p T cells were also shown to inhibit a G V H D . Strober et al identified a cell +  population in the mouse bone marrow that was able to suppress mixed lympohocyte reaction in vivo about two decades ago (Strober et al., 1987). Later, Strober's group further demonstrated that donor derived bone marrow N K T cells were able to suppress a G V H D when co-transferred with naive donor T cells i n an I L - 4 dependent manner (Zeng et al., 1999). Despite their suppressive activities on a G V H D , it is not feasible to enrich and harvest donor derived bone marrow N K T cells for transplantation. O n the other hand, the host N K T cells were also shown to regulate a G V H D i f they were properly activated (Hashimoto et al., 2005). Thus Strober's group developed a conditioning approach that specifically enriched the host N K T cells (Lan et al., 2001; Lan et al., 2003). The approach included multiple doses o f total lymphoid irradiation plus antibody-mediated T cell depletion. After this conditioning, the dominant cell population in the lymphoid organs was N K T cells. a G V H D was totally blocked when naive donor T cells were transferred to these recipients while the same number o f naive T cells caused severe lethal a G V H D when transferred to recipients receiving total body irradiation as conditioning. They showed i n a murine a G V H D model that one hundred times more donor T cells were required to mediate lethal a G V H D i n recipients conditioned by multiple total lymphoid irradiation and antibody-mediated T cell depletion  as  compared  to  mice receiving conventional total  body irradiation  conditioning. The lack o f a G V H D was due to the dominant suppression by N K T cells enriched by the conditioning approach since C D 1 knockout mice recipients lacking  43  N K T cells were not protected against a G V H D . The suppression was dependent on IL-4 as IL-4 knockout mice recipients were not protected either. In addition, suppression o f a G V H D did not interfere with the G V T responses. More importantly, the results o f this approach was highly reproducible in humans and very promising clinical data have been reported using this approach to prevent  aGVHD  ( L o w s k y et al., 2005).  Collectively, these studies show that suppression o f a G V H D b y Tregs may be a potential solution for a G V H D which remains the main barrier for the wide clinical use o f A H S C T to treat multiple diseases.  1.1.5.3. Infections and malignant tumors Although the immune system has evolved to develop many powerful measures to successfully fight infections caused by myriads o f microorganisms, immune evasion by pathogens is still common. Recently, much attention has been paid to the role o f Tregs i n the immune evasion by pathogens. M a n y pathogens can block the immune attack by inducing Tregs. Recent experimental studies have accumulated ample evidence showing that effective immune response against some pathogens could be blunted by pathogen-induced Tregs (Belkaid et a l , 2006; Lavelle et al., 2003; Marshall et al., 2003; M c G u i r k et al., 2002; Rouse et al., 2006). In support o f these experimental studies, virus specific Tregs have been found to be present clinically in patients infected with different viruses that suppressed the T h l responses to the viruses (Bolacchi et al., 2006; Franzese et al., 2005; Rushbrook et al., 2005; X u et al., 2006). The presence o f natural Tregs is another important factor that favors the persistence o f pathogens. Belkaid et al first demonstrated that natural Tregs were responsible for the  44  persistence o f intracellular pathogens (Belkaid et al., 2002) and removal o f natural Tregs could result in the enhanced immune response against these pathogens (Rad et al., 2006). Similar to microbial infections, Tregs also play a critical role i n the immune evasion o f malignant tumors. Malignant tumors have developed effective ways, such as production o f immunosuppressive cytokines IL-10 or T G F - P , to induce tumor antigenreactive T cells to become Tregs, favoring the persistence o f tumor cells (Gajewski et al., 2006; M u n n and Mellor, 2006). Furthermore, the presence o f natural Tregs likely suppresses the overall responses to malignant tumors, as natural Tregs have also been shown to suppress C D 4 and C D 8 T cells, N K cells and N K T cells, which may play a role i n mediating tumor immunity. The important role o f natural Tregs in the control o f tumor immunity is demonstrated by prolonged tumor survival i n the absence o f this population (Shimizu et al., 1999). Elucidating the roles o f Tregs i n the pathogenesis o f infections and malignant tumors has great impact on developing more effective therapies against them. In contrast to their role i n the therapy o f autoimmune diseases and transplantation, Tregs should be deleted to allow more powerful anti-infection and anti-tumor effects in most cases. Depletion o f Tregs has been shown to induce effective anti-tumor (Shimizu et al., 1999; Turk et al., 2004; V i e h l et al., 2006) and anti-infection immunity (Mendez et al., 2004; Taylor et a l , 2005) in experimental studies. Induction o f lymphopenia was shown to greatly enhance tumor immunity. One o f the mechanisms which has been proposed is the removal o f natural Tregs (Gattinoni et al., 2006). Vaccination is an effective way to prevent infections. Multiple groups have shown that elimination o f  45  natural Tregs prior to immunization effectively enhanced the efficacy o f vaccination against multiple infectious pathogens (Moore et al., 2005; Stober et a l , 2005; Tabbara et al., 2005; Toka et al., 2004). O f note, although vaccination with tumor antigens works well i n animal models with malignant tumors, to date it has not been remarkably effective in human subjects. Possibly, the presence o f natural Tregs may limit the effectiveness o f these strategies as their removal can greatly potentiate the efficacy o f vaccination against malignant tumors (Antony et al., 2005; Kudo-Saito et al., 2005).  1.2. Superantigens 1.2.1. Definition and categories o f superantigens For nominal antigens to be recognized by T cells, they must be taken up and processed by A P C s . The antigen-derived peptide w i l l bind to M H C molecules and are presented to T cells v i a interactions with the T cell receptor. Generally, only one in 10 -10 T cells w i l l be activated as a result. However, this process does not apply to 4  5  superantigens which are defined by their unique interactions with T cells. The term superantigen was first coined by Marrack et al to define their ability to directly crosslink the M H C class II molecules and the T cell receptor i n a V P specific manner and activate the T cells without the requirement o f prior uptake and processing by A P C s . Thus, due to their relatively non-selective actions as compared to the nominal antigens, superantigens can activate 5-30% o f the T cell population dependent on the size of a certain V P family o f T cells (White J et al., 1989). According to the sources o f superantigens, they can be divided into bacterial and viral superantigens. The gram-positive Staphylococcus aureus is one o f the major  46  microorganisms producing bacterial superantigens. V i r a l superantigens derived from the endogenous mouse mammary tumor viruses were the best characterized viral superantigens  (Leung  et  al., 1997). Additionally,  based  on  the  target  cells,  superantigens can also be divided into T cell superantigens and B cell superantigens. The definition above also applies to B cell superantigens since they are able to bind to the conserved region o f the antibody molecule and activate polyclonal B cells (Goodyear and Silverman, 2003; V i a u and Zouali, 2005). This action is different from the interactions o f nominal antigens and antibodies, where antigen recognition involves the complementarity determining regions o f the antibody. It should be noted that bacterial T cell superantigens are the main focus for discussion i n this thesis.  1.2.2. Biological effects o f superantigens A l l superantigens possess the ability to activate a large number o f T cells or B cells. In vitro, stimulation o f human or murine naive T cells with bacterial superantigens leads to massive proliferation and production o f T h l cytokines, such as IL-2, T N F - a , IFN-y (Arad et a l , 2001; Dauwalder et a l , 2006; K u m et al., 2001). Costimulation is required for the activation o f T cells by superantigens ( K u m W W et al., 2006; Saha et a l , 1996a; Saha et al., 1996b). This activation process in turn results in the deletion v i a apoptosis or anergy o f superantigen-reactive T cells (Kawabe and Ochi, 1991; MacDonald et al., 1991; Rellahan et a l , 1990; White et a l , 1989). C D 4 + T cells are the main cell type responding to superantigen stimulation, although C D 8 + T cells can also be activated by superantigens (Sabapathy et al., 1994). It is thought that the activation o f T cells by superantigens plays an important role i n the pathogenesis o f  47  toxic shock syndrome, both i n human cases and i n animal models o f toxic shock syndrome. Suppression o f antibodies against superantigens has been  documented,  which has been thought to be the result o f a preferential T h l polarization effect o f superantigens (Leung et al., 1997). The second biological activity o f superantigens is their ability to enhance endotoxin-induced lethal shock (Dinges et al., 2000; Leung et al., 1997). It has been shown i n mice and rabbits that injection o f bacterial superantigens could greatly enhance the response caused b y endotoxin by up to 50,000 fold. This enhancement has been hypothesized to be the direct induction o f hypersensitivity to endotoxin or indirect effects o f decreasing the clearance o f enodotoxin. Bacterial superantigens can also cause emesis i n clinical settings, which is characteristic o f S E A and S E B , the causative agents for food poisoning. However, the mechanism for this biological effect is unknown (Dinges et al., 2000; Leung et al., 1997). In addition, superantigens are able to directly interact with epithelial cells and activate them. This interaction involves receptors other than M H C Class II molecules and the T cell receptor, but the precise nature is elusive at the present time. The effects of superantigens on epithelial cells have also been implicated i n the pathogenesis o f toxic shock syndrome.  1.2.3. Staphylococcal superantigens Staphylococcus aureus produces various exoproteins that include toxic shock syndrome toxin-1 (TSST-1), the staphylococcal enterotoxins ( S E A , S E B , S E C n , S E D , S E E , S E G , S E H , and SEI), the exfoliative toxin A and B ( E T A and E T B ) , and  48  leukocidin.  These  proteins  play  an  important  role  in  the  pathogenesis  of  Staphylococcus aureus infections. TSST-1 and the staphylococcal enterotoxins are also called pyrogenic toxin superantigens (PTSAgs) due to their superantigenic properties (Dinges et a l , 2000).  1.2.3.1. TSST-1 Encoded by tstH present on the bacterial chromosome within a 15.2-kb mobile genetic element called staphylococcal pathogenicity island 1, the mature TSST-1 is a single polypeptide chain with a molecular weight o f 22,000 daltons. TSST-1 is highly soluble in water, although it contains a high percentage o f hydrophobic amino acids. TSST-1 is resistant to heat and proteolysis as can be shown by the fact that it can be boiled for more than 1 h without detectable loss o f biological activity, and it is not cleaved after  prolonged exposure  to trypsin (Dinges et al., 2000). TSST-1 is  antigenically distinct from other P T S A g s and does not have significant primary sequence homology to other known proteins, including other P T S A g s (Dinges et al., 2000). TSST-1 was first identified as the causative agent o f toxic shock syndrome associated with menstruation and use o f tampons. TSST-1 reacts with V p 2 o f human T cells (Proft and Fraser, 2003). In B 1 0 . B R mice, it has been reported that TSST-1 primarily stimulated VP 15 (Callahan et al., 1990). A s compared to other P T S A g s , TSST-1 possesses the unique ability to cross vaginal mucosal surfaces and is the only P T S A g s able to reactivate bacterial cell-wall induced arthritis. However, unlike the staphylococcal enterotoxins, TSST-1 does not possess the ability to cause emesis  49  (Dinges et al., 2000). Structurally, TSST-1 consists o f two adjacent domains. Domain A (residues 1 to 17 and 90-194) contains a long central a-helix (residue 125-140) surrounded by a five-strand P-sheet. Domain B (residue 18-89) consists o f five P strands forming a barrel motif (Acharya et al., 1994; Prasad G S et al., 1993). One study demonstrated a zinc-binding site in the TSST-1 molecule which might potentiate superantigenicity at low toxin concentration (Prasad et al., 1997) although the binding o f TSST-1 to M H C Class II molecules is Zinc-independent. Crosslinking o f the T cell receptor and M H C Class II molecules is required for the superantigenic  activities o f T S S T - 1 ,  as  demonstrated by mutational studies. Hurley et al first showed that the binding o f TSST-1 to the T cell receptor required the central a-helix, and an a-helix at the N terminal (Hurley et al., 1995). Studies in our laboratory identified several important binding sites o f TSST-1 to the M H C Class II molecules ( K u m W W et al., 1996; K u m W W et a l , 2000) . The interaction o f TSST-1 with M H C Class II molecule and the T cell receptor is illustrated in Figure 1.2-A.  1.2.3.2. S E A and other staphylococcal enterotoxins Staphylococcal enterotoxins consist o f a family o f structurally related exotoxins produced  by certain  Staphylococcus  aureus bacterial  strains.  They were  first  recognized as agents causing food poisoning and can be divided into serological types, S E A , S E B , SECi-3, S E D S E E , S E H and SEI. The mature form o f S E A consists o f 233 amino acids with a molecular weight o f 2 3 K D (Dinges et a l , 2000). S E A stimulates a wide range o f V P families o f human T cells, including V p i . l , 5.3, 6.3, 6.4, 6.9, 7.3, 7.4,  50  9.1, and 18 (Proft and Fraser, 2003). In B 1 0 . B R mice, S E A was found to react with V p 1, 3, and 11 families o f T cells (Callahan, 1990) S E A is structurally similar to T S S T - 1 . However, differences are also noted. H i g h affinity binding o f S E A to M H C Class II molecules is Zinc-dependent (Schad et a l , 1995) and a number o f studies suggested that S E A had two binding sites on M H C Class II molecules, as S E A competed with TSST-1 i n binding H L A - D R 1 while T S S T I did not inhibit S E A binding (Hudson et al., 1995; Thibodeau et al., 1994). S E A is the most  potent superantigen  among the  staphylococcal enterotoxins.  Recent  data  suggested that this might be due to its stronger Zinc-dependent binding to M H C class II molecules (Pless et al., 2005). The inferred binding o f S E A to M H C Class II molecule is shown i n Figure 1.2-B.  1.2.4. Role o f bacterial superantigens i n diseases Superantigens  have  been  implicated  in  multiple  human  diseases.  Staphylococcal enterotoxins are involved in food poisoning due to their ability to cause emesis. Toxic shock syndrome (TSS) is characterized clinically by hypotension, diffuse erythematous rash, generalized edema and lymphocytopenia. TSST-1 is the dominant toxin implicated in T S S , menstrual cases in particular, and other P T S A g s are important in non-menstrual T S S . The superantigenic activities o f these toxins cause massive polyclonal  activation o f T cells followed  by massive  production o f  proinflammatory cytokines that in turn result in severe hypotension, hyperthermia, and the devastating dysregulation o f multiple physiological systems and organs, leading to life-threatening multi-organ failure. The pathogenicity o f T S S T - 1 is also enhanced by  51  its unique ability to cross mucosal surfaces (Dinges et al., 2000; M c C o r m i c k et al., 2001). Other contributing factors o f superantigens to the initiation and progression o f TSS include the direct activation o f the innate immune system v i a their interactions with epithelial cells (Watson et al., 2005) and ability to potentiate the effects o f endotoxin (Blank C et al., 1997).  TSST-1  ___  Figure 1.2. Diagram o f binding o f superantigens to their ligands. A . A modeled trimolecular complex composed o f a T S S T - 1 , M H C Class II molecule and T cell receptor (Dinges et al., 2000). B . Interactions o f two S E A molecules with two M H C Class II molecules inferred from mutation studies (Fraser et al., 2000).  52  Due to their ability to activate polyclonal T cells which may include the autoantigen-reactive  T  cells,  superantigens  have  long  been  implicated  in the  pathogenesis o f autoimmune diseases. O n the one hand, clinical and experimental data showed that selective expansion o f VP family o f T cells has been detected in patients with various autoimmune disease, suggesting the exposure o f pathogen-derived superantigens, although direct identification o f pathogen-derived superantigens i n these studies was not successful (Conrad et al., 1994; Paliard et al., 1991). One study suggested that superantigens derived from human endogenous retroviruses could be a possible cause o f type I diabetes (Conrad et al., 1997). accumulated by studies showing that timely injection  Further evidence was  o f superantigens  caused  activation or acceleration o f autoimmune diseases i n multiple animal models (Brocke et al., 1993; Kageyama et al., 2001; Schiffenbauer et al., 1993; Schwab et al., 1993; Soos et al., 1995). In addition to the direct activation o f autoantigen-reactive T cells, the proinflammatory responses elicited by superantigens may indirectly facilitate the activation o f autoreactive T cells; thus leading to the initiation or progression o f autoimmune diseases. O n the other hand, the ability o f superantigens to mediate deletion o f T cells also provides a therapeutic strategy for the treatment o f autoimmune diseases. A s the superantigen-reactive T cells overlapped with autoreactive T cell clones that mediate autoimmune diseases, it has been suggested that administration o f superantigens could actually attenuate autoimmune diseases. This has also been well proven i n various animal models o f autoimmune disease (Kawamura et al., 1993; K i m et al., 1991; Prabhu Das et al., 1996; Soos et al., 1993). The timing o f superantigen administration was critical in generating the therapeutic effects. The same concept has  53  also been demonstrated using B cell superantigens i n autoimmune disease models where B cells play an important role (Viau and Zouali, 2 0 0 5 ) . The superior ability o f superantigens to induce activation o f the immune system is also favorable for the treatment o f malignant tumors where tolerance to the progressing tumors frequently occurs. Stimulation o f T cells i n draining lymph nodes o f mice bearing tumors could overcome the deficient function o f T cells, leading to tumor-specific effector T cells that were able to mediate tumor regression (Shu et al., 1994). Injection o f superantigens to mice could confer protection against tumor challenge or induce regression o f established tumors i n mice (Torres et al., 2001). Similarly, superantigens were able to enhance the efficacy o f other immunotherapeutic measures against malignant tumors (Gidlof et al., 1997; Pulaski et al., 2000; Wahlsten et al., 1998). A recent clinical trial showed that injection o f bacterial superantigen had favorable effects on pleural effusions caused by non-small cell .lung cancer and prolonged the survival o f patients (Ren et al., 2004). M o r e studies are required to further exploit the beneficial effects o f superantigens on the therapy against malignant tumors.  1.3. Connections o f bacterial superantigens and Tregs Research on superantigens related to their role i n diseases and in particular their interactions with T cells, has contributed much to our understanding o f the mechanisms of immune tolerance. After the intrathymic deletion o f immature self-reactive T cells was conclusively shown in the 80's, the mechanisms o f induction o f peripheral tolerance in mature T cells were still poorly understood. Studies i n the late 80's and  54  early 90's using superantigens as a paradigm for tolerizing antigens shed new light into this field. Exposure o f T cells to bacterial superantigens, both in vivo and in vitro, could result i n the clonal expansion o f the responding T cells bearing specific VP segments followed by the selective clonal deletion o f a large fraction o f the responding T cells mediated b y apoptosis. The remaining T cells showed an anergic state i n response to superantigen stimulation (Kawabe and Ochi, 1991; M a c D o n a l d et al., 1991; Rellahan et al., 1990; White et al., 1989). This was the first direct evidence showing recessive mechanisms played a role i n the peripheral immune tolerance to antigens. These findings supported the hypothesis that peripheral tolerance to autoantigens that were not expressed i n the thymus during negative selection was i n part mediated by clonal deletion or induction o f anergy o f autoreactive T cell clones. Later, some studies showed i n other experimental systems that anergic T cells actually could mediate active suppression o f responses o f naive T cells, suggesting that immune tolerance could also be induced or maintained b y T cell-mediated dominant mechanisms. This concept at that time was still controversial, although the idea o f regulating immune responses b y T cells with suppressive functions had been first raised about two decades before (Diaz-Gallo et al., 1992; Lombardi G et a l , 1994). However, i n the middle 90's, a series o f studies conducted b y Sakaguchi et al. convincingly demonstrated  that peripheral tolerance to autoantigens  indeed was  maintained by a dominant mechanism mediated by a subset o f C D 4 + T cells coexpressing the a unit o f the IL-2 receptor C D 2 5 (Asano et al., 1996; Sakaguchi et al., 1995).  55  Studies on the induction o f Tregs by superantigens were first performed by Sundstedt et al who showed that repeated injection o f S E A to mice induced high levels of IL-10 and low level o f proinflammatory cytokines I L - 2 , IFN-y and T N F - a in the serum. Furthermore, enhanced IL-10 production was actually responsible for the decreased levels o f IFN-y and T N F - a , but not I L - 2 . Depletion studies showed that C D 4 + T cells were responsible for the high production o f IL-10. Collectively, these data suggested that repeated administration o f a bacterial superantigen could induce a subset o f C D 4 + T cells with regulatory functions (Sundstedt et al., 1997). Following this study, consecutive reports confirmed that repeated injection o f low to moderate doses o f superantigens induced Tregs. C D 4 + T cells from mice treated with repeated injection o f superantigens not only changed their cytokine production profile and proliferative state, but also developed active suppressive functions on the responses o f naive cells to the superantigen stimulation (Feunou et al., 2003; Grundstrom et al., 2003; N o e l et al., 2001). For example, T cells from mice injected repeatedly with S E A or S E B produced lower levels o f IL-2 and IFN-y, but increased level o f IL-10. When co-cultured with naive splenocytes in the presence o f superantigen stimulation, these superantigen-primed T cells suppressed the ability o f naive splenocytes to proliferate and to produce IL-2 and IFN-y (Noel et al., 2001). These studies collectively suggested that i n addition to clonal deletion and anergy, a dominant mechanism mediated by superantigen-induced  Tregs  existed  for  the  control  o f immune  responses  superantigens.  56  to  Chapter 2: Hypothesis and specific aims M y studies focused on the study o f Tregs induced by the staphylococcal superantigen known as toxin shock syndrome toxin-1 (TSST-1). Previously, studies i n our laboratory showed that repeated injection o f TSST-1 was able to induce Tregs i n B A L B / c mice. However, most o f the basic features o f the T S S T - l - i n d u c e d Tregs were unclear, such as the specificity and mechanism o f action o f their suppressive activities, their capacity to proliferate and produce cytokines, and their susceptibility to apoptosis. The unifying hypothesis for the present project is: TSST-1 induces Tregs with distinctive properties Surrounding this hypothesis, the specific aims were: •  To determine whether repeated injection o f T S S T - 1 can induce regulatory T cells i n C 5 7 B L / 6 (B6) mice. Although previous studies i n our laboratory showed that repeated injection o f TSST-1 could induced Tregs i n B A L B / c mice, this finding should be further confirm using B 6 mice. In vitro experiments using a co-culture system and in vivo studies using a lethal shock model would be performed to assay the suppressive functions o f TSST-1-primed C D 4 + T cells.  •  To elucidate the differences o f Tregs induced by TSST-1 and S E A . It is not clear whether T S S T - l - i n d u c e d Tregs differ from those induced b y S E A . A side-by-side comparison o f Tregs induced b y T S S T - 1 and S E A would be performed. These Tregs would be compared for their cytokine profile (cytokine secretion and intracellular expression), proliferation response, cell  57  death, markers (surface and intracellular), and suppressive activities (in vitro and in vivo). To elucidate the mechanisms o f T S S T - l - i n d u c e d Tregs. The mechanisms o f T S S T - l - i n d u c e d Tregs were unclear and required further investigation. Studies would be performed to determine whether T S S T - l - i n d u c e d Tregs could induce cell death or inhibit the activation o f their target cells. The role o f infectious tolerance, soluble factors and cytokine competition i n the suppression mediated by T S S T - l - i n d u c e d Tregs would also be studied. To investigate the potential clinical application o f T S S T - l - i n d u c e d Tregs. Acute graft-vs-host disease ( a G V H D ) is the major complication following allogeneic bone  marrow transplantation  or transplantation  of  organs  containing lymphoid tissues, such as the small intestine. Triggered by the mismatched alloantigens expressed by the recipients, a G V H D is the major barrier for the wide use o f allogeneic bone marrow transplantation, which is otherwise the curative therapy for a number o f hematological diseases including leukemia (Goker et al., 2001). Multiple recent studies showed that immunotherapy, the use o f Tregs in particular, may provide a promising potential solution (Blazar and Taylor, 2005; Goker et al., 2001). Elucidating the effects o f TSST-1 induced Tregs on a G V H D w i l l provide a novel therapeutic strategy for a G V H D . The clinical application potential o f these cells would be addressed by using a murine a G V H D model.  58  Chapter 3: Methods and materials 3.1. M i c e , TSST-1 preparation and TSST-1 treatment Female C 5 7 B L / 6 (H-2 ) and B 6 D 2 F 1 ( H - 2 ) mice, 6-8 weeks old, were b  b/d  purchased from Charles River Laboratories (Wilmington, M A ) and housed at Jack B e l l Research Centre. Recombinant TSST-1 was purified from culture supernatants o f Staphylococcus aureus RN4220 containing plasmid pBS(tst) by a combination o f preparative isoelectric focusing and chromatofocusing. Our laboratory has previously demonstrated that highly pure TSST-1 preparations can be obtained using this method ( K u m et al., 1993). Purity o f toxin preparations was demonstrated b y silver staining after S D S - P A G E . S E A was purchased from T o x i n Technology (Sarasota, F L ) and was further purified by chromato-focusing, using a p H 6-8 gradient polybuffer exchanger. M i c e were subcutaneously injected with 4 pg TSST-1 or S E A i n 0.2 m L P B S for three times at the interval o f 4 days. Control mice were subcutaneously given the same volume o f P B S . Splenocytes or C D 4 + T cells from mice treated with T S S T - 1 , S E A or P B S were isolated 2 hours after the 3 injection. rd  3.2. C e l l isolation To isolate splenocytes, spleens from naive, T S S T - 1 , S E A or PBS-treated mice were removed and homogenized to single cell suspension on a cell strainer ( B D Pharmingen).  Cells were washed with R P M I  1640 culture medium (Stemcell  Technologies, Vancouver, B C ) and red cells were removed b y hypotonic lysis in Gey's solution for 1 minute. Splenocytes then were washed with R P M I 1640 culture medium twice and resuspended in complete R P M I 1640 culture medium at 8><10 /mL. C e l l 6  59  viability was determined b y typan blue exclusion. C D 4 + T cells were further isolated from the splenocytes by using a magnetic isolation kit according to the product manual (Stemcell Technologies, Vancouver, B C ) and adjusted to 2 x i o / m L for in vitro 6  studies. For studies o f T S S T - l - p r i m e d C D 4 + T cells to suppress lethal shock, C D 4 + T cells were adjusted to 5 x l 0  7  / m L . For studies o f T S S T - l - p r i m e d C D 4 + T cells to  suppress a G V H D , C D 4 + T cells were adjusted to 2 . 5 x l 0 / m L . The purity o f C D 4 + T 7  cells was greater than 90% as determined b y flow cytometry.  3.3. C e l l transfer To investigate the suppressive effects o f superantigen-primed C D 4 + T cells on lethal shock, C D 4 + T cells were isolated from mice repeatedly injected with T S S T - 1 , S E A or P B S . 2 x l 0  7  o f C D 4 + T cells i n 0.4 m L P B S were injected to each mouse v i a  the lateral tail vein. Two hours later, mice were subjected to lethal shock induction and survival was monitored. In some groups, C D 4 + T cell-depleted, splenocytes were recovered and injected into mice at the dose o f 8 x 1 0 /mouse.  3.4. In vitro culture For co-culture experiments, 4 x l 0 / w e l l naive splenocytes were co-cultured with 5  titrated numbers o f T S S T - 1 , S E A or PBS-primed C D 4 + T cells i n flat-bottom, 96-well plates i n the presence o f TSST-1 or S E A (1 n M ) . Cells were cultured at 37°C, 5% C 0 in 200 u L R P M I 1640 medium supplemented with 10% F B S (Hyclone, Logan, U T ) , 50 uM  2-mercaptoethanol ( 2 - M E ; Sigma, St Louis, M O ) , 10 m M H E P E S buffer, I m M  sodium pyruvate, antibiotics [100 U / m L penicillin; 100 m g / m L streptomycin (Stemcell  60  2  Technologies, Vancouver, B C ) ; and 2 u,g/mL polymycin B (Sigma, St Louis, M O ) ] . Supernatant was collected after 48 hours o f culture. In some experiments, T S S T - 1 , S E A or P B S primed splenocytes ( 4 x l 0 conditions. Supernatant  5  /well) were also cultured under these  was collected at different  time points post-stimulation.  Supernatant was stored at -70°C until assay for various cytokines b y E L I S A .  3.5. C F S E staining and proliferation assay To label cells with C F S E , splenocytes or C D 4 + T cells ( 5 x l 0 / m L ) were stained 6  with 5 n M C F S E (Molecular Probes, Eugene, O R ) in P B S at 37°C for 10 minutes. Cells were washed with cold R P M I 1640 medium for three times and then resuspended in completed R P M I 1640 medium containing 10% o f F B S for 30 minute. To assay the proliferation  o f superantigen-primed  CD4+  T cells,  4xl0  5  /well  CFSE-labeled  splenocytes from mice treated with TSST-1 or S E A were plated i n flat-bottom, 96-well plates and cultured at 37°C, 5% CO2 in 200 uL complete R P M I 1640 culture medium in the absence or  presence o f TSST-1 or S E A . A t different time points, cells were  harvested, washed and stained with a n t i - C D 4 antibody. Then, flow cytometry analysis was performed to determine the proliferation o f C D 4 + T cells. Degree o f proliferation of C D 4 + T cells was calculated as follows: [numbers o f proliferated C F S E and C D 4 positive cells (low C F S E staining cells)] / [total C F S E and C D 4 positive cells] x 100%. To determine the effects o f superantigen-primed C D 4 + T cells on the proliferation o f naive splenocytes, 4 1 0 / w e l l CFSE-labeled nai've splenocytes were co-cultured with x  5  titrated numbers o f T S S T - 1 , S E A or PBS-primed C D 4 + T cells i n flat-bottom, 96-well plates in the presence o f TSST-1 or S E A (1 n M ) as described i n the co-culture  61  experiments. 72 hours later, cells were harvested and washed twice with P B S followed by flow cytometry analysis o f proliferation o f the naive splenocytes. Degree o f proliferation o f naive splenocytes was calculated as follows: [numbers o f proliferated C F S E positive naive splenocytes (low C F S E staining cells)] / [total C F S E positive naive splenocytes] x 100%.  3.6. Assay o f cell death To assay cell death, cells were suspended i n P B S and stained with 7 - A A D (Sigma, St Louis, M O ) at the concentration o f 4ug/mL for 10 minutes at 4°C. Cells then were washed twice with P B S and were analyzed by flow cytometry.  3.7. Assay o f cytokine production Supernatants from in vitro cell cultures were tested for the presence o f IL-2, I L 12 (p70) and IL-10 by specific E L I S A , according to instructions from the manufacturer ( B D Pharmingen). T N F - a and IFN-y were measured by E L I S A also according to instructions from the manufacturer (eBioscience, San Diego, C A ) .  3.8. Lethal shock induction A two-hit lethal shock model was established based on a previous published study (Visvanathan et al., 2001). Each mouse was injected intraperitoneally with 20mg of D-galactosamine (Sigma-Aldrich, St Louis, M O ) , 0.01 pg o f endotoxin (from E . coli 055:B5, Sigma-Aldrich, St Louis, M O ) and 4 pg o f TSST-1 or S E A i n 0.2 m l o f P B S . The survival o f mice was then monitored for 72 hours post-injection. In initial studies,  62  we confirmed that this regimen could induce 100% lethality in na'ive mice within 24 hours post-stimulation.  3.9. Antibodies and flow cytometry The following conjugated  monoclonal antibodies were obtained from B D  Pharmingen (San Diego, C A ) : F I T C or PE-anti-CD4, P E or APC-anti-IL-10, F I T C or PE-anti-IFN-y, PE-anti-IL-4, PE-anti-TNF-a and P E - a n t i - C T L A - 4 . APC-anti-IL-2 and A P C - a n t i - C D 2 5 were purchased from Biolegend (San Diego, C A ) . FITC-anti-Foxp3 and biotin-anti-GITR antibody were purchased from eBioscience (San Diego, C A ) . For surface staining, cells were washed with P B S supplemented with 1% F B S and stained for 20 m i n at 4°C with optimal dilutions o f each antibody, washed again, and analyzed. For intracellular staining o f cytokines, Golgistop ( B D Pharmingen, San Diego, C A ) was added to the cell culture for the last 5 hours. Then, cells were harvested, washed, stained with anti-CD4 antibody and fixed with cytofix ( B D Pharmingen, San Diego, C A ) . The fixed cells were further peameabilized with Permwash ( B D Pharmingen, San Diego, C A ) and stained with various antibodies at 4°C for 30min. After staining, cells were washed and analyzed. Foxp3 and C T L A - 4 were stained using the intracellular staining methods mentioned above. For apoptosis detection, cells were incubated with 4 ug/mL o f 7 - A A D (Sigma-Aldrich, St Louis, M O ) for 20 m i n at 4°C. T o assay the T cell or antigen presenting cell chimerism o f B 6 D 2 F 1 that received B 6 splenocytes, cells were washed with P B S supplemented with 1% F B S and stained for 20 m i n at 4°C with optimal dilutions o f PE-anti-H-2D ( B D Pharmingen, San Diego, C A ) and A P C b  anti-mouse C D 4 , APC-anti-mouse C D 8 or APC-anti-mouse M H C class II antibodies.  63  Flow  cytometry analysis was performed  on a F A C S C a l i b u r  machine (Becton  Dickinson, San Jose, C A ) and data were analyzed using W i n M D I 2.8.  3.10. V p analysis Splenocytes were first stimulated with I n M o f TSST-1 or S E A for 3 days, and then they were washed with R P M I 1640 medium twice to remove the superantigens. Cells were then resuspended in complete R P M I 1640 medium at 37 °C, 5%  CO2  overnight for the reexpression o f T cell receptor (Makida et al., 1996). The proportion o f cells expressing V p 3, 8, 10, 11, 12 or 15 that reacted to either TSST-1 or S E A stimulation were analysed. Although V P 17 had been reported to respond to both TSST-1 and S E A in some mouse strains (Marrack and Kappler, 1990), it has been shown that C57B1/6 mice do not express functional V P 17 (Wade et al., 1988), and hence V P 17 expression was not analysed. Cells were stained with anti-CD4 antibody ( B D Pharmingen) and antibody against V p 3 ( B D Pharmingen, San Diego, C A ) , 8, 10, 11, or 12 (Cedarlane, Hornby, O N ) and analyzed by flow cytometry. Since there was no commercially available antibody against V p i 5 , R T - P C R was used to analyze the V P 1 5 expression. P B S , T S S T - 1 , or SEA-primed C D 4 + T cells were first stimulated by TSST-1 or S E A for three days, then they were isolated magnetically as mentioned above. Total R N A was then isolated from 1><10 superantigen-stimulated C D 4 + T cells 7  using the RNeasy Total R N A isolation kit (Qiagen Inc., Mississauga, Ontario) according to the manufacturer's  direction. R N A was reverse  transcribed using  Omniscript reverse transcriptase (Qiagen Inc., Mississauga, Ontario). V p i 5 specific primers were designed using PrimerQuest. software (http://www.idtdna.com/UBC-  64  NAPS/Login.aspx) and sequences accessed from the Nucleotide database at N C B I (Accession numbers A E 0 0 6 6 3 , A E 0 0 6 6 4 and AE00665). Primers were synthesized by Integrated D N A Technologies Inc. (Coralville, I A ) and purified by standard desalting. Primer  sequence  for  V(315  was:  AGCTTGGTATCGTCAATCGCCTCA;  forward reverse  primer primer  (5' (5'  to  3'):  to  3'):  A C T G T C A G C T T T G A G C C T T C A C C A . 4 u l (~118ng) o f c D N A was added to each lOOul P C R reaction containing 0.2 u M o f each forward and reverse primer, 10X Buffer, 10 m M dNTPs and 2.5 U o f Taq D N A polymerase (Qiagen Inc., Mississauga, Ontario). Samples were mixed on ice and a simple hot start was provided by an initial denaturation at 94°C for 3 minutes. This was followed by 4 cycles o f 1 minute each at 94°C, 55°C and 72°C, with a final extension o f 10 minutes at 72°C and then rest at 4 ° C . 22.5ul o f P C R product, along with 2.5 ul o f 10X gel loading solution (Ambion Inc., Austin, T X ) was run on a 2% agarose gel containing ethidium bromide at 100 V , along with a 100 bp D N A ladder (Invitrogen Corporation, Burlington, Ontario) to assess the presence o f various V(3 segments. Gels were viewed using AlphaEase F C system (Alpha Innotech) and images were saved and the molecular weight o f each band was determined using the same software. The 1-D M u l t i line densitometry analysis tool was used to measure the integrated area o f each band, which is representative o f the intensity o f each band.  65  3.11. Reisolation o f naive splenocytes co-cultured with T S S T - l - i n d u c e d Tregs and secondary co-culture To determine i f naive splenocytes would acquire regulatory functions after coculture with T S S T - l - i n d u c e d Tregs, T S S T - l - i n d u c e d Tregs were first stained with C F S E as mentioned above. Then l x l 0 / w e l l na'ive splenocytes were co-cultured with 7  the same number o f CFSE-labeled T S S T - l - i n d u c e d Tregs i n 6-well plates in 6 m L o f complete R P M I 1640 medium. In the control group, na'ive cells were co-cultured with CFSE-labeled PBS-primed CD4+ T cells. Cells were then cultured at 37°C, 5% C 0 in 2  the presence o f TSST-1 (1 n M ) . 40 hours later, cells were recovered and dead cells were removed by gradient centrifugation. The CFSE-negative naive splenocytes were then reisolated by F A C S . N e w na'ive splenocytes ( 2 x l 0 / w e l l ) were then co-cultured 5  with re-isolated splenocytes at the ratio o f 1:1 or 1:2 in 96-well flat-bottom plates in the presence o f TSST-1 and cultured at 37°C, 5% C 0  2  i n 200 u L complete R P M I 1640  culture medium. Concurrently, new na'ive splenocytes and re-isolated splenocytes were cultured alone i n the presence o f T S S T - 1 . 48 hours later, supernatant was collected and stored at -70 °C until assay by E L I S A .  3.12. Culture o f na'ive splenocytes with T S S T - l - i n d u c e d Tregs conditioned medium 2 x l 0 / w e l l T S S T - l - p r i m e d splenocytes were plated i n 24-well plates in l m L o f 6  complete R P M I 1 6 4 0 medium and cultured at 37°C, 5% C 0 i n the presence o f TSST-1 2  (1 n M ) for 48 hours. To imitate the na'ive splenocyte co-cultured with Tregs at high ratio (5:4), 1 . 2 x l 0 magnetically isolated T S S T - l - i n d u c e d C D 4 + Tregs were added to 6  2xl0  6  /well T S S T - l - p r i m e d splenocytes and cells were cultured under the same  66  condition for 48 hours. Then the supernatant was recovered and residual cells were removed by centrifugation. 200 ul o f supernatant was then used to culture 4><10 /well 5  naive splenocytes in 96-well flat-bottom plates for 48 hours, and supernatant was collected and stored at -70 °C until assay by E L I S A . TSST-1 was added to final concentration o f l n M .  3.13. Culture o f T S S T - l - p r i m e d splenocytes with naive splenocyte conditioned medium 2 x l 0 / w e l l naive splenocytes were plated i n 24-well plates in l m L o f complete 6  R P M I 1640 medium and cultured at 37°C, 5% C 0 i n the presence o f TSST-1 ( l n M ) 2  for 48 hours. Supernatant was collected and centrifuged to remove residual cells. Concurrently, TSST-1 or PBS-primed splenocytes were prepared. 4 x l 0 / well TSST-1 5  or PBS-primed splenocytes were cultured in the naive splenocytes conditioned medium for 48 hours, and supernatant was collected and stored at -70°C until assay by E L I S A . In some groups, anti-CD25 antibody and control antibody (Biolegend, San Diego, C A ) were added to the culture at the concentration o f 10 ug/mL.  3.14. Neutralization o f IL-10 Blockade o f IL-10 was achieved by the anti-IL-10 receptor monoclonal antibody (clone 1B1.2). This antibody has previously been shown to be able to neutralize actions o f IL-10. Together with its control antibody (clone GL113), it is a generous gift from Dr. A l i c e M u i (Department o f Surgery, University o f British Columbia) (O'Farrell A M et a l , 1998). For in vitro studies, anti-IL-10 receptor  67  antibody was added to the culture at the concentration up to 100 pg/mL. The control antibody was added at the same concentration. For in vivo blockade, l m g o f antibody was given intraperitoneally at day 0 and 0.5 mg o f antibody was given once a week 7 days after the first dose (Kingsley et al., 2002).  3.15. Tumor cells P815  cells were purchased from A T C C . They were frozen i n liquid nitrogen  upon being received. Three days before infusion, they were thawed and the cells were cultured  in Dulbecco's modified  Eagle's medium containing  1.5  g/L sodium  bicarbonate and 4.5 g/L glucose (Stemcell Technologies, Vancouver, B C ) , 10% F B S (Hyclone, Logan, U T ) at the density o f 1><10 / m L as recommended i n 2 5 m L tissue 5  culture plates. After three days, cells reached confluence and they were detached by vigorous pipetting. They were then washed with cold D M E M medium and mixed with B 6 or F l splenocytes to reach the dose o f 10,000 cells/mouse i n 0.2 m L R P M I 1640 medium.  3.16. M i x e d lymphocyte reaction For mixed lymphocyte reaction, recipients o f B 6 splenocytes were sacrificed at different time points after transplantation. Total splenic T cells were isolated. 2><10  5  /well splenic T cells were co-cultured with 2><10 /well mitomycin C-treated B 6 or F l 5  splenocytes. Cells were cultured at 37°C, 5% C 0 2  in 200 u L R P M I 1640 medium  supplemented with 10% F B S (Hyclone, Logan, U T ) , 50 m M 2-mercaptoethanol ( 2 - M E ; Sigma, St Louis, M O ) , 10 m M H E P E S buffer, I m M sodium pyruvate, antibiotics [(100  68  U / m L penicillin; 100 m g / m L streptomycin (Stemcell Technologies, Vancouver, B C ) ; and 2 u.g/mL polymycin B (Sigma, St Louis, M O ) ) . Supernatant was collected after 5 days o f co-culture and was stored at -70°C until assay by E L I S A .  3.17. a G V H D induction and monitoring a G V H D was induced as reported (Owens, Jr. and Santos, 1968). Briefly, recipient  B6D2F1  mice  were  conditioned  by  intraperitoneal  injection  of  cyclophosphamide (Sigma-Aldrich, St Louis, M O ) at the dose o f 300 mg/kg. 24 hours later, each recipient mouse was injected with 4 1 0 B 6 splenocytes v i a the lateral tail X  7  vein. The survival o f recipient mice was monitored and mice were weighed once a week. In some experiments, TSST-1 was given to recipient mice at the dose o f 4 ug/mouse/injection in 0.2 m L P B S subcutaneously. When two doses o f TSST-1 were given, the interval was 4 days. M i c e i n the control group were given P B S . In tumor challenge experiments, mice dying from tumor progression were determined by the lack o f a G V H D clinical signs o f a G V H D and the presence o f liver and spleen enlargement with obvious tumor nodules upon autopsy.  3.18. Assay o f endotoxin Serum endotoxin level was determined by a chromogenic limulus amebocyte lysate endpoint assay using a kit purchased from Cambrex (Walkersiville, M D ) according to the instructions from the manufacturer.  69  3.19. Histological studies Tissues harvested from  different groups o f mice were fixed with P B S  containing 10% formaldehyde and embedded i n paraffin. 5 um sections were then cut and stained with hematoxylin and eosin for histological examination. Sections were prepared by Waxit Inc. (University o f British Columbia, Vancouver, B C )  3.20. Statistical analysis Descriptive data were expressed as [mean ± standard error o f mean] and analyzed with Prism 4.0. P values<0.05 were considered statistically significant. Group comparisons were made by two-tailed Student's t test. Kaplan-Meier survival analysis was used to determine the differences o f survival i n groups o f mice.  70  Chapter 4: Generation and characterization of regulatory T cells induced by repeated injection of toxic shock syndrome toxin-1 4.1. Introduction Staphylococcal exotoxins, including TSST-1 and staphylococcal enterotoxin A , B , C l , C 2 , C 3 , D , E , G , H , and I, are called superantigens because they are able to crosslink the M H C class II molecules on the A P C s and T cell receptors bearing specific VP segments without the need o f prior antigen processing, thus activating large numbers o f T cells (Marrack and Kappler, 1990). The activated T cells then respond by potent proliferation and massive production o f proinflammatory cytokines, such as I L 1, IL-2, IFN-y, IL-12 and T N F - a ( K u m et al., 2001). S E A and S E B are among the most studied bacterial superantigens. Earlier studies using these two superantigens demonstrated that exposure o f T cells to bacterial superantigens, both in vivo and in vitro could result i n the clonal expansion o f the responding T cells bearing specific VP segments followed by the selective clonal deletion o f a large fraction o f the responding T cells mediated b y apoptosis. The remaining T cells showed an anergic state i n response to superantigen stimulation (Kawabe and Ochi, 1991; MacDonald et al., 1991; Rellahan et al., 1990; White et a l , 1989). Later studies further showed that repeated stimulation with these bacterial superantigens was also found to induce the toxin reactive T cells to become Tregs, as shown by the changed cytokine production profile, proliferative state, and the active suppressive effects on the responses o f naive cells to the superantigen stimulation (Feunou et al., 2003; Grossman et a l , 2004; N o e l et a l , 2001). For example, T cells from  mice  injected  repeatedly  with  staphylococcal  enterotoxin  A ( S E A ) or  71  staphylococcal enterotoxin B ( S E B ) produced lower levels o f I L - 2 and IFN-y, but increased level o f IL-10 (Noel et al., 2001). When co-cultured with naive splenocytes in the presence  o f superantigen  stimulation, these superantigen-primed  T cells  suppressed the ability o f naive splenocyte to proliferate and to produce I L - 2 and IFN-y (Feunou et a l , 2003). TSST-1  is another  well-characterized bacterial superantigen.  Despite  the  structural differences with S E A or S E B , previous reported studies indicated that its biological activities mostly overlap with those o f S E A or S E B , in terms o f its ability to induce the cytokine and proliferation responses in T cells, and also the ability to induce lethal shock in mice. However, recent data i n our laboratory suggest the presence o f important  differences  i n biological  activities between  them. For instance,  our  laboratory showed that TSST-1 did not induce apoptosis o f human P B M C at low concentration as compared to S E B (Huang, 2004). Furthermore, until now, most o f the data regarding superantigen-induced Tregs were accumulated using S E A or S E B , while the characteristics o f T S S T - l - i n d u c e d Tregs are less clear. Therefore, the differences between Tregs induced by TSST-1 or other bacterial superantigens are also unknown. Thus  to  gain further  understanding  of TSST-l-induced  Tregs,  a  side-by-side  comparative study o f Tregs induced by TSST-1 and other superantigens is warranted. In this study, the biologic properties o f TSST-1 vs SEA-induced Tregs i n C 5 7 B L / 6 mice are directly compared.  72  4.2. Experimental Design M i c e were injected subcutaneously with TSST-1 or S E A at the dose o f 4 p.g/mouse /injection at the interval o f 4 days. M i c e in the control group were injected with P B S . T w o hours after the 3  rd  injection, splenocytes were isolated from these mice.  Cell surface or intracellular markers were analyzed by flow cytometry. Splenocytes were stimulated with TSST-1 or S E A and the cytokine, proliferation and apoptosis profiles were then analyzed by E L I S A or flow cytometry ( F C M ) . To assay the suppressive functions, C D 4 + T cells were further enriched from the splenocytes. They were then co-cultured with naive splenocytes in the presence o f TSST-1 or S E A , and their effects on cytokine and proliferation responses were determined b y E L I S A and F C M respectively. To investigate the in vivo suppressive functions, C D 4 + T cells were adoptively transferred to naive mice which then were subjected to lethal shock induction.  Survival o f the recipients was monitored. A flow  diagram o f the  experimental approach is depicted below. T S S T - 1 , x3 Splenocytes  *  analysis by FCM  S E A , x3  PBS, 3 x  >>  C D 4 + T  ce  n  s  by E L I S A and / I f FCM / Effects on lethal Cytokine responses Proliferation shock assayed by E L I S A responses assayed by F C M V  Figure 4.1. Experimental design o f studies described i n Chapter 4.  73  4.3. Results 4.3.1. Generation o f superantigen-induced Tregs i n C 5 7 B L / 6 mice Previous studies i n our laboratory have established that repeated injection o f TSST-1 in B A L B / c (H-2 ) mice at the dose o f 4pg/mouse/injection at the interval o f 4 d  days induced C D 4 + Tregs (Cameron, 2004). However, i n the present study, which w i l l also examine the effects o f Tregs in an a G V H D model, C 5 7 B L / 6 (B6, H - 2 ) mice were b  used as donor m i c e from w h i c h Tregs were derived. Thus i n i t i a l studies were performed to confirm that the same protocol that generated C D 4 + Tregs in B A L B / c mice can also do so in B 6 mice despite differences i n genetic background between these two strains o f mice. To generate TSST-1 or SEA-induced Tregs in B 6 mice, TSST-1 or S E A was administrated subcutaneously for three times at the interval o f 4 days at the dose o f 4pg/mouse/injection in 0.2mL P B S . In the control group, the same v o l u m e o f P B S was given i n the same way. T w o hours after the 3  r d  injection,  splenocytes from mice treated with T S S T - 1 , S E A or P B S were isolated and CD4+ T cells were further enriched by magnetic separation. Co-culture experiments were then performed to assay the suppressive activities o f these superantigen-primed C D 4 + T cells. Titrated numbers o f superantigen-primed C D 4 + T cells were co-cultured with fixed number o f na'ive splenocytes in the presence o f TSST-1 or S E A for 48 hours and then the supernatant was assayed for levels o f various cytokines by E L I S A . A s shown in Figure 4.2, the addition o f T S S T - l - p r i m e d C D 4 + T cells to naive splenocytes in the presence o f TSST-1 led to decreased levels o f IL-2, IL-12 and IFN-y. Similarly, coculture o f SEA-primed C D 4 + T cells with na'ive splenocytes in the presence o f S E A  74  Stimulation with TSST-1  Stimulation with SEA 150-1  20(H  IL-2  IL-2 100H lOOH 50H  O-l  —i  150-1 0  i  5:0 5:0.06 5:0.25 5:1  A— 5:4  5:2  o  "s-0 5:0'.06 5:0'.25 5M~  5:2  5:4  5:2  5:4  s  o  ** 100' c  o o  5o^ c (LI  o s-  0  4)  OH  ~sTo  210-1  5:0'.06 5:0'.25 iTl  5:2  i  5:4  i  i  5:0 5:0.06 5:0.25 5:1 250n IFN-y  IFN-y  140H  70H  —I—  —r-  5:0.06 5:0.25 5:1 5:2 5:4 5:0 5:0.06 5:0.25 5:0 • Ratio of naive splenocytes : CD4+ T cells  Figure induced  4.2.  Initial  Tregs  studies  in vitro.  to  Titrated  repeated injection o f T S S T - 1 splenocytes was  assayed  cells  S E A - p r i m e d were • :  i n the presence for various  down-regulated C D 4 +  expressed  Naive  primed  or S E Awere o f T S S T - 1  cytokines  T  were  Naive  cells;  from 2 independent  suppressive  o f C D 4 +  • :  +  T  from  co-cultured with a  b y E L I S A .  A s  hours,  shown,  levels  able to d o w n - r e g u l a t e  o f  superantigen-  mice  fixed  treated  number o f  and then  T S S T - l - p r i m e d  induced  b y  +  P B S - p r i m e d  C D 4 + C D 4 +  with naive  supernatant C D 4 +  T S S T - 1 ,  T  while  those induced b y S E A . Data  o f cytokine production b y naive splenocytes  T S S T - l - p r i m e d  Naive  activities  cells  or S E A for 48  proinflammatory cytokine  T cells  A :  the  numbers  as the p e r c e n t a g e  alone;  C D 4 +  determine  5:4  5:2  5:1  T  cells; T  cells.  T:  Naive  Data  alone.  + S E A -  accumulated  experiments.  7 5  resulted in decreased levels o f IL-2, IL-12 and IFN-y. Thus these data confirmed that our protocol was able to generate superantigen-induced Tregs i n B 6 mice and that the generated Tregs could down-regulate responses o f naive splenocytes to their cognate antigens. Moreover, these initial studies also set up a co-culture system to assay and investigate the mechanisms o f the suppressive functions o f superantigen-induced Tregs described in Chapter 5. Following these experiments, a comparative study investigating the cytokine profile, proliferative ability, apoptosis, and suppressive activities o f T S S T 1 vs SEA-induced Tregs was then performed.  4.3.2. Cytokine production profile o f TSST-1 and SEA-induced Tregs A s Tregs are associated with distinct cytokine production profiles, the cytokine production profile o f TSST-1 or SEA-induced Tregs were compared. To study their cytokine profile, splenocytes from mice treated with repeated injection o f TSST-1 or S E A were stimulated with TSST-1 or S E A for various periods o f time, then E L I S A was performed to quantify the cytokine levels i n the supernatant and flow cytometry analysis was performed to further determine the intracellular cytokine expression by C D 4 + T cells. i) PBS-primed CD4+ T cells: PBS-primed splenocytes stimulated with TSST-1 or S E A produced mainly proinflammatory cytokines, such as IL-2, IFN-y and T N F - a , while the level o f anti-proinflammatory cytokine IL-10 was low. Their production was timedependent, as the levels o f these cytokines increased as the culture time increased (Figure 4.3). F l o w cytometry analysis confirmed the expression o f these cytokines at 40 hours post-stimulation. A s shown i n Figure 4.4, no cytokine positive cell population  76  2500 IL-2 2000 1500 1000 500 0  -^S-Tregs+T -¥-S-Tregs+S  12h 24h 48h 72h 96h  IFN-y  bt) C  12h 24h 48h 72h 96h  e  10007505002500-_  12h 24h 48h 72h 96h  12h 24h 48h 72h 96h 1250-1  1250-1  10007505002500L_^ .  .  .  .  12h 24h 48h 72h 96h  12h 24h 48h 72h 96h 400-1  '& SJ  u e u C  o U 3000 IL-10  2 4 0 0  1800 1200 600 0  12h 24h 48h 72h 96h 12h 24h 48h 72h 96h 12h 24h 48h 72h 96h 3000 3000-1 2400 24001800 18001200-1 1200600 6000 012h 24h 48h 72h 96h 12h 24h 48h 72h 96h 12h 24h 48h 72h 96h  Figure 4.3. Cytokine levels in supernatant of TSST-1 or SEA-induced Tregs following reactivation with TSST-1 or SEA. Splenocytes from mice treated with TSST-1, SEA or PBS were stimulated with TSST-1 or SEA for various periods of time, then supernatant was harvested and assayed by ELISA for IL-2, IFN-y, IL-10, and TNF-a. T-Tregs: TSST-l-induced Tregs; S-Tregs: SEA-induced Tregs; P-CD4: PBS-primed CD4 T cells; T: TSST-1; S: SEA. Data were accumulated from four independent experiments, n=4.  77  P-CD4+R  P-CD4+T  P-CD4+S  3%  14.5%  23.3%  mm  o o  & 10°  10'  1  10"  10'  io  io  1  3  T-Tregs+T  jfc.;20.1°/o  22.6%  IK  D  ID  1  T-Tregs+R  Q  io  10  io'  io  1  io  3  10°  10*  io  1  io  0'  o*  0°  io  1  io  3  T-Tregs+S  3  S-Tregs+R  S-Tregs+T  S-Tregs+S  8.4%  32%  24.4%  IP'*  IL-2 io°  io'  Ho  1  IL-2-APC  io  1  1D  D  Figure 4.4. Expression o f cytokines by TSST-1 or SEA-induced Tregs. A . Representative raw data o f expression o f IL-2 by TSST-1 or SEA-induced Tregs assayed by flow cytometry. Splenocytes from mice treated with T S S T - 1 , S E A or P B S were cultured i n the absence or presence o f TSST-1 or S E A for 45 hours, then intracellular staining o f I L - 2 followed by flow cytometry analysis was performed. Data shown was representative o f flow cytometry analysis o f I L - 2 expression by TSST-1 or SEA-induced Tregs, or PBS-primed C D 4 + T cells from one o f four independent experiments, n=4. T-Tregs: T S S T - l - i n d u c e d Tregs; S-Tregs: S E A induced Tregs; P - C D 4 : PBS-primed C D 4 T cells; T: T S S T - 1 ; S: S E A ; R : R P M I 1640 medium.  78  45n  B  IL-2  X  3(H  15H  RPMI TSST-1 5(H  SEA  IFN-Y  404  X  £ 1 30  =  +  U ©  a  o fl 3  o  60-  RPMI  SEA  RPMI  40-  RPMI TSST-1  SEA  RPMI  SEA  X  TSST-1  SEA  X  TL  TSST-1  SEA  2<H  RPMI  TSST-1  SEA  RPMI  TSST-1  35n  u PM  TSST-1  IL-10  1  s  X  SEA  2010'-1  Tift  RPMI TSST-1  284  TNF-a  X  2H  X  14H  RPMI  TSST-1  SEA  PBS-CD4 T  RPMI TSST-1  SEA  TSST-l-Tregs  RPMI TSST-1  SEA  SEA-Tregs  Figure 4.4. B . Data presented were aggregated from four independent experiments, n=4. The bar graphsrepresented the percent o f cytokine positive C D 4 + T cells. *, P O . 0 5 , compared with SEA-Tregs+R, t-test.  79  was detected in PBS-primed C D 4 + T cells cultured in the absence o f stimulation while cytokine positive populations were detected in TSST-1 or SEA-stimulated PBS-primed C D 4 + T cells, showing that the cytokine expression by PBS-primed C D 4 + T cells depended on the induction by superantigen stimulation. ii) TSST-l-induced Tregs: In response to TSST-1 stimulation, I L - 2 was only detected at the 12 hour time point, but undetectable at other time points. IFN-y and T N F - a were detected in the supernatant at levels comparable to those i n PBS-primed C D 4 + T cells stimulated with T S S T - 1 , although the levels were lower at later time points. IL-10 production was greatly enhanced at all time points. When T S S T - l - i n d u c e d Tregs were stimulated with S E A , IL-2 was not detectable at any time point while IFN-y, T N F - a and IL-10 were detected, although their levels were generally lower than those i n the supernatant o f T S S T - l - i n d u c e d Tregs stimulated with T S S T - 1 (Figure 4.3). Unlike the PBS-primed C D 4 + T cells and SEA-induced Tregs, whose cytokine expression was triggered by superantigen stimulation, flow cytometry analysis found that T S S T - l induced Tregs expressed all cytokines assayed even i n the absence o f superantigen stimulation. A significantly higher percent o f cytokine positive C D 4 + T cells was found in T S S T - l - i n d u c e d Tregs than i n SEA-induced Tregs i n the absence o f stimulation. When T S S T - l - i n d u c e d Tregs were stimulated with TSST-1 or S E A , the percentage o f cytokine positive CD4+ T cells was only slightly increased as compared to that in the absence o f stimulation (Figure 4.4). iii) SEA-induced Tregs: Levels o f I L - 2 , IFN-y and T N F - a greatly decreased in the supernatant o f SEA-induced Tregs stimulated with S E A as compared to PBS-primed C D 4 + T cells stimulated with S E A , while the level o f IL-10 was elevated. However,  80  when SEA-induced Tregs were stimulated with T S S T - 1 , the cytokine production profile was rather similar to that o f the PBS-primed C D 4 + T cells stimulated with S E A , which was high I L - 2 , IFN-y and T N F - a and l o w I L - 1 0 (Figure 4.3). Interestingly, despite the low levels o f IL-2, IFN-y and T N F - a in the supernatant o f SEA-induced Tregs stimulated with S E A , flow cytometry analysis showed that these cytokines were actually expressed by SEA-induced Tregs, as cytokine positive C D 4 + T cells were readily detectable in SEA-induced Tregs. Similar to PBS-primed C D 4 + T cells, S E A induced Tregs cells did not show marked expression o f these cytokines i n the absence of superantigen stimulation, although the percent o f cytokine positive SEA-induced Tregs i n the absence o f stimulation was slightly increased as compared to PBS-primed C D 4 + T cells (Figure 4.4). These data show that cytokine expression by SEA-induced Tregs was also dependent on superantigen stimulation.  4.3.3. Proliferation profile o f TSST-1 or SEA-induced Tregs It has been shown that superantigen stimulation resulted i n T cell anergy, which was demonstrated by reduced incorporation o f [ H]thymidine. Moreover, an anergic 3  state has been shown to be a feature o f Tregs. To study the proliferation profile o f superantigen-primed C D 4 + T cells, splenocytes from mice treated with repeated injection o f TSST-1 or S E A were labeled with C F S E and cultured i n the absence or presence o f TSST-1 or S E A . The proliferation o f C D 4 + T cells was determined by flow cytometry analysis at different time points. A s shown in Figure 4.5, PBS-primed C D 4 + T cells and SEA-induced Tregs did not proliferate i n the absence o f superantigen  81  A  P-CD4+S  P-CD4+T  P-CD4+R  lif  S<5  1.2% in1  10°  in  in  1  46.6%  35% in  3  0  in'  n  '1 n  3  n*  3  o"  io'  o  1  io  3  T-Tregs+T  T-Tregs+R  u 64% 10  D  ID'  10  3  I O 10  S-Tregs+R  7.9% 10°  3  10'  10°  10'  10  3  10  3  1 0'  S-Tregs+S  54.4%  io'  ^  CD4  Figure 4.5. Proliferation response o f superantigen-induced Tregs. A . Representative raw data o f proliferation response o f superantigen-induced Tregs. Splenocytes from mice treated with T S S T - 1 , S E A or P B S were first stained with C F S E and then cultured in the absence or presence o f TSST-1 or S E A . A t different time points poststimulation, cells were harvested and analyzed by flow cytometry for the proliferation o f C D 4 + T cells. Representative data shown were proliferation of C D 4 + T cells stimulated in the absence or presence o f l n M superantigens for 3 days. T-Tregs: T S S T - l - i n d u c e d Tregs; S-Tregs: SEA-induced Tregs; P - C D 4 : PBS-primed C D 4 T cells; T: T S S T - 1 ; S: S E A ; R: R P M I 1640 medium. PBS-primed C D 4 + T cells cultured i n the absence o f superantigen (P-CD4+R) were used as control to determine the C F S E intensity o f non-proliferating cells.  82  B  Day 1  Day 2  Day 3  Day 5  Day 1  Day 2  Day 3  Day 5  Time of stimulation (day) Figure 4.5. B . Data presented are accumulated data from four independent experiments and show the proliferation of C D 4 + Tcells i n the absence of stimulation or in the presence of different concentration o f superantigen stimulation. • : T S S T - l - i n d u c e d Tregs+RPMI; • : SEA-induced Tregs+RPMI; 0: PBS-primed C D 4 T cells + R P M I ; A : TSST-l-induced Tregs+TSST-1; T : SEA-induced Tregs+SEA; o PBS-primed C D 4 T cells+TSST-1; • : PBS-primed C D 4 T cells+SEA. *, P O . 0 5 , compared with S-Tregs+S, t-test, n=4. :  83  stimulation and proliferation was detected when they were stimulated with TSST-1 or S E A , showing that the proliferation o f PBS-primed C D 4 + T cells and SEA-induced Tregs was dependent on superantigen stimulation. Interestingly, i n the absence o f superantigen  stimulation, TSST-l-induced  Tregs could undergo  proliferation as  compared to the control PBS-primed C D 4 + T cells and SEA-induced Tregs over the 5day period. The maximum degree o f proliferation o f T S S T - l - i n d u c e d Tregs in the absence o f TSST-1 was comparable to that in the presence o f TSST-1 at day 5. In the presence o f cognate antigen stimulation, both TSST-1 and SEA-primed C D 4 + T cells proliferated more potently than the control PBS-primed C D 4 + T cells. However, the proliferation o f T S S T - l - p r i m e d C D 4 + T cells at day 2 and day 3 was significantly higher than that o f SEA-induced Tregs when cells were stimulated with different doses o f cognate antigen. A t day 5, difference o f proliferation between these two types o f superantigen-primed CD4+ T cells was only observed at 0.1 n M .  Therefore, the  proliferation o f T S S T - l - i n d u c e d Tregs was antigen-independent, and i n the presence o f the cognate antigen stimulation, the proliferation response o f T S S T - l - i n d u c e d Tregs was more sensitive than that o f SEA-induced Tregs.  4.3.4. Other phenotypic properties o f TSST-1 and SEA-induced Tregs Previous studies showed that superantigen stimulation o f T cells might lead to deletion o f superantigen-reactive T cells v i a cell death. Thus the cell death o f TSST-1 and SEA-induced Tregs in response to their cognate antigen stimulation in vitro was also investigated. A s showed in Figure 4.6, in the absence o f antigen stimulation, P B S primed C D 4 + T cells and SEA-induced Tregs underwent increased cell death as the  84  P-CD4+R  P-CD4+T  21.2%  P-CD4+S 12.3%:  : 14.4%  8^*v  io a  io'  ia 1  1D  io 3  D  >  .^  io  ia  3  i  3  T-Tregs+T  T-Tregs+R M  io'  23.7%  23%  |||t!l  Q % U  10°  in'  in 2  S-Tregs+R  in 3  1  S-Tregs+S  26.9%  19.7%  io°  -•  io  1  io 7-AAD z  io  3  7-AAD  Figure 4.6. C e l l death o f superantigen-induced Tregs in vitro. A . Representative raw data o f cell death o f superantigen-induced Tregs i n vitro. Splenocytes from mice treated with T S S T - 1 , S E A or P B S were cultured i n the absence or presence o f T S S T 1 or S E A . A t different time points post-stimulation, cells were harvested, stained with 7 - A A D and PE-anti-CD4 antibody and analyzed by flow cytometry for the cell death of C D 4 + T cells. Representative data shown were the cell death o f C D 4 + T cells stimulated in the absence or presence o f I n M superantigens after 1-day culture. TTregs: T S S T - l - i n d u c e d Tregs; S-Tregs: SEA-induced Tregs; P - C D 4 : PBS-primed C D 4 T cells; T: T S S T - 1 ; S: S E A ; R : R P M I 1640 medium.  85  B 75n  75' 6  RPMI  O.lnM 50H  s  50U  H  +  25H  25-  o u  +  Day 1  Day 3  Day 5  0  Day 1  75n  75n  I  8  50H  50H  25-  25H  PH  Dayl  Day 3  Day 5  Day 5  Day 3  Day 5  lOnM  lnM t--  Day 3  0'  Day 1  T i m e of stimulation (days) Figure 4.6. B . Data presented are accumulated data from four independent experiments showing the cell death of CD4+ Tcells i n the absence o f stimulation or i n the presence of different concentration o f superantigen stimulation. • : T S S T - l - i n d u c e d Tregs+RPMI; • : SEA-induced Tregs+RPMI; 0: PBS-primed C D 4 T cells + R P M I ; A: TSST-l-induced Tregs+TSST-1; T : SEA-induced Tregs+SEA; o PBS-primed C D 4 T cells+TSST-1; • : PBS-primed C D 4 T cells+SEA. *, P O . 0 5 , compared with S-Tregs+S, t-test, n=4. :  86  culture time increased. In contrast, T S S T - l - i n d u c e d Tregs did not show this time dependent cell death, suggesting that TSST-1 is anti-apoptotic or pro-survival. A significantly lower percent o f C D 4 + T cells undergoing cell death was detected i n T S S T - l - i n d u c e d Tregs at day 3 and day 5 than in SEA-induced Tregs. In the presence of l n M and 10 n M cognate superantigen stimulation, the survival o f TSST-1 and S E A induced Tregs was better than PBS-primed C D 4 + T cells as shown by the lower percent o f C D 4 + T cells undergoing cell death at day 3 and day 5. N o significant difference in the percent o f C D 4 + T cells undergoing cell death was detected between TSST-1 and SEA-induced Tregs at any time point. In response to 0.1 n M superantigen stimulation, a significant difference was detected at day5 between TSST-1 and S E A induced Tregs. Thus T S S T - l - i n d u c e d Tregs showed a better ability for survival i n the absence o f antigen stimulation or in the presence o f low concentration o f antigen stimulation. A s the expression o f Foxp3, C T L A - 4 , C D 2 5 or G I T R has been associated with the Treg phenotype, the expression o f these molecules on TSST-1 or SEA-induced Tregs was investigated. Results showed that the percentages o f Foxp3 and G I T R positive C D 4 + T cells were not significantly different between T S S T - l - p r i m e d and SEA-primed C D 4 + T cells. In contrast, the percentages o f C D 2 5 and C T L A - 4 positive cells were significantly higher in T S S T - l - p r i m e d C D 4 + T cells than i n SEA-primed C D 4 + T cells (Figure 4.7 and Table 4.1).  The expression o f C D 2 5 or C T L A - 4 i n  superantigen-primed C D 4 + T cells was also higher than that i n the PBS-primed control C D 4 + T cells.  87  TSST-l-Tregs  U  IP"'-  PBS-CD4  SEA-Tregs  11.3%  12.7%  9.4%  3  I Foxp3 44.5%  13.1%  18.4% l"  s  CTLA-4 24%  17.8%  21.5%  Ml lift GITR 35%  11.3%  22%  CD25 ID"  i  Figure 4.7. Phenotype o f superantigen-induced Tregs. Representative raw data of phenotypical analysis o f superantigen-induced Tregs. F l o w cytometric analysis was performed to determine the expression o f Foxp3, C T L A - 4 , G I T R and C D 2 5 on splenic C D 4 + T cells from mice injected with 3 times o f TSST-1 or S E A . Representative data of four independent experiments were shown. Left column, T S S T - l - i n d u c e d Tregs. Middle column, SEA-induced Tregs. Right column, PBS-primed C D 4 + T cells.  Table 4.1. Expression o f Foxp3, C T L A - 4 , G I T R or C D 2 5 on C D 4 + T cells o f mice treated with T S S T - 1 , S E A or P B S . *, p < 0.05, compared with SEA-Tregs, t-test, n=4. TSST-l-Tregs  SEA-Tregs  PBS-CD4  8.8=1=1.0  13.0±1.8  8.4±1.3  49.4±3.4*  25.9±3.8  12.6±1.8  GITR  24.1±6.3  20.6±0.5  14±1.7  CD25  45.3±4.1*  22.2=bl  8.9±1  Foxp3 CTLA-4  88  4.3.5. T S S T - l - i n d u c e d Tregs had broader suppressive activities than SEA-induced Tregs Tregs are defined by their suppressive activities over the immune responses; thus, elucidation o f the differences between the suppressive activities o f Tregs induced by TSST-1 or S E A was pivotal for the comparison. In vitro studies were first performed to determine the suppressive actions o f superantigen-induced Tregs on cytokine production b y co-culture experiments. T o this end, a fixed number o f naive splenocytes were co-cultured with a titrated number o f purified C D 4 + TSST-1 or S E A induced Tregs i n the presence o f TSST-1 or S E A for 48 hours and cytokine levels i n the supernatant were assayed b y E L I S A . A s shown i n Figure 4.8-A, the addition o f T S S T - l - i n d u c e d C D 4 + Tregs to na'ive splenocytes resulted i n decreased levels o f IL-2, IL-12 and IFN-y induced by either TSST-1 or S E A i n a dose-dependent manner. In contrast, the addition o f SEA-induced C D 4 + Tregs to the culture lead to decreased levels o f IL-2, IL-12 and LFN-y induced by S E A , but not by T S S T - 1 . Therefore, these in vitro data showed that the suppressive function o f T S S T - l - i n d u c e d C D 4 + Tregs was broader than that o f SEA-primedCD4+ T cells. To further confirm the in vitro findings, a lethal shock model was used to test the in vivo suppressive function o f TSST-1 or SEA-induced Tregs. C D 4 + Tregs from mice treated with TSST-1 or S E A were adoptively transferred to naive recipient mice and then they were subjected to lethal shock challenge b y TSST-1 or S E A . The results showed that mice without transfer o f T cells all died quickly after challenge (Figure 4.8-B, mean survival TSST-1:9.5±1.3 h, SEA:8±0.9 h , n=5). Survival o f recipient mice o f T S S T - l - i n d u c e d Tregs was significantly prolonged i n response to either T S S T -  89  Stimulation with SEA  Stimulation with TSST-1 200  1 IL-2  I50 IL-2 n  150 100  150n  IL-12  150  5:0.06 5:0.25 IL-12  o  u  c o u  1004  100  G a u sw  PH  — i —  5:0.06 5:0.25 5:1  5:2  5:4  IFN-y  5:0 5:0.06 5:0.25 5:1 5:0 5:0.06 5:0.25 5:1 5:2 5:4 Ratio of naive splenocytes : CD4+ T cells  —I—  5:2  —I—  5:4  Figure 4.8. Suppressive function o f superantigen-induced Tregs in vitro and in vivo. A : Suppressive activities o f superantigen-induced Tregs i n vitro. Titrated numbers of C D 4 + T cells from mice treated with repeated injection o f TSST-1 or S E A were cocultured with a fixed number o f naive splenocytes i n the presence o f TSST-1 or S E A for 48 hours, and then supernatant was assayed for various cytokines by E L I S A . A s shown, T S S T - l - i n d u c e d C D 4 + Tregs down-regulated proinflammatory cytokine levels induced b y TSST-1 or S E A , while SEA-induced C D 4 + Tregs were only able to down-regulate those induced b y S E A but not by T S S T - 1 . Data were expressed as the percentage o f cytokine production by naive splenocytes alone. • : Naive alone; • : Na'ive + T S S T - l - i n d u c e d Tregs; T : Na'ive + SEA-induced Tregs;*: Na'ive + PBS-primed C D 4 + T cells. *, P<0.05, compared with na'ive splenocytes alone, paired t-test, n=4.  90  Challenge with TSST-1  B  Navie Naive+T-Tregs Navie+S-Tregs Naive+T n o n - C D 4 Naive+P-CD4 T  0  S  Challenge with SEA ©  Naive  fl  Naive+T-Tregs  o u  Naive+S-Tregs Naive+S non-CD4  CL  Naive+P-CD4 T  0  10  20  30  Time post lethal shock induction (h) Figure 4.8. B . Suppressive effects o f TSST-1 or SEA-induced Tregs i n vivo. M i c e were first transferred with C D 4 + T cells (2><10 /mouse) isolated from mice treated with repeated injection o f T S S T - 1 , S E A or P B S . T w o hours after cell transfer, they were challenge with lethal shock induced by TSST-1 or S E A . Survival was monitored for 72 hours post-challenge. *, P O . 0 5 , compared with naive mice (no cell transfer), n=5. 7  91  c  Stimulation with TSST-1  45  0  T  Stimulation with SEA  • Naive •A- N a i v e + T - T r e g s  5:0  5:0.06 5:0.25  •  5:1  5:2  5:4  0  5:0  Naive  5:0.06 5:0.25  5:1  5:2  5:4  Ratio of naive splenocytes : CD4+ T cells  Figure 4.8. C . Effects o f superantigen-induced Tregs on the proliferation o f naive splenocytes. To study the effects o f superantigen-induced Tregs on the proliferation responses o f na'ive splenocytes, na'ive splenocytes were first stained with C F S E and then were co-cultured with titrated number o f magnetically isolated C D 4 + T cells from mice treated with three injections o f T S S T - 1 , S E A or P B S i n the presence of TSST-1 or S E A for 7 2 hours. Then the proliferation o f na'ive splenocytes was assayed by flow cytometry. A s shown, addition o f TSST-1 or SEA-induced CD4+ Tregs did not lead to decreased proliferation o f naive splenocytes, rather, they increased the proliferation o f na'ive splenocytes. Thus, although TSST-1 or S E A induced Tregs could down-regulate the cytokine responses o f na'ive splenocytes, they did not suppress the proliferation responses o f their target cells. Results shown were accumulated data from three independent experiments (n=3). Navie: naive splenocytes; T-Tregs: T S S T - l - i n d u c e d Tregs; S-Tregs: SEA-induced Tregs; P - C D 4 T: PBS-primed C D 4 + T cells.  92  1 (18±2.9 h) or S E A (16.5±2.7 h) challenge. In contrast, survival o f mice that received SEA-induced Tregs was prolonged only when challenged with S E A (13.9±2.5 h), but not with TSST-1 (9.4±0.8 h). Transfer o f C D 4 + T cell-depleted TSST-1 or S E A primed splenocytes, or PBS-primed C D 4 + T cells did not confer protection, further confirming the suppressive effect o f superantigen-induced C D 4 + Tregs. Interestingly, the addition o f TSST-1 or SEA-induced Tregs to naive splenocytes stimulated with TSST-1 or S E A did not lead to the inhibition o f naive splenocytes in vitro. Instead, proliferation o f naive splenocytes was increased i n the presence o f TSST-1 or SEA-induced Tregs i n a dose-dependent manner. This effect was also more remarkable in T S S T - l - i n d u c e d Tregs (Figure 4.8-C). In summary, the data above further  confirmed findings in the initial studies  showing that i n our  repeated  superantigen injection system, C D 4 + Tregs could be induced. Moreover, T S S T - l induced Tregs had broader suppressive abilities than SEA-induced Tregs.  4.3.6. VP usage by TSST-1 and S E A induced Tregs—Cross-reactive suppressive function o f T S S T - l - i n d u c e d Tregs was not due to shared VP families of T cells reactive to both TSST-1 and S E A Previous studies have reported that in mice TSST-1 can react with T cells bearing VP 15 while S E A can react with T cells bearing VP3, 10, 11, 12. A s shown above, since T S S T - l - i n u d c e d Tregs could suppress responses triggered by S E A , it was possible that T S S T - l - i n d u c e d Tregs contained Vp families o f T cells that were also reactive to S E A . To exclude this possibility, Tregs induced by TSST-1 or S E A were analyzed for the T cell receptor VP repertoire. To this end, splenocytes from mice  93  treated with TSST-1 or S E A were stimulated with their cognate antigen in vitro and flow cytometry was performed to detect the expansion of specific VP family o f T cells. A s shown in Figure 4.9 and Table 4.2, stimulation o f PBS-primed splenocytes with SEA  resulted in the  expansion of VP3  and V p i l  C D 4 + T cells which  was  demonstrated by the increased percent of V p 3 (40.4±3.4 vs 4.9±0.5) and VP 11 CD4+ T (25±3.4 vs 5.8±0.8) as compared to PBS-primed splenocytes in the absence o f toxin stimulation. However, stimulation o f PBS-primed splenocytes with TSST-1 resulted in the decreased percent of V P 3 positive CD4+ T cells which was due to the expansion o f other V P families, thus showing that V p 3 C D 4 + T cells did not react to T S S T - 1 . Moreover, we could not detect the expansion of V p l O and V p i 2 C D 4 + T cells as reported in PBS-primed splenocytes stimulated with S E A . A s expected, stimulation of SEA-induced Tregs led to the expansion of V p 3 and V p i l C D 4 + T cells. In contrast, stimulation  o f TSST-l-induced Tregs with the cognate antigen did not lead  to  expansion of any V P family o f CD4+ T cells that we analyzed. Since it was not possible to assay the V p l 5 expression due to lack of commercially available antibody against this segment, R T - P C R was used to determine the expression o f V p i 5 of T S S T 1 and SEA-induced Tregs. A s shown in Figure 4.10, stimulation o f PBS-primed CD4+ T cells or TSST-l-induced Tregs with TSST-1 led to increased expression of V p i 5 while stimulation o f these cells with S E A did not. Thus, TSST-1 reacted with V p i 5 family T cells. Taken together, the data showed that SEA-induced Tregs consisted o f VP3 and 11 families o f T cells while TSST-l-induced Tregs consisted o f V p i 5 family o f T cells. Therefore, the data here demonstrated that the suppression o f SEA-triggered  94  P-CD4+R  10°  10'  ID  1  10  P-CD4+T  A  10  8  10'  m  0.5%  IP  10"  I D 0  10'  10*  10'  T-Tregs+T  T-Tregs+R  U  P-CD4+S  10"  S-Tregs+R  io'  0.8%  io  2  io  3  S-Tregs+S  36.5%  20.8%  Vp3 io"  io"  io^  io*  io*  io°  io'  io'  itP  io*  Figure 4.9. VP analysis of superantigen-induced Tregs. Splenocytes from mice treated with T S S T - 1 , S E A or P B S was cultured in the absence or presence o f T S S T 1 or S E A for 3 days. Then cells were washed to remove the toxin and cultured in the absence o f toxin overnight followed by flow cytometry analysis o f VP family of the C D 4 + T cells. R a w data shown was one o f three experiments o f V p 3 analysis. T-Tregs: T S S T - l - i n d u c e d Tregs; S-Tregs: SEA-induced Tregs; P - C D 4 : P B S primed C D 4 T cells; T: T S S T - 1 ; S: S E A ; R : R P M I 1 6 4 0 medium.  95  PBS-CD4  Cells stimulant  RPMI  TSST-1  T S S T - -Tregs SEA  SEA-Tregs  RPMI  TSST-1  RPMI  SEA  Vp3  5.0±0.5  2.4±0.4  40.4±3.4  1.9±0.9  1.5=1=0.5  14.9±3.3  36.1=1=2.9  Vp8  19.1=1=1.3  10.6±2.8  5.0=1=1.2  4.6=1=1.5  4.3=1=1.2  19.8±2.8  12.9±1.3  vpio  5.3±0.3  3.4±0.6  3.9±0.3  1.1=1=0.1  1.5±0.1  3.1±0.3  2.6±0.1  vpn  5.8=1=0.8  3.9±1.1  25.1=1=3.4  1.7=1=0.5  2.0±0.3  8.8±0.5  12.9=1=1.6  Vpl2  11.1±2.9  4.8±2.2  7.3±1.5  2.5±0.4  2.6±0.3  9.7±2.6  6.9±1.8  Table 4.2. VP analysis o f superantigen-induced Tregs. Splenocytes from mice treated with T S S T - 1 , S E A or P B S were cultured i n the absence or presence of TSST-1 or S E A for 3 days. Cells were then washed to remove the toxin and cultured in the absence o f toxin overnight followed by flow cytometry analysis of VP family o f the C D 4 + T cells. The numbers shown represented the percent of specific VP positive cells i n C D 4 + T cells. Data shown were accumulated from three independent experiments, n=3.  96  25-i  .  j  es  2  15-  £ 09 2  io-  u  to  5-  o - L - ^  P-CD4+T  1  ' — i — '  1  P-CD4+S T-Tregs+T S-Tregs+S  Figure 4.10. Assay o f V p i 5 expression by R T - P C R . Total R N A was purified from TSST-1 or SEA-induced Tregs and PBS-primed C D 4 + T cells stimulated with either TSST-1 or S E A . R T - P C R then was performed to amplify the V p i 5 segment o f the T cell receptor. Constant region o f the T cell receptor served as a control. Product o f R T - P C R was then measured for it density. The density o f each group was then corrected by the density o f product o f constant region o f the T cell receptor. Density o f product o f stimulated C D 4 + T cells was then divided by that o f the unstimulated C D 4 + T cell to calculate the fold o f increase o f expression o f V p i 5 . A s shown, stimulation o f C D 4 + T cells with T S S T - 1 , but not S E A , led to increased expression of V p l 5 .  97  responses by TSST-l-induced Tregs was not because they contained CD4+ T cells reactive to S E A .  4.4. Discussion To characterize the TSST-l-induced Tregs, a side-by-side comparison o f Tregs induced b y two different bacterial superantigens, namely TSST-1 and S E A was performed. The results showed that repeated injection o f TSST-1 or S E A could induce CD4+ Tregs that were able to down-regulate the cytokine responses o f na'ive splenocytes in vitro and protected against lethal shock challenge in vivo. These data demonstrated that the suppressive activities o f SEA-induced Tregs were highly antigen specific, as they only suppressed the SEA-triggered responses in vitro and in vivo, but were unable to suppress the TSST-l-induced responses. Similar results were found i n a previous study, where repeated mucosal administration o f S E A was protective against subsequent SEA-mediated, but not TSST-l-mediated lethal shock challenge, although the authors did not determine i f repeated mucosal administration o f S E A was able to induce Tregs that were responsible for these effects (Collins et al., 2002). In contrast to SEA-induced Tregs, the suppressive functions o f T S S T - l - i n d u c e d Tregs were more broadly reactive as they suppressed responses triggered b y either TSST-1 or S E A . In a previous study, N o e l et al. showed that splenocytes o f C57BL/6 mice repeatedly injected with S E A or S E B had a similar cytokine profile o f high IL-10 and low IL-2 i n response to either S E A or S E B stimulation in vitro. This profile was not observed i n B A L B / c mice. They hypothesized that this was because S E A and S E B could both stimulate T cells bearing V p 3 , which B A L B / c mice lacked (Noel et al., 2001).  98  However, data o f V P screening by flow cytometry and R T - P C R did not reveal that TSST-1 or S E A could stimulate a common VP family o f T cells. In addition, i f there is a T cell VP family population reactive to both TSST-1 and S E A , the SEA-induced Tregs should also be activated by TSST-1 and suppressed the TSST-1-triggered responses, which was not observed. Therefore, the ability o f T S S T - l - i n d u c e d Tregs to suppress SEA-induced responses was unlikely to be the presence o f V p family o f T cells reactive to both TSST-1 and S E A . Thus, our data for the first time showed that repeated injection o f TSST-1 could generate Tregs with broader suppressive activities as compared to S E A . Interestingly, our data showed that neither TSST-1 nor SEA-induced Tregs suppressed the proliferation o f naive splenocytes, which was different from what was previously reported (Feunou et al., 2003; Grundstrom et al., 2003). The discrepancy may be due to the different experimental systems used. In Feunou's studies, Tregs were induced by three intraperitoneal injections o f S E B in B A L B / c mice while repeated intravenous injection o f S E A or S E B were administered to V P 3 or V p 8 transgenic mice to generate Tregs i n the other report. The different toxin preparations, dosing regimen, strains o f mice and toxin administration methods i n these studies may have resulted in these disparate findings. Despite the difference in the ability to control the proliferation of target cells, Tregs generated in our system and others all showed suppressive activities on cytokine responses by naive splenocytes. Generally, as antigen-induced proliferation is an inevitable stage that a naive T cell must undergo when it is activated and develops into an effector T cell, proliferation is often used as a readout for the response o f T cells when co-culture experiments are performed to determine the  99  suppressive activities o f Tregs. However, it must be borne i n mind that proliferation is only an indirect reflection o f T cell response, but not a direct index for T cell effector functions. Thus the use o f proliferation as the only readout o f T cell response i n some cases may not accurately reflect the T cell effector functions, as is shown in two recent reports. Ehrenstein et al showed that natural Tregs isolated from autoimmune disease patients prior to anti-TNF-a therapy were able to suppress the proliferation o f naive T cells but unable to suppress the cytokine production. However, natural Tregs isolated from patients responding positively to anti-TNF-a therapy were able to suppress both proliferation and cytokine production. These data clearly showed that cytokine response was a more accurate readout for T cell effector functions than the proliferative response i n certain circumstances (Ehrenstein et al., 2004). Moreover, L i n et al showed that Tregs generated using non-depleting anti-CD4 antibody were able to maintain transplantation  tolerance  without suppressing  the proliferation o f C D 8 T cells  mediating rejection in vivo ( L i n et al., 2002). It has been also demonstrated that superantigens are able to induce T cell responses that included massive proliferation and proinflammatory cytokines production. However, it is the cytokine responses that mediate the pathological processes triggered b y superantigens, such as lethal shock (Arad et a l , 2001; Dinges and Schlievert, 2001; Faulkner et al., 2005; Matthys et a l , 1995). Thus i n the current study, cytokine as readout for T cell responses was more accurate to reflect the suppressive activities o f TSST-1 or SEA-induced Tregs. Therefore, although TSST-1 and SEA-induced Tregs did not suppress the proliferation o f their target cells, they were still able to down-regulate cytokine levels in vitro and inhibit lethal shock in vivo.  100  Previous studies have established that repeated exposure o f T cells to S E A or S E B resulted i n the increased production o f IL-10 and the decreased production o f IL-2 and IFN-y. This cytokine profile has also been noted in Tregs induced in different systems, with T r l cells as the most typical example. Our results o f cytokine levels in the supernatant o f SEA-induced Tregs stimulated with S E A were i n line with the previously reported data (Feunou et al., 2003; Grundstrom et al., 2003; N o e l et al., 2001). In contrast, stimulation o f T S S T - l - i n d u c e d Tregs with TSST-1 not only led to decreased level o f IL-2 and increased levels o f IL-10, but also significantly increased levels o f IFN-y and T N F - a . Recently, several studies showed that T cells with regulatory functions also produced IFN-y (Hong et al., 2005; Kemper et al., 2003; Riemekasten et al., 2004; Sawitzki et al., 2005; Stock et al., 2004). The authors in these studies also demonstrated an essential role o f IFN-y either i n the regulatory functions, or in the induction o f Foxp3+ Tregs. However, the roles o f  IFN-Y  m  both the  generation and regulatory functions o f T S S T - l - i n d u c e d Tregs need to be further investigated. In line with the increased level o f IL-10 in the supernatant, flow cytometry analysis showed an increased percent o f IL-10 positive C D 4 + T cells in TSST-1 or SEA-induced Tregs stimulated with their cognate antigens. However, unexpectedly, flow cytometry analysis demonstrated that these Tregs did express other proinflammatory cytokines, such as  IFN-Y,  T N F - a and I L - 2 , regardless o f changed  levels i n the supernatant. These data for the first time showed that the decreased levels of proinflammatory cytokines in the supernatant o f stimulated superantigen-induced Tregs might not be due to lack o f expression, but might be due to inhibited release or enhanced consumption o f them by the Tregs themselves.  101  When TSST-1 or SEA-induced Tregs were not stimulated with their cognate antigens, the cytokine responses in vitro were i n some degree similar to their suppressive activities. While SEA-induced Tregs stimulated with TSST-1 produced I L 2, LFN-y and T N F - a comparable to the control, T S S T - l - i n d u c e d Tregs stimulated with S E A produced neglectable level o f I L - 2 and low level o f LFN-y. Thus, the cytokine responses o f SEA-induced Tregs were more antigen-specific while those o f T S S T - l induced Tregs were more broadly suppressive. This might be due to the antigen-nonspecific suppressive activities o f T S S T - l - i n d u c e d Tregs, so that the responses o f S E A reactive C D 4 + T cells i n the T S S T - l - p r i m e d splenocytes were suppressed when they were stimulated with S E A . The differences i n the cytokine secretion profile further confirmed the findings concerning the differences o f the suppressive activities between TSST-1 and SEA-induced Tregs. It has been previously shown that exposure to superantigen led to the activation of superantigen-reactive T cells followed b y deletion and anergy o f these T cells (Kawabe and Ochi, 1991; MacDonald et al., 1991; Rellahan et al., 1990; White et al., 1989). However, in our system, TSST-1 or SEA-induced Tregs did not become anergic in response to superantigen restimulation in vitro. Instead, they proliferated more vigorously than the control PBS-primed C D 4 + T cells, having characteristics o f antigen-experienced T cells. Although the proliferation response o f PBS-primed C D 4 + T cells to TSST-1 or S E A was similar, T S S T - l - i n d u c e d Tregs proliferated more rigorously than SEA-induced Tregs i n response to their cognate antigens. This was demonstrated by higher proliferation o f T S S T - l - i n d u c e d Tregs i n response to different concentrations o f cognate superantigen stimulation at day 2 and day 3. This  102  proliferative state o f superantigen-induced Tregs was intriguing. Multiple previous studies showed that generation o f Tregs, including superantigen-induced Tregs, was often associated with the development o f an anergic state o f the Tregs i n response to antigen restimulation (Groux et al., 1997; Jonuleit et al., 2000; Levings et al., 2005), although hyper-proliferative Tregs was reported occasionally (Kemper et al., 2003). Our data for the first time showed that superantigen-induced Tregs could also be hyperproliferative. The factors that might be responsible for this discrepancy with results from previous studies included the toxin preparation, the injection protocol o f superantigen, and the strains o f mice used. In terms o f the cytokine profile and proliferative response, one remarkable difference between TSST-1 and SEA-induced Tregs was the responses o f these cells in the absence o f cognate superantigen stimulation. While SEA-induced Tregs showed very modest proliferation and cytokine expression i n the absence o f stimulation and required SEA-stimulation to fully trigger cytokine expression and proliferation, T S S T l-induced Tregs showed remarkable cytokine expression and proliferation i n the absence o f TSST-1 stimulation. Thus it seemed that T S S T - l - i n d u c e d Tregs showed a more activated phenotype which was characterized by antigen-independent activities. It could be argued that this might be due to the residual binding o f the third dose o f superantigen to A P C s o f the splenocytes which triggered these activities o f T S S T - l induced Tregs in vitro as these cells were isolated two hours after the last dose o f superantigen administration. However, this was unlikely because S E A has been shown to have a stronger binding to A P C s than TSST-1 (Pless et al., 2005), yet SEA-induced Tregs did not show similar antigen-independent activities as T S S T - l - i n d u c e d Tregs.  103  Thus, this feature o f T S S T - l - i n d u c e d Tregs was probably due to a unique activation process induced by repeated injection o f TSST-1 in vivo, which was also suggested by the analysis o f cellular markers o f these Tregs. This more activated phenotype o f T S S T - l - i n d u c e d Tregs may also be partially responsible for their broader suppressive activities because possessing this phenotype might enable them to respond more easily to other stimuli and exert their suppressive functions. W h i l e SEA-induced Tregs did not have this phenotype, they might be required to be activated first by their cognate antigen to exert their suppressive function, thus they were not able to suppress the responses triggered by TSST-1 i n the absence o f S E A . Further studies are required to test this hypothesis. The cell death o f TSST-1 or SEA-induced Tregs was also investigated in this study. In the control group, data showed that PBS-primed C D 4 + T cells underwent progressive cell death in the absence o f stimulation over a five-day in vitro culture. Addition o f superantigen did not increase the cell death o f C D 4 + T cells, but decreased the percentage o f C D 4 + T cells undergoing cell death. These data demonstrated that in this system lack o f stimulation caused B 6 C D 4 + T cell death while provision o f stimulation was a survival signal for C D 4 + T cells. However, it should be kept in mind that the in vitro culture was a dynamic system i n which cell death caused by lack o f stimulation, cell death caused by activation and increased cell number due to proliferation induced by stimulation co-exist. The percent o f death cells measured was a net result o f these three processes. The results showed that i n the absence o f stimulation, TSST-l-induced Tregs demonstrated a decreased overall percentage o f cell death as compared to SEA-induced Tregs and PBS-primed C D 4 + T cells. This was  104  probably because T S S T - l - i n d u c e d Tregs were able to undergo  antigen-independent  proliferation while the other two were not, as i f T S S T - l - i n d u c e d Tregs were stimulated by their cognate  antigen. This factor thus sustained their survival and added  proliferated C D 4 + T cells into the culture, which might offset the cells undergoing cell death due to lack o f stimulation or stimulation-induced cell death. A s the concentration of superantigens increased the difference o f cell death between TSST-1 and S E A induced Tregs began to disappear, suggesting that their survival in vitro i n response to higher doses o f stimulation was similar. However, their survival in vivo was unknown and further studies are required to investigate this issue. Foxp3 has been shown to be a marker associated with the development o f naturally Tregs (Hori et a l , 2003). Although it is a reliable marker for natural Tregs, it can not serve as a reliable marker for antigen-induced Tregs. This is because it has been shown that antigen-induced Tregs did not show enhance expression o f Foxp3 (Vieira et al., 2004). In this study, we did not detect a significantly increased expression i n TSST-1 or SEA-induced Tregs, which was i n line with the previous study where the expression o f Foxp3 m R N A i n SEB-induced Tregs was assayed (Feunou et al., 2003). These data also further confirm that Foxp3 is not a good marker for antigen-induced Tregs. In contrast, the expression o f C D 2 5 and C T L A - 4 on T S S T l-induced Tregs was one fold higher than that on SEA-primed C D 4 + T cells. Because increased expression C D 2 5 and C T L A - 4 normally could be seen i n stimulated T cells, these data indicated that the generation o f TSST-1 or SEA-induced Tregs in vivo b y repeated  injection o f these superantigens  was associated with a stimulation and  activation process, which has been proposed as a mechanism o f antigen-induced Tregs  105  in the periphery (Graca et a l , 2005). However, as the expression o f these activation markers was much higher in T S S T - l - i n d u c e d Tregs, this suggested that repeated injection o f TSST-1 in vivo has a more potent stimulatory signal than S E A . Moreover, our data also demonstrated  that antigen-independent  proliferation and cytokine  expression was much more remarkable i n T S S T - l - i n d u c e d Tregs than i n SEA-induced Tregs. A s previous studies showed that stimulated T cells could proliferate without further antigen stimulation (Duthoit et al., 2004; Lee et al., 2002), the proliferation and cytokine profile o f T S S T - l - i n d u c e d Tregs also indicated that the generation o f T S S T l-induced Tregs in vivo was associated with a distinct stimulation and activation process from that elicited by S E A . W h y and how this stimulation and activation process elicited by repeated injection o f TSST-1 differ from that elicited by S E A definitely needs more extensive investigation. In summary, our data clearly showed that T S S T - l - i n d u c e d Tregs possess properties distinct from those o f SEA-induced Tregs. Generation o f Tregs holds great therapeutic potential for various clinical conditions, such as transplant rejection and autoimmune diseases (June and Blazar, 2006). Unique properties o f T S S T - l - i n d u c e d Tregs, including their proliferative potential revealed i n this comparative study, may have important implications for selecting T S S T - l - i n d u c e d Tregs as a better candidate to be further studied for their clinical application potential. A s TSST-1 could induce Tregs with broader suppressive activities, they might be applied to the treatment o f more diseases regardless o f the antigens that cause those diseases. Their more potent proliferative response would be another advantage because recent data suggested that better proliferative ability o f Tregs in vivo was correlated with more favorable outcome  106  preventing disease (Trenado et al., 2003).  Chapter 5: Possible mechanisms of suppressive properties of TSST-linduced Tregs 5.1. Introduction Regulatory T cells (Tregs), including naturally occurring CD4+CD25+Foxp3+ Tregs (Sakaguchi, 2004; Sakaguchi, 2005) and antigen-induced Tregs (Bluestone and Abbas, 2003; Vigouroux et al., 2004), have been shown to play an important role in modulating the immune response. While naturally occurring C D 4 + C D 2 5 + Tregs are generated in the thymus, antigen-induced Tregs are generated i n the periphery after antigen encounter. Mechanisms underlying suppression by different types o f Tregs are poorly defined. Possible mechanisms included: 1) cytotoxic pathway (Grossman et al., 2004; Janssens et al., 2003; Zhao et a l , 2006); 2) inhibition o f activation o f target cells (Barthlott et al., 2005; de la Rosa M et al., 2004; George T C et al., 2003; Trenado et al., 2006); 3) infectious tolerance, referring to the ability o f Tregs to render their target cells to acquire regulatory functions (Alpan et al., 2004; Dieckmann et al., 2002; Dieckmann et al., 2005; Jonuleit et al., 2002; Unger et al., 2003); 4) soluble factors mediating suppression (Asseman et al., 1999; Fahlen et al., 2005; Grundstrom et al., 2003; M c G u i r k et al., 2002); and 5) cytokine competition (Barthlott et al., 2005; de la Rosa M et al., 2004; Gunther J et al., 1982; Scheffold A et al., 2005). Staphylococcal superantigens,  such as T S S T - 1 , activate C D 4 + T cells by  crosslinking M H C class II molecules and the T cell receptor bearing specific VP segments (Dinges et al., 2000). Experimentally, chronic exposure to superantigen has been shown to effectively induce Tregs (Feunou et al., 2003; Grundstrom et al., 2003; M i l l e r et al., 1999; N o e l et al., 2001). In the previous chapter it has been shown that  108  repeated injection o f TSST-1 i n C 5 7 B L / 6 mice could induce C D 4 + Tregs with distinct characteristics from those induced by S E A . T S S T - l - i n d u c e d Tregs  demonstrated  suppressive activities on both TSST-1 and S E A in vitro and in vivo. They downregulated the cytokine responses induced by TSST-1 or S E A in vitro, and suppressed lethal shock induced by either TSST-1 or S E A . In this chapter, the possible mechanisms o f T S S T - l - i n d u c e d Tregs are examined.  5.2. Experimental Design The  possible mechanisms o f T S S T - l - i n d u c e d Tregs were investigated. T o  determine i f T S S T - l - i n d u c e d Tregs caused cell death or inhibited the activation o f target cells, Tregs were co-cultured with CFSE-labelled naive splenocytes i n the presence o f TSST-1 and then the cell death or expression o f activation markers on naive splenocytes were assayed b y F C M . To study i f target cells acquired regulatory functions after co-culture with Tregs, the naive splenocytes were co-cultured with CFSE-labelled Tregs and reisolated for assay o f their cytokine profile and suppressive functions. To determine the role o f soluble factors i n the suppression by Tregs, splenocytes containing Tregs were stimulated with TSST-1 and then the supernatant was transferred to co-culture with naive splenocytes. Cytokine responses were assayed by E L I S A . T o investigate i f Tregs could consume cytokines produced by target cells, naive splenocytes were stimulated with TSST-1 or S E A , then the supernatant was transferred to co-culture with Tregs. Cytokines i n the supernatant were determined b y E L I S A . A flow diagram o f the experimental design is shown below.  109  T S S T - 1 , x3  ^ 2 Tregs  /  x3  ^O-  C F S E staining Naive  ®  PBS,  Naive mice  splenocytes  ^^-j  Co-culture  Apoptosis o f naive S C by FCM  Assay cytokines o f supernatant  expression on naive S C by F C M  Naive S C were reisolated by F A C S  i Assay o f cytokine profile and suppressive functions  Transfer o f supernatant  1  Activation marker  ®  Splenocytes  Transfer o f supernatant Assay cytokines o f supernatant  Stimulated with TSST-1  Figure 5.1. Experimental design o f studies described i n Chapter 5.  110  5.3. Results 5.3.1. T S S T - l - i n d u c e d Tregs did not enhance cell death o f the target cells To investigate whether T S S T - l - i n d u c e d Tregs exerted their functions by inducing cell death o f their target cells, naive splenocytes were stained with C F S E and co-cultured with TSST-1-indcued Tregs i n the presence o f T S S T - 1 . C e l l death o f the target cells was then determined by flow cytometry at different time point post stimulation. A s shown, in Figure 5.2, addition o f T S S T - l - i n d u c e d Tregs to naive splenocytes did not lead to enhanced cell death o f the target cells at either day 1 or day 2 post stimulation as compared to na'ive splenocytes alone or naive splenocytes cocultured with PBS-primed control C D 4 + T cells. Therefore, these data showed that apoptosis induced by T S S T - l - i n d u c e d Tregs was not responsible for their suppressive activities.  5.3.2. T S S T - l - i n d u c e d Tregs enhanced the activation o f target cells Activation is a critical step o f the T cell response to the cognate antigens, and is subject to the regulation by Tregs. Thus, efforts were made to determine i f T S S T - l induced Tregs inhibited the functions o f their target cells b y suppressing the activation o f the target cells i n response to TSST-1 stimulation. To this end, naive splenocytes were first labeled with C F S E and then co-cultured with TSST-1-indued Tregs. The activation state o f the target cells was then determined by analysis o f the expression o f C D 2 5 and C D 6 9 on target CD4+ T cells at different time points post stimulation. A s shown i n Figure 5.3, surprisingly, co-culture o f naive splenocytes w i t h T S S T - l induced Tregs resulted in enhanced, rather than decreased, expression o f C D 2 5 and  111  Naive+T  59.4%  63.9%  Naive+TSST-lTre2f5:4)+T  NaTve+PBS-T(5:4)+T  W  t/i  safi  u  io  Q  in'  vi^-.;-.  ID  2  5  0  io  >  8  ::  %  51.7%  3  7-AAD  B  Naive splenocytes TSST-l-Tregs (5:4) TSST-l-Tregs (5:1)  +  PBS-CD4 T (5:4) PBS-CD4 T (5:1) TSST-1 RPMI  +  +  +  +  +  +  +  Figure 5.2. Effect o f T S S T - l - i n d u c e d Tregs on the cell death o f target cells. Naive splenocytes ( 2 x l 0 / w e l l ) were stained with C F S E and co-cultured with TSST-1 inducd Tregs or control PBS-primed C D 4 + T cells (ratio o f na'ive splenocytes to T S S T - l - i n d u c e d Tregs or PBS-primed C D 4 + T cells at 5:4 and 5:1) i n the presence o f T S S T - 1 . A t 24 and 48 hours, cells were stained with 7 - A A D followed by flow cytometry analysis. A . Representative data o f flow cytometry analysis o f cells at day L B . Accumulated data from four independent experiments. A s shown, addition o f T S S T - l - i n d u c e d Tregs did not result i n enhanced cell death o f the na'ive splenocytes. Data shown were accumulated results o f four independent experiments (n=4). 5  112  A Naive+TSST-1  Naive+RPMI  tf|8  • ..".'"".rT;  10%  ;  o Naive+TSST-1 Treg(5:4HT  i5%  Naive+TSST-1 Treg(5:l)+T  35%  '50%;  k Naive+PBS CD4 T(S:1)+T  .Naive+PBS CD4 T(5:4HT  pi  31%  ta U  20%  i CD25  F i g u r e 5.3.  Effect o f T S S T - l - i n d u c e d  Representative naive ( 4 x l 0 Tregs  raw  splenocytes 5  / w e l l ) or  control  presence and were  of  stained  with  T r e g s o n the activation o f target cells.  cytometry  A t  data  24  C D 4 +  and  48  followed o f  flow  analysis  T S S T - l - i n d u c e d  C F S E  or P B S - p r i m e d  antibodies  representative  n a i v e c e l l s at d a y  flow  P B S - p r i m e d Tregs  T S S T - 1 .  anti-CD4  o f  co-cultured with  were  T S S T - l - i n d u c e d  data  and T  b y  C D 2 5  cells  (ratio T  cells  flow  cytometry  expression  Tregs. N a i v e  co-cultured with  C D 4 +  hours,  of  o f  cells were  stained  cytometry analysis  5:4  splenocytes  splenocytes and  5:1)  with  analysis. o f  on  T S S T - l - i n d u c e d  naive  at  A .  C D 2 5  in  to the  anti-CD25  Data  shown  expressed  on  2.  113  B  + Q U  ^  >  w  cu -  •a JS +  Z  Q U  50n  + Q U  40H  Qi  + ON  Q  Day 1 Day 2  CD69  ^ 0  H U  s  30H  *_§*§  20H 10H  Naive Splenocytes TSST-l-Tregs (5:4)  +  +  TSST-l-Tregs (5:1)  + +  JJ +  +  PBS-CD4 T (5:1) TSST-1 +  +  +  +  +  PBS-CD4 T (5:4)  RPMI  +  i  +  +  + +  Figure 5.3. B. Data shown were accumulated results of four independent experiments (n=4). *, PO.05, compared with naive splenocytes, paired t-test (Day 1 or Day 2). §, PO.05, compared with nai've splenocytes+PBS-primed CD4+ T cells, paired t-test (Day 1 or Day 2).  114  C D 6 9 on na'ive C D 4 + T cells as compared to naive splenocytes alone. In the control groups, adding PBS-primed splenocytes to na'ive splenocytes also increased the expression o f C D 2 5 and C D 6 9 on naive C D 4 + T cells. Therefore, T S S T - l - i n d u c e d Tregs enhanced, rather inhibited, the activation o f their target cells.  5.3.3. T S S T - l - i n d u c e d Tregs did not render their target cells to acquire regulatory functions Since T S S T - l - i n d u c e d Tregs enhanced the activation o f their target cells, it was then asked i f T S S T - l - i n d u c e d Tregs could render their target cells to acquire regulatory function. To test this hypothesis, the T S S T - l - i n d u c e d Tregs were first stained with C F S E and then co-cultured with na'ive splenocytes for 48 hours. The naive splenocytes were then reisolated by F A C S and analyzed for their suppressive activities by coculture experiments. A s shown in Figure 5.4, reisolated naive splenocytes had a proinflammatory cytokine production profile similar to the control groups i n response to TSST-1 re-stimulation. Co-culture o f the re-isolated splenocytes with fresh na'ive splenocytes did not result in any down-regulation o f proinflammatory cytokine levels. Concurrent  re-isolated  TSST-l-induced  Tregs  still  had  regulatory  functions,  demonstrating that the lack o f regulatory activities i n the co-cultured naive splenocytes was not due the isolation procedure (data not shown). Thus, infectious tolerance was not involved in the activities o f T S S T - l - i n d u c e d Tregs.  115  1250n 1000-  IL-2  JL  750-  s  5002500  =  400i  IL-12  T  300-  S  200-  =  10O  o  U  0  800-  IFN-y  6004002000-  Naive SC(4xl0 ) 5  +  Tregs-cocultured SC (8x10 ) s  + +  Tregs-cocultured SC (4xl0 )  +  5  PBS-T-cocultured SC (8x10 ) s  PBS-T-cocultured SC (4xl0 ) TSST-1  +  5  +  +  +  + + +  +  + + +  + +  +  Figure 5.4. T S S T - l - i n d u c e d Tregs did not render the target cells to acquire regulatory functions. Naive splenocytes were first co-cultured with CFSE-stained T S S T - l - i n d u c e d Tregs or PBS-primed C D 4 + T cells at the ratio o f 5:4 i n the presence o f TSST-1 for 48 hours. Then dead cells were removed and naive splenocytes were re-isolated by F A C S . The re-isolated splenocytes were cocultured with freshly prepared naive splenocytes at two ratios or cultured alone i n the presence o f TSST-1 for 48 hours. Various cytokines i n the supernatant were assayed by E L I S A . A s shown, splenocytes co-cultured with T S S T - l - i n d u c e d Tregs still showed a proinflammatory cytokine profile i n response to TSST-1 stimulation as compared to freshly prepared na'ive splenocytes and splenocytes cocultured with PBS-primed C D 4 + T cells. They did not show any suppressive activity on the cytokine response o f the freshly prepared na'ive splenocytes. Data shown were accumulated results o f three independent experiments (n=3).  116  5.3.4. Supernatant o f stimulated T S S T - l - i n d u c e d Tregs did not mediate suppressive activities Tregs have been well-known for producing soluble immunosuppressive factors, such as IL-10, to mediate their suppressive activities. To determine the role o f soluble factors i n mediating their regulatory functions, T S S T - l - p r i m e d splenocytes, which contained T S S T - l - i n d u c e d Tregs and A P C s , were first stimulated with T S S T - 1 , then the supernatant was harvested and used to culture na'ive splenocytes to determine i f the responses o f target cells could be down-regulated in the presence o f medium containing soluble factors produced by stimulated T S S T - l - i n d u c e d Tregs. A s shown i n Figure 5.5, transfer o f TSST-1-conditioned medium from T S S T - l - p r i m e d splenocytes to na'ive splenocytes did not result in decreased levels o f IL-2 and IL-12, compared to na'ive splenocytes cultured i n control medium. Increasing the number o f T S S T - l - i n d u c e d Tregs up to four-fold did not render the supernatant to have suppressive ability. Thus, these data showed that the suppressive activities o f T S S T - l - i n d u c e d Tregs were not solely mediated by soluble factors produced by these Tregs.  5.3.5. Suppressive activities o f T S S T - l - i n d u c e d Tregs were cell contact-dependent Since the experiments above did not reveal an exclusive role for soluble factors in the suppressive functions o f T S S T - l - i n d u c e d Tregs, it was reasoned that T S S T - l induced Tregs may depend on cell-contact with their target cells to exert their functions. To test this hypothesis, a permeable transwell was added to the co-culture o f T S S T - l induced Tregs and naive splenocytes to determine i f separation o f these two populations w o u l d lead to the reversal o f suppression. A s shown i n Figure 5.6,  117  150Oi  IL-2  lain  -S  sow  ~5JD a. e  o  c s J.  6001  IL-12  c  « a o U  40tH  20CH 0  Naive Splenocytes Fresh RPMI  I  I  + +  +  I  +  +  Tregs supernatant (8><10) 5  Tregs supernatant (2*10 ) 6  TSST-1  +  +  +  +  Figure 5.5. The role o f soluble factors i n the suppressive function o f T S S T - l induced Tregs. T S S T - l - i n d u c e d Tregs ( 8 x l 0 / w e l l or 2 x l 0 / w e l l ) together with antigen presenting cells were first stimulated with TSST-1 for 48 hours. Supernatant then was collected and used to culture naive splenocytes ( 4 x l 0 /well) for another 48 hours i n the presence o f T S S T - 1 . Supernatant then was assayed by E L I S A for I L - 2 and I L - 1 2 levels. Naive splenocytes cultured i n the Tregs supernatant produced similar level o f IL-2 and I L - 1 2 as compared to naive splenocyte i n fresh medium or control medium, showing that soluble factors contained in the Tregs supernatant were not responsible for the suppressive effects o f T S S T - l - i n d u c e d Tregs. Data shown were accumulated results o f four independent experiments (n=4). 5  6  5  118  2000-1 fl  c a e  «  1500-  •  X  1000-  1  *  c  IL-2  500-  w  U  0  Naive Splenocytes  + +  TSST-l-Tregs T r a n s w e l l membrane  +  +  TSST-1  + + + +  PBS-CD4 T  I  1  ecococo  +  ccococo  + + ccococo  Figure 5.6. Cell contact-dependency o f T S S T - l - i n d u c e d Tregs. Naive splenocytes ( 2 x l 0 / w e l l ) were co-cultured with T S S T - l - p r i m e d splenocytes i n the presence o f TSST-1 at the ratio o f 1:1 and were separated by a permeable transwell for 48 hours. Supernatant was assayed for IL-2 and IL-12 by E L I S A . A s shown, high levels o f IL-2 and IL-12 were detected in supernatant o f naive splenocytes stimulated with T S S T - 1 . Addition o f T S S T - l - p r i m e d splenocytes containing Tregs to naive splenocytes decreased levels o f IL-2 and IL-12. Separation o f naive splenocytes and T S S T - l - p r i m e d splenocytes with a permeable transwell resulted in restoration o f 80% o f IL-2 and 100% o f IL-12 levels. Addition o f PBS-primed splenocytes did not led to decreased levels o f IL-2 and IL-12. The data showed that suppressive effects o f TSST-l-induced Tregs were cell-contact dependent or required the Tregs to be in the vicinity o f their target cells. Data shown were accumulated results o f four independent experiments (n=4). 6  119  separation o f T S S T - l - i n d u c e d Tregs and naive splenocytes with a permeable transwell membrane did lead to the reversal o f suppression, as IL-2 and IL-12 levels in this group were restored to 80% and 100% o f control levels, respectively. Therefore, cell contact was required for T S S T - l - i n d u c e d Tregs to exert their functions.  5.3.6. Blockade o f IL-10 with an anti-IL-lOR monoclonal antibody did not reverse the suppressive functions o f TSST-l-induced Tregs IL-10 was one o f the major immunosuppressive cytokines that mediated the suppression by Tregs. In general, soluble IL-10, such as that produced by T r l cells, is responsible for their suppressive activities. However, a recent study did clearly show that IL-10 had to exert its suppressive effects in the context o f cell contact (Alpan et al., 2004). Thus, although it was demonstrated  above that soluble factors did not  exclusively mediate the suppression by T S S T - l - i n d u c e d Tregs and cell contact was actually required for their suppressive activities, it was still possible that IL-10 might play a role in the suppression by T S S T - l - i n d u c e d Tregs. To further determine the role o f IL-10, a monoclonal antibody against the IL-10 receptor was added to the co-culture of T S S T - l - i n d u c e d Tregs and naive splenocytes to block the actions o f IL-10. The results showed that addition o f the anti-IL-10 receptor antibody did not reverse the suppressive activities o f T S S T - l - i n d u c e d Tregs as levels o f proinflammatory cytokines were still down-regulated (Figure 5.7). This was not due to insufficient concentration of anti-IL-10 receptor antibody which had been raised to 10 times the concentration  120  1500T  IL-2  IOOOI 500-  o400"  & a  o  v» «s  u  fl w fl o  U  IL-12  300" 200100-  o-' 800-  IFN-y  600" 400" 200"  TSST-1 Naive Splenocytes TSST-l-Tregs a-IL-lOR A b Control A b  0  J  + +  + + +  + + + +  + + + +  Figure 5.7. Effect o f addition o f anti-LL-10 receptor antibody on the suppressive function o f TSST-l-induced Tregs. Naive splenocytes were co-cultured with Tregs at the ratio o f 5:1 (for IL-2 and IL-12) or 5:4 (for IFN-y) i n the presence o f anti-IL-10 receptor antibody for 48 hours and supernatant was assayed for various cytokines by E L I S A . Addition o f anti-IL-10 receptor antibody (lOOpg/mL) did not reverse the levels o f IL-2, IL-12 and IFN-y as compared to those in the absence o f this antibody or in the presence o f control antibody. Data shown were accumulated results o f four independent experiments (n=4).  121  previously shown to be sufficient to block IL-10 actions in vitro (O'Farrell A M et al., 1998). Taken together, these data suggest that T S S T - l - i n d u c e d Tregs do not depend on IL-10 to exert their suppressive activities.  5.3.7. Intracellular cytokine expression by naive splenocytes was not suppressed in the presence o f T S S T - l - i n d u c e d Tregs Results o f the experiments  above led us to look for other  mechanisms  responsible for the suppressive activities o f T S S T - l - i n d u c e d Tregs. Several reports have shown that cytokine competition was an important mechanism by which natural Tregs controlled the responses o f their target cells (Barthlott et al., 2005; de la Rosa M et al., 2004). Thus, it was hypothesized that cytokine competition also played a role in the suppressive activities o f TSST-1-induded Tregs. To this end, it was essential to first determine the cytokine production by naive splenocytes i n the co-culture with T S S T - l induced Tregs. In the co-culture system E L I S A was performed to assay the cytokine levels in the supernatant. However, this method could only determine the net result o f cytokine production and consumption. Therefore, a reduction in the level o f a specific cytokine might be due to reduced production or enhanced consumption. Intracellular staining o f cytokine followed by flow cytometry analysis was performed to address this issue. Naive splenocytes were first stained with C F S E and then co-cultured with T S S T l-induced Tregs for 45 hours, with Golgistop added to the culture for the last 5 hours to stop the cytokine secretion. Cells were then stained for cytokines and analyzed by flow cytometry. A s shown in Figure 5.8, co-culture o f T S S T - l - i n d u c e d Tregs with naive splenocytes did not lead to the decreased expression o f either IL-2 or IFN-y on  122  Nai've+RPMI p  Naive+TSST-1  -1  111 |g;v;.>'-;,.MFI:42 11%  2.5%  10°  io  io*  1  io  io  3  IL-2-APC  Naive+TSST-1 Treg(5:l)+T  |*U  16% MFI:22  19% MFI:24  |;>:  io°  io  io  1  2  io  1  io  Naive+PBS CD4 T(5:l)+T  tvv. 11% ^;f:- -viVlFI:37  17% pf v;>v ''MFI:37  :  ;  :  u  t  10"  10'  10IL-2-APC  • •  10  J  io°  io  1  io-  io  J  io  IL-2-APC  IL-2  Figure 5.8. Expression o f IL-2 by naive splenocytes i n the presence o f T S S T l-induced Tregs. A . Representative raw data o f expression o f IL-2 by naive splenocytes in the presence o f T S S T - l - i n d u c e d Tregs. CFSE-labelled naive splenocytes were co-cultured with TSST-l-induced Tregs at the ratio o f 5:1 or 5:4 in the presence o f TSST-1 for about 45 hours. Golgistop was added for the last 5 hours. Intracellular staining was performed for I L - 2 as described in Methods and Materials followed by flow cytometry analysis. Data shown were IL-2 expression by na'ive splenocytes in the presence o f T S S T - l induced Tregs from one o f four independent experiments. Na'ive splenocytes cultured in the absence o f TSST-1 stimulation served as the negative control for IL-2 exoression.  123  4>  20-i  >  B  IL-2  o  15-  a. _  |8  10-  ^s  5-  o U  0-  40-  30-  22 °  X  X  X  20-  10-  E  e  0-  Na'ive Splenocytes TSST-l-Tregs (5:4)  "-1—  +  TSST-l-Tregs (5:1)  + +  +  +  +  +  PBS-CD4 T (5:4) PBS-CD4 T (5:1) TSST-1  +  +  + +  + + +  Figure 5.8. B . Intracellular I L - 2 expression by naive splenocytes i n the presence o f T S S T - l - i n d u c e d Tregs. Data shown were accumulated from four independent experiments, n=4.  124  > ^ O  20i  S i  15H  £s O  X  ^  IFN-Y  10  U  a. 40i o "S  .s +  30H  X  201  X  X  x  10 HH HH  W B  0  Naive Splenocytes TSST-l-Tregs (5:4)  J  +  TSST-l-Tregs (5:1)  + +  +  +  +  +  PBS-CD4 T (5:4) PBS-CD4T(5:1) TSST-1  +  +  + +  +  + +  Figure 5.8. C . Intracellular IFN-y expression by naive splenocytes i n the presence o f T S S T - l - i n d u c e d Tregs. Data shown were accumulated from three independent experiments, n=3.  125  stimulated splenocytes. The mean fluorescence intensity ( M F I ) o f the IL-2 positive naive splenocytes co-cultured with T S S T - l - i n d u c e d Tregs was decreased as compared to that o f naive splenocytes alone, while the M F I o f IFN-y positive naive splenocytes co-cultured with T S S T - l - i n d u c e d Tregs was unaffected. Thus these data suggested that although the levels o f proinflammatory cytokines in the supernatant were decreased by T S S T - l - i n d u c e d Tregs, the expression o f these cytokines by their target cells might not be suppressed. The results o f these experiments led to the hypothesis that cytokines produced by naive splenocytes were either prevented from being released to the supernatant or were consumed by T S S T - l - i n d u c e d Tregs. The latter possibility then was investigated in the following studies.  5.3.8. T S S T - l - i n d u c e d Tregs competed for IL-2 with naive splenocytes To investigate i f T S S T - l - i n d u c e d Tregs could consume cytokines produced by naive splenocytes, naive splenocytes were first stimulated with TSST-1 or S E A for 48 hours and then the supernatant was harvested and cultured with T S S T - l - i n d u c e d Tregs for another 48 hours. Cytokine levels in the supernatant were assayed by E L I S A . A s shown i n Figure 5.9, culture o f T S S T - l - p r i m e d splenocytes containing T S S T - l induced Tregs i n TSST-1 or SEA-conditioned supernatant resulted i n decreased levels of IL-2, but not IL-12 or IFN-y as compared to the supernatant cultured with no cells or with PBS-primed splenocytes. To further prove that I L - 2 in the conditioned medium was consumed, an anti-CD25 antibody was added to the culture to block the interaction of IL-2 and its high affinity receptor C D 2 5 . Addition o f this antibody restored the level o f IL-2 i n the TSST-l-conditioned supernatant cultured with T S S T - l - i n d u c e d Tregs  126  TSST-l-conditioned supernatant 2000'  IL-2  1500' 1000' 500'  s  o  "Si 3  1000-  o  800-  S3  IL-12  600e u e  4002000-  o U  1000'  IFN-y  800' 600' 400' 200'  TSST-l-conditioned supernatant TSST-l-Tregs (0.8x10 ) TSST-l-Tregs (3.2*10 ) s  5  PBS-CD4 T (0.8x10 ) s  PBS-CD4 T (3.2x10 ) s  0-  -r-  + + + +  +  + + +  +  Figure 5.9. Competition o f I L - 2 by T S S T - l - i n d u c e d Tregs. Naive splenocytes ( 2 x l 0 / w e l l ) were stimulated with TSST-1 or S E A for 48 hours. Then the supernatant was collected and used to cultured T S S T - l - i n d u c e d Tregs ( 0 . 8 x l 0 / w e l l or 3 . 2 x l 0 / w e l l ) i n the presence o f antigen presenting cells for another 48 hours. Supernatant was then assayed for various cytokines by E L I S A . A . Culture o f TSST-1 conditioned naive splenocyte supernatant with T S S T - l induced Tregs resulted i n the significantly decreased level o f IL-2 as compared to supernatant cultured with no cells or with PBS-primed C D 4 + T cells. I L - 1 2 and IFN-y levels were not affected by the co-cultured o f TSST-1 or S E A conditioned naive splenocyte supernatant with T S S T - l - i n d u c e d Tregs. The data indicated that T S S T - l - i n d u c e d Tregs could consume I L - 2 produced by the naive splenocytes. *, P<0.05, compared with TSST-l-conditioned supernatant alone, paired t-test. Data shown were accumulated results from four independent experiments (n=4). 6  5  5  127  B  2000'  SEA-conditioned supernatant IL-2  1500' 1000-  ta s  500-  "bio  0-  C-  1000  C  IL-12  800  o  S3 i-  600  s <j B © U  400 200 0 1000-  IFN-y  8006004002000-  SEA-conditioned supernatant TSST-l-Tregs (0.8x10 ) s  TSST-l-Tregs (3.2x 10 ) s  PBS-CD4 T (0.8x10 ) s  PBS-CD4 T (3.2xl0 ) 5  + + + + + + +  +  +  Figure 5.9. B . Culture o f SEA-conditioned naive splenocyte supernatant with T S S T - l - i n d u c e d Tregs resulted i n the significantly decreased level o f IL-2 as compared to supernatant cultured with no cells or with PBS-primed C D 4 + T cells. IL-12 and IFN-y levels were not affected by the co-cultured o f TSST-1 or SEA-conditioned naive splenocyte supernatant with T S S T - l - i n d u c e d Tregs. The data indicated that TSST-l-induced Tregs could consume IL-2 produced by the naive splenocytes. *, P O . 0 5 , compared with SEA-conditioned supernatant alone, paired t-test. Data shown were accumulated results from four independent experiments (n=4).  128  TSST-l-conditioned supernatant  2000'  —1  1500'  s  1000'  a  o iC w u C  o  500' 0-»-r 2000'  SEA-conditioned supernatant  1500' 1000' 5000'  TSST-l/SEA-conditioned supernatant  +  TSST-l-Tregs (0.8xl0 )  +  5  TSST-l-Tregs (3.2xl0 ) 5  a-CD25Ab Control Ab  +  +  +  +  '+ +  + +  + +  +  +  + +  +  Figure 5.10. Effect o f anti-CD25 antibody on the uptake o f I L - 2 by T S S T - l induced Tregs. In experiments described i n F i g 5.9, groups were set up where monoclonal antibody against C D 2 5 , the high affinity receptor subunit o f IL-2 receptor, was added to the co-culture o f TSST-1 or SEA-conditioned naive splenocyte supernatant and TSST-l-induced Tregs. A s shown, TSST-1 or S E A conditioned naive splenocyte supernatant contained high level o f I L - 2 when it was cultured with no cells for 48 hours. In the presence o f T S S T - l - i n d u c e d Tregs, IL-2 level decreased to undetectable level while addition anti-CD25 antibody (10|j.g/mL) restored the level o f IL-2 to that in the absence o f any cells. Control antibody was not able to restore the I L - 2 level. The data further confirmed that I L - 2 was uptaken v i a I L - 2 receptor. Data shown were accumulated results o f four independent experiments (n=4). *, P<0.05, paired ttest, n=4.  129  TSST-l-Tregs  PBS-CD4 T cells  Figure 5.11. Expression o f C D 2 5 on T S S T - l - i n d u c e d Tregs i n the coculture with naive splenocytes. In experiments described i n Figure 5.3, the expression o f C D 2 5 on T S S T - l - i n d u c e d Tregs was also assayed simultaneously. T S S T - l - i n d u c e d Tregs could be identified as C F S E negative and C D 4 positive population. A s shown, T S S T - l - i n d u c e d Tregs had higher expression o f C D 2 5 than PBS-primed C D 4 + T cells. Data were accumulated from four independent experiments. *, P<0.05, compared with P B S - C D 4 T cells, t-test, n=4.  130  (Figure 5.10). In addition, the expression o f C D 2 5 on T S S T - l - i n d u c e d Tregs in the coculture was also assayed. A s shown in Figure 5.11, C D 2 5 was highly expressed on T S S T - l - i n d u c e d Tregs. It could be as high as 70% o f the T S S T - l - i n d u c e d Tregs at days o f the co-culture. Thus collectively, these data showed that T S S T - l - i n d u c e d Tregs down-regulated the IL-2 level by active consumption o f this cytokine.  5.3.9. Addition o f exogenous IL-2 did not reverse the suppression o f T S S T - l - i n d u c e d Tregs Since IL-2 plays an important role in the T cell responses and T S S T - l - i n d u c e d Tregs were able to deprive their target cells o f this important cytokine, it was reasonable to ask i f TSST-l-induced Tregs suppressed their target cells by preventing their target cells from accessing IL-2. T o address this issue, exogenous I L - 2 was added to the co-culture o f naive splenocytes and T S S T - l - i n d u c e d Tregs. A s shown i n Figure 5.12, the addition o f exogenous IL-2 did not reverse the suppressive activities o f T S S T l-induced Tregs, suggesting that the presence o f additional mechanisms to control cytokine responses o f naive splenocytes by T S S T - l - i n d u c e d Tregs.  131  800n  IL-12  600400-  -J  s  200-  "ex  0-  3  a  .©  « ts U  c  600-  V  400'  o  IFN-y  800n  200'  Naive Splenocytes  +  +  + +  +  + +  + +  +  + +  TSST-l-Tregs TSST-1 IL-2 (10 ng/ml)  Figure 5.12. Effect o f addition o f exogenous IL-2 on the suppression mediated by TSST-l-induced Tregs. Naive splenocytes were co-cultured with TSST-l-induced Tregs at the ratio o f 5:4 in the presence o f TSST-1 and exogenous IL-2 (lOng/mL) for 48 hours and supernatant was assayed for I L 12 and IFN-y by E L I S A . A s shown, IL-12 and LFN-y were produced by naive splenocytes in response to T S S T - 1 . Addition o f exogenous IL-2 (lOng/mL) to naive splenocytes did not alter the IL-12 and LFN-y- Co-culture of naive splenocytes with TSST-l-induced Tregs resulted i n decreased levels of IL-12 and LFN-y. Addition o f IL-2 to the co-culture did not restore the levels o f IL-12 and IFN-y. The data suggested that regulation o f IL-12 and LFN-y by TSST-l-induced Tregs differed from that o f IL-2. Data shown were accumulated results o f four independent experiments (n=4).  132  5.4. Discussion Previous studies on the mechanisms o f Tregs induced by superantigens, such as S E A or S E B , showed that these cells mainly rely on soluble factors to mediate their effect (Miller et al., 1999; N o e l et a l , 2001). In the last chapter, our studies showed that co-culture o f TSST-l-induced Tregs with naive splenocytes in vitro led to decreased levels o f proinflammatory cytokines. Thus i n this chapter, the possible mechanisms o f TSST-l-induded Tregs were investigated. W e first determined i f T S S T l-induced Tregs induced apoptosis o f their target cells. This mechanism has been noted in Tregs in several published studies with different antigens (Grossman et al., 2004; Janssens et al., 2003; Zhao et a l , 2006). However, i n our system, co-culture o f T S S T l-induced Tregs did not result in enhanced cell death o f the target cells, showing that induction o f cell death o f target cells was not the major mechanism. It has been demonstrated in vivo and in vitro that activation o f target cells could be inhibited by naturally occurring Tregs as up-regulation o f activation markers, such as C D 2 5 , was suppressed. Interestingly, co-culture o f T S S T - l - i n d u c e d Tregs did not lead to down-regulation o f C D 2 5 and C D 6 9 expression; instead, up-regulation o f C D 2 5 and C D 6 9 was noted. These data showed not only that the activation o f naive C D 4 + T cells was not inhibited, but also that T S S T - l - i n d u c e d Tregs might cause activation o f their target cells. W e then further determined i f co-culture o f these Tregs with their target cells could render the target cells to acquire regulatory functions. However, re-isolated naive splenocytes co-cultured with T S S T - l - i n d u c e d Tregs did not show any suppressive activities in secondary co-culture experiments, indicating that  133  activation o f target cells i n the presence o f T S S T - l - i n d u c e d Tregs did not render them to acquiring regulatory functions. Soluble factor-mediated suppression was the most well-documented mechanism o f Tregs, where IL-10 was one o f the major immunosuppressive cytokines. Our previous data did show that T S S T - l - p r i m e d splenocytes produced high levels o f IL-10 and a high percentage o f TSST-l-induced Tregs expressed IL-10 by flow cytometry analysis. Yet, our data in this study did not reveal any role for soluble factors or IL-10 in the suppression mediated by T S S T - l - i n d u c e d Tregs. Thus, in contrast to previous reports with S E A or SEB-induced Tregs (Noel et a l , 2001), T S S T - l - i n d u c e d Tregs in C 5 7 B L / 6 mice did not depend exclusively on IL-10 or other soluble factors to mediate their suppressive effects, as least in vitro. These data are i n line with the observation that the suppressive activities o f T S S T - l - i n d u c e d Tregs are mainly cell-contact dependent, as separation o f T S S T - l - i n d u c e d Tregs from their target cells using a permeable membrane almost completely reversed their suppressive effects. For some types o f Tregs, their cytokine profile is highly associated with their suppressive activities. For instance, T r l cells depend on the production o f high levels o f soluble IL-10 to exert their functions, as neutralization o f IL-10 can totally block their suppression (Groux et al., 1997; M c G u i r k et al., 2002). However, some groups have reported that soluble IL-10 or other immunosuppressive factors were not exclusively responsible for the suppressive effects mediated by Tregs in different systems (Chen et al., 2003b; Steinbrink et al., 2002; Tang et a l , 2004). Thus, there are non-redundant mechanisms for Treg-mediated suppression, depending on the local environment and immune response.  134  Cytokine competition has long been recognized as one o f the important arms i n the control mechanisms for immune responses. Several recent papers have suggested that cytokine competition may be one o f the major mechanisms for the regulation mediated by CD4+CD25+ natural Tregs (Barthlott et al., 2005; de la Rosa M et al., 2004). These naturally occurring Tregs constitutively express I L - 2 receptor C D 2 5 but fail to produce IL-2 when activated. However, they can compete with their target cells for I L - 2 , leading to suppression o f activation or effector function o f their target cells. Our data clearly showed that T S S T - l - i n d u c e d Tregs could consume IL-2, but not other cytokines produced by target cells stimulated with TSST-1 or S E A . This can explain why T S S T - l - i n d u c e d Tregs could down-regulate IL-2 level when they were cocultured with naive splenocytes stimulated with either TSST-1 or S E A . In contrast to inhibition by naturally occurring Tregs, where the addition o f IL-2 to the co-culture can reverse the inhibition, addition o f exogenous IL-2 did not reverse the suppression mediated by T S S T - l - i n d u c e d Tregs. This could be explained by the role o f IL-2 in different suppression assay systems. In the co-culture system used to assay the suppressive functions o f natural Tregs, proliferation o f target cells, which is highly I L 2 dependent, is used as the readout o f response (de la Rosa M et al., 2004). However, for superantigen-induced proliferation and cytokine production, I L - 2 might not play a central role as T cells from IL-2 or IL-2 receptor deficient mice still responded normally to S E B stimulation as compared to w i l d type mice (Jin et al., 2006). Thus, although exogenous I L - 2 was added to offset the decreased level o f IL-2 available to the target cells due to consumption by T S S T - l - i n d u c e d Tregs, suppression o f IL-12 and IFN-y by T S S T - l - i n d u c e d Tregs still existed.  135  These findings do not necessarily mean that IL-2 is not important i n the superantigen-induced  responses.  I L - 2 is one o f the main cytokines induced by  superantigens (Rink L et al., 1996). The role o f IL-2 i n superantigen-induced responses could be shown b y the suppressive effects o f cyclosporine A on lethal shock, which targeted the activation o f T cells by blocking IL-2 production (Miethke T et al., 1993). Additionally one study also demonstrated that blockade o f I L - 2 b y antibody in vivo could partially inhibit lethal shock induced by superantigens (Mountz et al., 1995). Thus, it is possible that consumption o f IL-2 b y T S S T - l - i n d u c e d Tregs was one o f the major mechanisms for these cells to exert their suppressive functions. Further studies are required to confirm its role in vivo. In contrast to the regulation o f IL-2 by T S S T - l - i n d u c e d Tregs, mechanisms o f regulation o f IL-12 and IFN-y by these cells are less clear. Our data excluded cytokine competition as the responsible mechanism. Although flow cytometry analysis showed that target cells did express these cytokines induced b y T S S T - 1 , i n the presence o f T S S T - l - i n d u c e d Tregs, the levels o f these cytokines i n the supernatant were decreased. Taken together, these data suggest that T S S T - l - i n d u c e d Tregs may inhibit the release o f these cytokines b y the target cells. The identification o f specific molecules that mediate this effect w i l l be one o f the focuses i n future studies. In summary, we sought to elucidate the mechanisms o f suppression mediated by T S S T - l - i n d u c e d Tregs. W e found that T S S T - l - i n d u c e d Tregs regulated the actions of IL-2 b y competition for and consumption o f IL-2 from their target cells. They suppressed cytokine responses other than IL-2 v i a a cell-contact dependent manner.  136  The molecules responsible for the latter mechanism are unknown and require further investigation.  137  Chapter 6: Potential application of TSST-l-induced Tregs in the control of acute graft-versus-host disease 6.1. Introduction Although allogeneic hematopoietic stem cells transplantation ( A H S C T ) is a curative therapy for certain blood malignancies, a G V H D remains a major complication and main barrier to A H S C T . A s mentioned earlier, superantigen administration has been shown to regulate certain diseases, including transplantation rejection (Pan et al., 2003). One previous study investigated the effects o f S E B administration on acute graft-vs-host response using a non-lethal murine model where parent splenocytes were infused to unconditioned F l mice (Takenaka et al., 2001). They showed that S E B administration could down-regulate the allogeneic responses due to the ability o f S E B to induce deletion o f SEB-reactive T cells which happened to be alloreactive. However, this model is o f limited clinical relevance and it was not known i f superantigen-induced Tregs could mediate dominant suppression on a G V H D or not. In the previous chapter, it was shown that TSST-l-induced Tregs were able to suppress the T h l cytokine response by naive splenocytes. Because T h l cytokine response has been shown to play a critical role i n the pathogenesis o f a G V H D , it can be hypothesized that T S S T - l induced Tregs may also be able to suppress a G V H D . In this chapter, a chemotherapyconditioned parent-to-Fl lethal a G V H D model is used to test this hypothesis.  6.2. Experimental Design To determine i f T S S T - l - i n d u c e d Tregs are able to suppress a G V H D , splenocytes containing Tregs from B 6 (H-2 ) mice injected with 3 doses o f TSST-1 were b  138  adoptively transferred to B 6 D 2 F 1 recipient mice ( H - 2 ) that were conditioned by b/d  cyclophosphamide (300 mg/kg) 24 hours before. Recipient mice were then given either no T S S T - 1 , one dose, or two doses o f TSST-1 (4 pg/mouse/injection at 4-day interval). Survival was monitored. If suppression o f a G V H D was observed, the effects o f Tregs on the donor T cells, host antigen presenting cells, cytokine production and serum endotoxin levels were further studied. The role o f IL-10 and G V T effects were also investigated. A flow diagram o f the experimental design is shown below.  B 6 (H-2 ), T S S T - 1 , x3 b  »  ' ^ m  D  2 9  "  z  Jd!  Cyclophosphamide (300 mg/kg)  r^. , I f  Splenocytes containing Tregs  24 h later  N o stimulation  One dose o f TSST-1  T w o doses o f TSST-1  v  V  J  Survival was monitored I  Effects o f Tregs on donor T cells and host APCs  I f  a  G  V  H  D  J  as  suppressed  M  Effects o f Tregs on cytokine production and serum endotoxin level  . ~ ~  G V T effects -_  .-  Figure 6.1. Experimental design o f studies described i n Chapter 6.  139  6.3. Results 6.3.1. Development and optimization o f an a G V H D animal model Multiple murine models exist for the study o f a G V H D (Korngold R et al., 2005). Most o f the models use total body irradiation as the conditioning approach because irradiation is the most commonly used conditioning method clinically i n human A H S C T . Irradiation is used because it can effectively ablate the immune system and hematopoietic cells o f the immune-competent host, allowing the engraftment o f donor T cells and hematopoietic cells. A s the donor T cells are not rejected, they become activated and thus can mediate lethal a G V H D . M i c e undergoing a G V H D in these models have clinical symptoms, such as hunched back, diarrhoea, body weight loss and death. Survival and body weight were reliable gross indices for the severity o f a G V H D . However, when the proposed studies were planned, unfortunately  no  irradiation facility could be accessible. Thus, a non-irradiation conditioning approach was utilized for this study. Several such models exist. One model o f a G V H D is established by infusing parent mice splenocytes to their hybrid F l recipient mice. According to the literature, when this donor-recipient combination is used, lethal a G V H D can happen whether conditioning is used or not used (Ellison et al., 1998; H i l l et al., 1998). In reports where no conditioning is used, a large number o f parent lymphoid cells are infused to unconditioned F l hosts. a G V H D develops and is followed by lethality 4 weeks after cell transfer (Ellison et a l , 1998). For nonirradiation conditioning approaches,  cyclophosphamide, is commonly used,  and  infusion o f a much smaller number o f parent lymphoid cells resulted in lethal a G V H D (Owens, Jr. and Santos, 1968; Rybka W B , 1994). Thus, based on these reported studies,  140  initial studies were performed first to establish a cyclophosphamide conditioned model for this study. Out o f the many parent and F l strain combinations, the B 6 parent and B 6 D 2 F 1 recipient combination is one o f the most well established ones. Lethal a G V H D has been reported in this donor and recipient combination whether conditioning is used or not. Thus, this combination was chosen for the initial study. In one experiment, a large number o f B 6 splenocytes ( 1 . 2 x l 0 /mouse) were infused into the unconditioned 8  B 6 D 2 F 1 mice, and clinical symptoms o f a G V H D , lethality and body weight were monitored. A s shown, in Figure 6.2, infusion o f this high number o f splenocytes did lead to transient a G V H D symptoms, such as hunched back, body weight loss, while recipients o f F l splenocytes did not show any o f these symptoms. B o d y weight was observed to drop from day 7 to day 20, and then recovered to that prior to cell transfer. Lethality only occurred in one o f 6 mice. In another experiment, recipient mice were first given cyclophosphamide at the dose o f 300 mg/kg intraperitoneally for conditioning 24 hours prior to cell transfer. Then, B 6 splenocytes were transferred to these conditioned recipients at the dose o f 4><10 /mouse. After cell transfer, clinical symptoms o f a G V H D , such as hunched back, 7  were observed as early as day 4 post cell transfer. B o d y weight showed persistent loss in recipients o f B 6 splenocytes, and lethality occurred i n 5 out o f 5 recipients. The body weight did not recover and this was well correlated with lethality (Figure 6.3). Moreover, histological studies o f the target organs demonstrated typical pathological changes o f a G V H D (Figure 6.4). Recipients o f F l splenocytes did not show any clinical symptoms o f a G V H D and no lethality was observed. Therefore, based on the  141  100' 03  75H  3 W5  50H  c  25H  SC Fl SC B6  u 0'  —r—  —i—  0  10  20  30  40  T i m e p o s t - t r a n s p l a n t (days) Figure 6.2. a G V H D in B 6 ^ u n c o n d i t i o n e d B 6 D 2 F 1 mouse model. 1 . 2 x l 0 B 6 splenocytes were intravenously transferred to unconditioned B 6 D 2 F 1 mice ( A , n=6/group). In the control group, same number o f F l splenocytes was transferred (•, n=3/group). Survival (A) and body weight (B) were monitored. A s shown in A , one out o f six mice died. Body weight o f recipients o f B 6 splenocytes decreased from day 7 to day 20, indicating a G V H D . However, after three weeks post cell transfer, body weight began to increase to that prior to cell transfer. N o lethality and body weight loss were observed i n the control group. 8  100  03  > a  SC Fl SC B6  75H  SC Fl SC  B6  50H  Vi  ~ 25H c  u su  —i—  PH  30  —i—  40  50  Time  60  0  10  20  —i—  —i—  —i—  30  40  50  - i  60  p o s t - t r a n s p l a n t (days)  Figure 6.3. a G V H D in B6^cyclophosphamide-conditioned B 6 D 2 F 1 mouse model. 4 x l 0 B 6 splenocytes were intravenously transferred to B 6 D 2 F 1 mice receiving 300mg/kg cyclophosphamide 24 hours ago ( A , n=6/group). In the control group, same number o f F l splenocytes was transferred (•, n=3/group) to cyclophosphamide conditioned F l mice. Survival (A) and body weight (B) were monitored. A s shown in A , five out o f five mice died. Body weight o f recipients o f B 6 splenocytes decreased persistently after cell transfer, indicating progressive severe a G V H D . In control group, no lethality was observed. Body weight o f mice in the control group dropped slightly at day 3 post cell transfer and then rose to level prior to cell transfer promptly. 7  142  Figure 6.4. Histology of target organs o f mice undergoing a G V H D . Recipient B6D2F1 mice were first injected with cyclophosphamide (300 mg/kg) and then injected with B 6 splenocytes (4><10 /mouse). M i c e i n the control group received conditioning and F l splenocytes. Survival was monitored. a G V H D target organs were removed at the time when mice died o f a G V H D . Tissues were fixed, sectioned and stained with hematoxylin and eosin. A and C. Liver and intestine of control mice respectively. N o pathological changes were observed. B and D . Liver and intestine o f mice undergoing a G V H D respectively. B . Marked lymphocyte infiltration o f the portal space was observed, which was typical pathology o f the liver related to a G V H D (arrow). D . Remarkable atrophic v i l l i and hyperplastic crypts with lymphocyte infiltration were observed in the intestine mucosa ( 125). 7  x  143  preliminary results, the model using cyclophosphamide as conditioning was chosen for the subsequent studies because o f the following advantages as compared to other models: 1) based on lethality, clinical symptoms, and histological studies, a G V H D in this model was more similar to that i n models where irradiation is used  for  conditioning; 2) lethal a G V H D occurred more consistently i n this model, suggesting that survival may be used as a reliable index for the severity o f a G V H D following treatment with Tregs; 3) persistent body weight loss was well correlated to a G V H D progression, making it also a reliable index for the severity o f a G V H D ; 4) since cyclophosphamide-conditioned mice were much more  sensitive to lethal shock  induction, fewer numbers o f donor cells were needed and the observation period could also be shortened; 5) cyclophosphamide is commonly used in human A H S C T clinically and thus this conditioning method is also clinically relevant.  6.3.2. Activation o f T S S T - l - i n d u c e d Tregs led to suppression o f a G V H D Since the studies in Chapter 4 showed that T S S T - l - i n d u c e d Tregs have potent ability to control proinflammatory cytokine responses, we asked i f these Tregs could also be used to inhibit a G V H D where proinflammatory cytokines play a critical role in its pathogenesis. To test this hypothesis, we used a well-characterized parent-to-Fl lethal murine a G V H D model, where splenocytes from B 6 mice receiving three injections o f TSST-1 were infused into cyclophosphamide-conditioned B 6 D 2 F 1 mice. A s shown in Figure 6.5-A, transfer o f 4 x l 0 splenocytes from F l mice, either TSST-1 7  or P B S primed, did not cause any clinical signs o f a G V H D and mice survived the 60-  144  A - A - T S S T - 1 B6 SC - A - PBS B6 SC - 0 - TSST-1 F l SC - • • - P B S F l SC  10  20  30  -1—  —1—  40  50  B 100-I--  60  70  - « - TSST-1 B6 SC + T(xl) - D - TSST-1 B6 SC + P(xl) - • - P B S B6 SC + T(xl) - O PBS B6 SC + P(xl)  50H i  25H  !  11  ^ — i i  - r  •k —t— 50  -  10  20  40  30  60  - • - T S S T - 1 B6 SC+T(x2) * - O - T S S T - 1 B6 SC+P(x2) *  - « - P B S B6 SC+T(x2) - O P B S B6 SC+P(x2)  10  20  i  i  i  30  40  50  i  —  60  70 Fig 6.5.  Time post-transplant (days)  145  21n  T i m e post-transplant (days) Figure 6.5. Activation o f TSST-l-induced Tregs attenuated a G V H D . B 6 D 2 F 1 ( H 2 ) recipient mice were conditioned by cyclophosphamide at the dose o f 300mg/mouse 24 hour before transfer o f B 6 (H-2 ) donor splenoytes. 4 x l 0 T S S T - l primed or PBS-primed B 6 or F l splenocytes, were transferred to the B 6 D 2 F 1 recipient mice intravenously 24 hours later. A . TSST-1 or PBS-primed B 6 or F l splenocytes were transferred to conditioned F l recipient mice and survival was monitored. Transfer o f TSST-1 (0) or PBS-primed (•) F l splenocytes did not result in a G V H D (n=5/group). A l l recipients o f T S S T - l - p r i m e d B 6 splenocytes ( A ) , which contained T S S T - l - i n d u c e d Tregs, died from a G V H D as recipients o f P B S primed B 6 spleocytes (A). Survival o f both group was not significantly different (15.6±0.3 vs 16.2±0.25 days, P>0.05, n=10/group). B . Recipient mice o f TSST-1 or PBS-primed B 6 splenocytes were given one injection o f TSST-1 (4u.g/mouse) or P B S at day 0. Recipients o f T S S T - l - p r i m e d B 6 splenocytes with one dose o f T S S T 1 ( • and dotted line) had significantly longer survival as compared to recipients o f T S S T - l - p r i m e d B 6 splenocytes with P B S injection ( • and dotted line) (26.3±1.9 vs 16.6±0.6 days, P O . 0 5 , n=10/group). Recipient mice o f PBS-primed B 6 splenocytes and given TSST-1 ( • and dotted line) or P B S (o and dotted line) all died from a G V H D (16.4±2.2 or 15.3±0.25 days, n=8/group). C . Recipient mice were given two doses o f TSST-1 or P B S after transfer o f TSST-1 or PBS-primed B 6 splenocytes. Survival time o f recipients o f T S S T - l - p r i m e d B 6 splenocytes with two doses o f TSST-1 at day 0 and day 4 (•) was significantly prolonged as compared to recipients of T S S T - l - p r i m e d B 6 splenocytes with injection o f two doses o f P B S (•) (52.1±5.3 vs 15.1±0.26, P O . 0 5 , n=10/group). Survival rate was also increased (80% vs 0%). Survival o f recipient mice o f PBS-primed B 6 splenocytes and two doses o f TSST-1 (•, n=9/group) or P B S (o, n=8/group) was also significantly shorter than that o f recipient o f T S S T - l - p r i m e d B 6 splenocytes and two doses o f TSST-1 (21.6±4.9 or 16.5±0.7 days, P O . 0 5 ) . *, P O . 0 5 , compared with T S S T - l - p r i m e d B 6 splenocytes+one/two dose/doses o f P B S , or PBS-primed B 6 splenocytes+/one/two dose/doses o f P B S , or PBS-primed B 6 splenocytes+one/two dose/doses o f T S S T - 1 . D . B o d y weight o f mice receiving B 6 splenocytes. M i c e receiving T S S T - l - p r i m e d B 6 splenocytes showed less severe body weight loss. Groups were the same as those inC. b d  b  7  146  day observation period (n=5/group). In contrast, adoptive transfer o f same numbers o f PBS-primed B 6 splenocytes resulted in clinical signs o f a G V H D , such as body weight loss and hunched- back, and all recipient mice died by 21 days post transplant (mean survival 16.2±0.25 days, n=10). Recipient mice o f T S S T - l - p r i m e d splenocytes, which contained TSST-1-indued Tregs, showed the similar morbidity and mortality as compared to recipient mice o f PBS-primed splenocytes (15.6±0.3 days, n=T0, P>0.05). Therefore, inclusion o f TSST-l-induced Tregs i n the graft alone did not confer protection" against a G V H D . The lack o f protection by the inclusion o f T S S T - l - i n d u c e d Tregs i n the graft led us to hypothesize that this might be due to the lack o f activation T S S T - l - i n d u c e d Tregs in vivo. W e then tested this hypothesis by giving one dose or two doses o f T S S T 1 to recipient mice post cell transfer. A s shown i n Figure 6.5-B, recipient mice o f T S S T - l - p r i m e d B 6 splenocytes receiving one dose o f T S S T - 1 at the time o f cell transfer had prolonged survival (26.3±1.9 days, n=10, P O . 0 5 ) as compared to mice receiving TSST-1  (16.6±0.6 days, n=10)  or PBS-primed  (15.3±0.25, n=8)  B6  splenocytes and one dose o f P B S , although all mice still died from a G V H D . This beneficial effect required the presence o f T S S T - l - i n d u c e d Tregs i n the graft, as recipients o f PBS-primed splenocytes receiving one dose o f TSST-1 all died from a G V H D (16.4±2.2, n=8). When two doses o f TSST-1 were given to mice receiving T S S T - l - p r i m e d splenocytes, the survival time was further significantly prolonged (52.1±5.3 days, n=10, P O . 0 5 ) as compared to that o f mice receiving T S S T - l - p r i m e d B 6 splenocytes and two doses o f P B S (15.1±0.26 days, n=9) or mice receiving P B S -  147  primed B 6 splenocytes and two doses o f P B S (16.5±0.7 days). A g a i n , this effect was not due to the two doses o f TSST-1 alone, as most o f the recipient mice o f PBS-primed B 6 splenocytes  receiving two doses o f TSST-1 still died from a G V H D by day 25  (21.6±4.9, n=9). Moreover, in contrast to one dose o f T S S T - 1 , mice receiving T S S T - l primed B 6 splenocytes and two doses o f TSST-1 had a much higher survival rate than those receiving T S S T - l - p r i m e d B 6 splenocytes and two doses o f P B S (80% vs 0%) (Figure 6.5-C). In addition, recipients o f T S S T - l - p r i m e d B 6 splenocytes and two doses of TSST-1 also demonstrated less severe body weight loss as compared to other control groups (Figure 6.5-D).  Therefore, these data showed that activation o f T S S T - l -  induced Tregs i n the graft could attenuate a G V H D .  6.3.3. T S S T - l - i n d u c e d CD4+ Tregs mediated suppression o f a G V H D To further confirm that the T S S T - l - i n d u c e d C D 4 + Tregs were responsible for this protection, C D 4 + T cells ( l x l 0 / m o u s e ) were isolated from splenocytes o f mice 7  treated with three doses o f TSST-1 and co-transferred with naive B 6 splenocytes (4xl0 /mouse) to B 6 D 2 F 1 recipient mice. In the control groups, C D 4 + T cell-depleted 7  T S S T - l - p r i m e d B 6 splenocytes ( 3 x l 0 /mouse) or PBS-primed B 6 C D 4 + T cells 7  ( l x l 0 / m o u s e ) were co- transferred with naive splenocytes to recipient mice. C o 7  transfer groups all received two doses o f TSST-1 at day 0 and day 4 since our earlier experiments have shown that activation o f T S S T - l - i n d u c e d Tregs led to control o f a G V H D . A s shown in Figure 6.6, co-transfer o f l x l O  7  T S S T - l - i n d u c e d C D 4 + Tregs  with two doses o f TSST-1 post cell transfer prolonged the survival o f recipient mice (46.2 ± 5.8 days, n=10, P O . 0 5 ) and increased the survival rate o f the recipient mice  148  100-  naive  naive+Treg+T(2x) naive+Treg-depleted SC+T(2x)  75H  naive+PBS-CD4+T(2x)  50H 254  JL  0-L 0  10  20  30  1  1  i  1  40  50  60  70  Time post-transplant (days)  Figure 6.6. Activation o f TSST-l-induced C D 4 + Tregs mediated.suppression o f a G V H D . B 6 D 2 F 1 recipient mice were conditioned with 300 mg/kg cyclophosphamide 24 hour before transfer o f B 6 (H-2 ) donor splenoytes. Recipient mice then were injected intravenously with 4 x l 0 naive B 6 splenocytes alone (•) or with l x l O T S S T - l - i n d u c e d Tregs (•), l x l O P B S primed B 6 C D 4 + T cells (0) or 3 x l 0 C D 4 + T cell-depleted T S S T - l - p r i m e d B 6 splenocytes (•). M i c e i n the later three groups all received two injections of TSST-1 (4 u.g/mouse) at day 0 and day 4. Survival time o f recipients o f T S S T - l primed C D 4 + T cells with two doses o f TSST-1 at day 0 and day 4 was significantly increased as compared to that o f recipients o f naive B 6 splenocytes alone (46.2±5.8 vs 16.4±1.8 days, P O . 0 5 , n=10/group). Co-transfer o f C D 4 + T cell-depleted T S S T - l - p r i m e d B 6 splenocytes (17.2±1 days) or PBS-primed C D 4 + T cells (17±0.61 days) did not confer protection, showing that T S S T - l induced C D 4 + Tregs were responsible for suppression o f a G V H D . *, P O . 0 5 , compared with naive alone, naive+Tregs-depleted T S S T - l - p r i m e d B 6 splenocytes + two doses o f T S S T - 1 , or naive+PBS-primed C D 4 + T cells+two doses o f T S S T - 1 . b  7  7  7  7  149  (60%) as compared to mice receiving naive splenocytes alone (16.4±1.8 days, 0% survival, n=10). In contrast, co-transfer o f PBS-primed C D 4 + T cells with two doses o f TSST-1 (17±0.61 days, n=10) or CD4+ T cell depleted T S S T - l - p r i m e d B 6 splenocytes with 2 doses o f TSST-1 (17.2±1 days, n=10) did not confer protection. Thus, these data further confirmed that donor derived T S S T - l - i n d u c e d C D 4 + Tregs when activated in vivo with two doses o f T S S T - 1 .  6.3.4. Activation o f TSST-l-induced Tregs did not affect donor T cell expansion or enhance host antigen presenting cell elimination in vivo A m p l e studies have shown that donor T cells and host A P C s are critical for a G V H D as host A P C s initiate and donor T cells maintain a G V H D . Thus it was logical to determine the effects o f activated TSST-induced Tregs on these cellular components. To this end, the splenocytes o f recipient mice were analysed at day 3, 7 and 14 posttransplant to track the donor T cells and host A P C s . Four groups were included i n this study. The experimental group consisted o f recipient mice receiving T S S T - l - p r i m e d splenocytes with two doses o f TSST-1 injection post transplant, and was previously shown to confer protection against a G V H D . Recipient mice o f T S S T - l - p r i m e d splenocytes with P B S injection post transplant served as a control group. The other two control groups were cyclophosphamide-conditioned mice receiving no cell transfer but TSST-1 injection or receiving T S S T - l - p r i m e d F l splenocytes with two injections o f T S S T - 1 . Donor and host cells could be distinguished by their expression o f H - 2  d  antigens, which was only expressed on host cells. Results in Figure 6.7 showed that there was no significant difference between the absolute numbers o f donor C D 4 and  150  A. Donor CD4+ T cells  e u o  T-B6 SC+T T-B6 SC+P  T-B6 SC+T:  «  T-B6 SC+P:  s  T:  i-  3  C "33  fl o  Day 3  Day 7  Day 14  T-Fl SC+T:  Recipients of TSST-lprimed B6 SC + two doses of TSST-1 Recipients of TSST-lprimed B6 SC + two doses of PBS No cell transfer + two doses of TSST-1 Recipients ofTSST-1primed F l SC + two doses of TSST-1  B. Donor CD8+ T cells  20-  CZ3T-B6 SC+T • l-llii SC+P  T  1510-  ss  5-  "3  0-  e  U  Day 3  fl  8 IO.O O  7.5'  i-  5.0  S  2.5'  Day 14  C. Host M H C 11+ APCs  a ve  1 ni  Day 7  -r  C H T - B 6 SC+T • T-B6 SC+P CUT  v X) 3  C rj  0.0'  • 1 T - F 1 SC+T  D  Day 3  Day 7  Time post-transplant (days)  Figure 6.7. Effect o f activation o f TSST-l-induced Tregs on donor T cells and host A P C s . B6D2F1 mice were conditioned with 300mg/mouse cyclophosphamide and transferred with T S S T - l - p r i m e d B 6 or F l splenocytes. T w o doses o f TSST-1 or P B S were given to the recipient mice. A t day 3, 7 and 14, spleens o f the recipient mice were removed and the percent o f donor C D 4 and C D 8 T cells and host A P C s were determined b y flow cytometry analysis. The absolute number o f these cells was then calculated. Donor cells were identified as H - 2 negative cells while host cells were H 2 positive. A P C s were identified as M H C Class II positive cells. A donor C D 4 T cell. B . donor C D 8 T cell. C . host A P C s . Four mice were included i n each group for each time. N o significant difference was found between the T - B 6 SC+T and T - B 6 SC+P groups in terms o f the number o f donor C D 4 and C D 8 T cells and host A P C s . d  d  151  C D 8 T cells recovered in the spleens o f recipient mice at all three time points, except for C D 8 T cells at day 7 when there was significantly higher number o f C D 8 T cells in recipients o f T S S T - l - p r i m e d B 6 splenocytes and two doses o f T S S T - 1 . These data suggested that activation o f TSST-l-induced Tregs did not affect the engraftment and expansion o f donor T cells in vivo. Analysis o f host A P C s by their expression o f M H C class II molecules also revealed no significant difference between the experimental group and the a G V H D control group (Figure 6.7-C).  Therefore, the activation o f  T S S T - l - i n d u c e d Tregs did not influence the donor T cell engraftment and expansion, and did not enhance the elimination o f host A P C s .  6.3.5. Activation o f T S S T - l - i n d u c e d Tregs modulated cytokine production and serum endotoxin level Proinflammatory T h l cytokines, such as T N F - a and IFN-y, have been shown to play critical roles in the pathogenesis o f a G V H D . Thus, it would be important to determine how the levels o f these cytokines were changed i n the presence o f activated T S S T - l - i n d u c e d Tregs. In the same experiment to study the effects o f activated Tregs on the donor T cells and host A P C s , sera o f recipients were obtained when the mice were sacrificed. Serum levels o f I L - 2 , IL-12, IL-10, T N F - a and IFN-y were then assayed by E L I S A . A s shown i n Figure 6.8-A, serum level o f T N F - a in both experimental and a G V H D control groups increased from day 3 to day 14, with no difference detected at each time point between these two groups. A t day 7 and 14, levels o f T N F - a in the experimental group appeared to be higher than that o f the a G V H D group, although the difference was not statistically significant. In contrast,  152  levels o f serum IFN-y in the experimental group (61.5±28.6 pg/mL, n=4, P<0.05) were significantly lower than i n the a G V H D group (280±44 pg/mL, n=4) at day 7 post transplant. These data suggest that activation o f T S S T - l - i n d u c e d Tregs might control a G V H D by down-regulating levels o f IFN-y. Amounts o f IL-10 were not significantly different between the experimental and a G V H D groups at all time points. In the control groups, no cell transfer or transfer o f T S S T - l - p r i m e d F l splenocytes with two doses o f TSST-1 did not lead to detectable levels o f T N F - a and IFN-y, confirming that the increased levels o f T N F - a and IFN-y in the other two groups were due to alloresponses mediated by donor cells. Serum levels o f IL-2 and IL-12 were below detection levels (data now shown). In vitro mixed lymphocyte reactions ( M L R ) using T cells recovered from the recipient mice at different time points post transplant as responder cells further confirmed the effect o f activated T S S T - l - i n d u c e d Tregs on the production o f cytokines induced by alloresponses. A s shown in Figure 6.8-B, when T cells from recipient mice o f T S S T - l - p r i m e d B 6 splenocytes and two doses o f TSST-1 were stimulated by F l A P C s , production o f IL-2, IL-12 and LFN-y was significantly lower than those o f T cells from the a G V H D group at day 7 and day 14 post transplant. T N F - a production was not affected. IL-10 level was higher in the experimental group than in the a G V H D group only at day 3 post transplant. Endotoxin translocation has been shown to be a critical step i n the pathogenesis o f a G V H D . Efforts were made to study how the activated T S S T - l - i n d u c e d Tregs affected endotoxin translocation. A s shown in Figure 6.8-C, serum levels o f endotoxin in recipient mice o f the experimental group were significantly lower than those o f the  153  — I  Day 3  1  Day 7  " l  Day 14  T i m e post-transplant (days)  Figure 6.8. Effect o f activated T S S T - l - i n d u c e d Tregs on cytokine production in vivo and in vitro and serum endotoxin levels. A . In the same experiment described i n F i g 6.7, sera were collected at each time point when mice were sacrificed to remove the spleens o f recipient mice. Serum levels o f cytokines were determined b y E L I S A . A . A s shown, no significant difference o f serum T N F - a between recipients o f T S S T - l - p r i m e d B 6 splenocytes and two doses o f TSST-1 (•, experimental group) and o f T S S T - l - p r i m e d B 6 splenocytes and two doses o f P B S ( • , a G V H D group) at three time points post transplant. In contrast, significant lower level o f serum LFN-y was detected in the experimental group as compared to the a G V H D group (61.5±28.6 vs 2 8 0 ± 4 4 pg/mL, P O . 0 5 , n=4). I L 10 levels o f the experimental group and a G V H D group were not significantly different at all time points. *, P O . 0 5 , compared with T - B 6 SC+P (2*). T ( T ) : N o cell transfer + two doses o f T S S T - 1 ; T - F l S C + T (•): Recipients o f T S S T - l primed F l S C + two doses o f T S S T - 1 .  154  350  TIL-2  300  300-1  CUB  250  200-  150 IOOH  0  1 1 III nlll Day 3  600 500  Day 7  I  Mi' til c = i A  I  JL  T  I  Day 3  Day 14  ft  IFN-y  A  CD B  -111 ill  200  50  •  IL-12  Day 7  -i *  IL-10  IB  Q  300-  A  CUB  750'  400-  i  Day 14  500  200-  250'  100-  0  Day 3  200  Day 7  150-1 100  50 0  Day 3  Day 7  Day 14  Day 14  IC  X  Day 3  Day 7  Day 14  T i m e post-transplant (days)  Groups  Responder T  A  cells  B 6 T  B  C  from recipients o f  splenocytes  cells  B 6  cells  + two  Na'ive B 6 T  + two  T S S T - l - p r i m e d  doses o f  from recipients o f  splenocytes  Stimulator  T S S T - 1  cell-depleted  F l  splenocytes  T S S T - l - p r i m e d  doses o f  T  cells  P B S  T  cell-depleted  F l  splenocytes T  cells  cell-depleted  F l  splenocytes  Figure mice doses  6.8.  were  B . M i x e d  transferred with  of T S S T - 1  spleens  o f  b y  T S S T - l - p r i m e d  or P B S . A t different time  recipient  treated F l o r B 6 assayed  l y m p h o c y t e reactions.  mice  were  splenocytes  E L I S A .  Splenic  when  cells  of  stimulated mice  from  with  B 6  splenocytes  points with  post  T  and  received  transplant, T  cell-depleted  cells  T  cells  mitomycin  a G V H D  g r o u p at e a c h t i m e . * , P O . 0 5 ,  containing activated T S S T - l - i n d u c e d  levels o f I L - 2 , I L - 1 2 a n d I F N - y  allogeneic group  stimulator cells (group  group A  B).  4  as  mice  at d a y  were  included  C -  were Tregs  7 and  c o m p a r e d to  compared with group B ,  two from  for 5 days. C y t o k i n e levels in the supernatants  (group A ) produced decreased 14  stimulated  Cyclophosphamide-conditioned F l  day  splenic for  T  each  n=4.  155  Figure 6.8. C . Serum levels o f endotoxin i n recipients o f activated T S S T - l induced Tregs. In separate experiments as described i n F i g 6.7, sera were collected at each time point and serum endotoxin level was determined. A s shown, serum endotoxin level in recipient mice o f T S S T - l - p r i m e d B 6 splenocytes plus two doses o f TSST-1 (•) was significantly lower than that o f the a G V H D mice (•).*, P O . 0 5 , compared with • : T - B 6 SC+P (2x), n=4. T ( T ) : N o cell transfer + two doses of T S S T - 1 ; T - F l S C + T (•): Recipients ofTSST-l-primed F l S C + two doses o f T S S T - 1 .  156  a G V H D group at day 14 post transplant, indicating that at this point i n time, the presence o f activated T S S T - l - i n d u c e d Tregs resulted i n suppression o f a G V H D leading to attenuated tissue injury o f the target organs and diminished translocation o f endotoxin from the gut.  6.3.6. IL-10 blockade with IL-10 receptor monoclonal antibody did not reverse the suppression mediated by T S S T - l - i n d u c e d Tregs It has been shown that suppression mediated by T r l cells in vivo, which is dependent on the production o f IL-10, was not associated with increased serum IL-10 levels (Cottrezet al., 2000). In addition, elevated IL-10 levels were detected i n the mixed lymphocyte reaction described i n Figure 6.7-B, suggesting that IL-10 might play a role i n the suppression mediated by activated T S S T - l - i n d u c e d Tregs on a G V H D . Thus, the r o l e . o f IL-10 was further studied using a monoclonal antibody, I B 1.2, against the IL-10 receptor which has been previously shown to block the actions o f I L 10. A s shown i n Figure 6.9, recipients o f T S S T - l - p r i m e d B 6 splenocytes with two doses o f P B S all died from a G V H D by day 20 (14±0.45 days, n=8) while recipients receiving the same cells with two doses o f TSST-1 showed prolonged survival (55±5 days, n=8) and increased survival rate (87.5%). Administration o f anti-IL-10 receptor antibody did not shorten the survival (55±4.4 days, n=8) time or decrease the survival rate (87.5%) o f the recipients as compared to recipients without the antibody or recipients o f control antibody. Therefore, these data suggest that activated T S S T - l induced Tregs did not depend on IL-10 exclusively to exert their suppression on a G V H D in vivo.  157  El  •  -•-T-B6 SC+P(2x) - Q -«-T-B6SC+T(2x) - A - T-B6 SC+T(2x)+anti-rL-10R Ab -O- T-B6 SC+T(2x)+control Ab  •  o  i i i  25H  0 i  q A, —  — i 1 1 1 1 1 0 10 20 30 40 50 60 70 T i m e post-transplant (days) Figure 6.9. Role o f IL-10 in the suppression mediated by T S S T - l - i n d u c e d Tregs on a G V H D . Conditioned B 6 D 2 F 1 recipients mice were transferred with 4*10 T S S T - l - p r i m e d B 6 splenocytes and received two doses o f T S S T - 1 . They were further injected with anti-IL-10 receptor antibody ( A , lmg/mouse at day 0 and 0.5mg/mouse/week thereafter). M i c e receiving T S S T - l - p r i m e d B 6 splenocytes and two doses o f P B S served as a G V H D positive group (•). Other control groups included mice receiving T S S T - l - p r i m e d B 6 splenocytes and two doses o f TSST-1 (•) and mice receiving T S S T - l - p r i m e d B 6 splenocytes, two doses o f TSST-1 and control antibody ( o ) . A s shown, mice i n a G V H D group (•) died by 20 days posttransplant while mice receiving T S S T - l - p r i m e d B 6 splenocytes and two doses of TSST-1 had prolong survival (•). Neutralization o f IL-10 did not reverse the suppression by activated TSST-l-induced Tregs ( A ) as compared to group without neutralization (•) or mice receiving control antibody ( o ) . 7  158  6.3.7. Control o f a G V H D by activation o f T S S T - l - i n d u c e d Tregs did not block graftvs-tumor ( G V T ) response a G V H D and G V T effect are two closely related processes that are triggered by alloresponses. Data above demonstrated that activation o f T S S T - l - i n d u c e d Tregs was able to suppress a G V H D mediated by donor T cells. Thus, it was important to further investigate i f G V T effects were preserved or not because G V T effects have been shown to be associated with a favourable outcome i n clinical A H S C T . To address this issue, l x l O P815 tumor cells, which are derived from D B A 2 mice, were added to the 4  inoculums for each recipient mouse. Results demonstrated that recipients o f T S S T - l primed F l splenocytes with two doses o f TSST-1 did not confer protection against the tumor attack. Lethality began from 11 days post transplant and all recipient mice died by day 14 (11.8=1=0.3 days, n=8). Due to syngeneic transplantation mice did not suffer from a G V H D as reflected by the l a c k . o f clinical signs and severe body weight loss typical o f a G V H D (Figure 6.10-A and B ) . Histological study and autopsy demonstrated the typical infiltration o f tumor in the liver (Figure 6.11-B). Recipients o f T S S T - l primed B 6 splenocytes with P B S survived a little longer, demonstrating the presence o f G V T effects, however, they all died from a G V H D (15±0.38 days, n=8) as revealed by clinical signs and severe body weight loss (Figure 6.10-B). Autopsy and histological study did not reveal the typical tumor nodules i n the liver (Figure 6.11-C). In contrast, recipients o f T S S T - l - p r i m e d B 6 splenocytes had significantly prolonged survival as compared to the two control groups described above (45.6±5.3 days, n=10, P O . 0 5 ) . Three out o f ten mice died from a G V H D , as determined by clinical signs and the absence o f infiltration o f tumor cells to the liver. Three mice ultimately died from  159  * - T-B6 SC+T(2x)+P815 O - T-B6 SC+P(2x)+P815 • - T - F 1 SC+T(2x)+P815 T-Fl SC+T(2x)  100 75H cs  > 50H  c ii u s-  25H  4  i -LX.  OH 10  20  30  —r-  40  22.5-  — l  -  60  70  50  B 20.0- • - T - B 6 SC+T(2x)+P815 -B- T-B6 SC+P(2x)+P815 - • - T - F l SC+T(2x)+P815 - 0 - T - F l SC+T(2x)  .SP17.S-  V, 15.0tt 12.5-  0  3  7  14  21  28  35  42  49  56  63  Time post-transplant (days) Figure 6.10. Effect o f activated T S S T - l - i n d u c e d Tregs on G V T response. After conditioning with 300mg/kg cyclophosphamide, recipient F l mice were injected with T S S T - l - p r i m e d B 6 splenocytes and given two doses o f TSST-1 or P B S , or with T S S T - l - p r i m e d F l splenocytes and given two doses o f T S S T - 1 . 10,000 P815 cells were added to each mouse. A . Survival o f mice. Recipient mice o f T S S T - l primed F l splenocytes, two doses o f TSST-1 and P815 cells (•) all died from tumor progression b y day 14 (11.8±0.3 days, n=8). Survival o f recipients o f T S S T - l primed B 6 splenocytes, two doses o f P B S and P815 cells (•) was slightly prolonged (15±0.38 days, n=8), but they all died from a G V H D . In contrast, recipients o f T S S T l-primed B 6 splenocytes, two doses o f TSST-1 and P815 cells (•) were significantly prolonged (45.6±5.3 days, n=10, P O . 0 5 ) as compared to the other groups. Three mice died from a G V H D and three mice died from tumor progression at day 50. Thus, activation o f T S S T - l - i n d u c e d Tregs could control a G V H D while sparing G V T response. *, P O . 0 5 , compared with T - F l SC+T(2x)+P8 15(D). Control mice receiving T S S T - l - p r i m e d F l splenocytes and two doses o f T S S T - l without tumor cells (0) were healthy and survived the whole observation period. B . Body weight o f the mice as a reflection o f a G V H D severity. Recipient mice o f T S S T - l - p r i m e d F l splenocytes, two doses o f T S S T - l with (•) or without (0) P815 cells did not showed persistently severe body weight loss, indicating absence o f a G V H D . Recipients o f T S S T - l - p r i m e d B 6 splenocytes, two doses o f T T S T - 1 and P815 cells (•) showed less severe body weight loss as compared to a G V H D recipients o f T S S T - l - p r i m e d B 6 splenocytes, two doses o f P B S and P815 cells (•), indicating attenuated a G V H D .  160  Figure 6.11. Histological study o f tumor infliltration o f the liver undergoing tumor challenge. Liver tissues were taken from experiments described i n F i g 6.10. Typical tumor nodules were formed in the livers o f recipients o f T S S T - l - p r i m e d F l splenocytes, two doses o f TSST-1 and P815 tumor cells ( B , arrow pointed to the tumor nodule). Tumor nodules were not found i n recipient mice o f T S S T - l primed B 6 splenocytes, two doses o f either TSST-1 or P B S and P815 tumor cells (C, D ) , or in recipient mice o f T S S T - l - p r i m e d F l splenocytes and two doses of TSST-1 without tumor cells (A). H . E stain, x l 2 5 .  161  tumor progression at day 50 as they had tumor infdtration o f the liver but lacked clinical signs o f a G V H D . Surviving mice did not show any signs o f a G V H D or tumor infiltration o f the liver at day 60 post-transplant. The overall less severe body weight loss o f this group further demonstrated the attenuated a G V H D (Figure 6.10-B) and prolonged survival against tumor infiltration. Histological studies o f the liver taken from this group also did not reveal tumor nodule formation (Figure 6.11-D). Recipients o f T S S T - l - p r i m e d F l splenocytes without tumor cells did not show any signs o f illness through the whole observation period except a slight body weight loss induced by the conditioning  protocol, which recovered  promptly  by  day  7  (Figure  6.10-B).  Histological study did not reveal any pathological changes i n the liver (Figure 6.11-A). The cause o f death o f each group was summarized in Table 6.1. Thus, activation o f T S S T - l - i n d u c e d Tregs could attenuate a G V H D while sparing G V T effects mediated by the allogeneic donor T cells.  Table 6.1. Summary o f cause o f death in each group o f mice. GVHD mortality  Tumor mortality  Overall Survival  T - F l SC+T(2x)  0/4 (0%)  0/4 (0%)  4/4(100%)  T - F l SC+T(2x)+P815  0 (0%)  8/8 (100%)  0 (0%)  T-B6 SC+T(2x)+P815  3/10(30%)  3/10 (30%)  4/10 (40%)  T-B6 SC+P(2x)+P815  8/8  0 (0%)  0 (0%)  162  6.4. Discussion Our data showed that transplantation o f allogeneic graft containing T S S T - l induced Tregs with two doses o f TSST-1 administration post-transplant remarkably attenuated the mortality and morbidity o f a G V H D using a parent-to-Fl murine model. This was due to the active suppression mediated by the T S S T - l - i n d u c e d C D 4 + Tregs rather than the effects o f superantigen administration alone. In a previous study, Takenaka et al showed that administration o f S E B could attenuate the graft-vs-host response due to the deletion o f SEB-reactive T cells bearing V p 8 which were alloreactive in unconditioned B 6 D 2 F 1 recipients transplanted with B 6 donor cells (Takenaka et al., 2001). However, in our model, C D 4 + T cells bearing V p i 5 were TSST-1 reactive but not alloreactive (Patterson and Korngold, 2001). Deletion or anergy induced by TSST-1 on donor T cells was not responsible for the attenuation o f a G V H D . This was further supported by the finding that recipients o f nai've allogeneic donor T cells plus two doses o f TSST-1 were not protected against a G V H D . Moreover, the data that co-transfer o f TSST-l-induced C D 4 + Tregs but not the C D 4 + T cell depleted T S S T - l - p r i m e d splenocytes protected the recipient mice further demonstrated the presence o f active suppression mediated b y T S S T - l - i n d u c e d Tregs on a G V H D . The  lack o f alloresponses mediated by C D 4 + T S S T - l - i n d u c e d Tregs might also  account for the requirement o f cognate antigen stimulation for their suppressive effects, as these cells were not stimulated by alloantigens at an early stage after transplantation. Therefore, our results demonstrated that T S S T - l - i n d u c e d C D 4 + Tregs could mediate active bystander suppression on a G V H D .  163  Bystander suppression has been noted i n multiple animal disease models. For example, the typical T r l cells could mediate bystander suppression o f colitis (Groux et al., 1997), Th2 response (Cottrez et al., 2000) or immune-mediated cardiovascular pathologies (Mallat et al., 2003) only when the mice were fed or immunized with O V A to activate the T r l cells. IL-10 was required for T r l cells to exert the bystander suppression effects as neutralization o f IL-10 blocked the bystander  suppression  (Cottrez et al., 2000; Groux et al., 1997). Oral tolerance has also been shown to elicit potent bystander suppression on autoimmune diseases, such as experimental allergic encephalitis (Chen et al., 1996). One o f the mechanisms o f inhibition o f autoimmune diseases by oral tolerance has also been shown to be the induction o f Tregs, such as Th3  cells which exerted their functions by TGF-p (Chen et al., 1994). In'the  transplantation context, it has recently been shown that induction o f CD4+CD25+ Tregs specific for a model antigen in recipients could prevent rejection o f the M H C mismatched donor skin graft (Karim et al., 2005). Hashimoto et al demonstrated that activation o f N K T cells by their cognate ligand could also attenuate a G V H D (Hashimoto et al., 2005). What these Tregs had in common was that the Tregs must be activated by their cognate antigens to suppress a response elicited b y antigens different from their cognate antigens. Despite these studies, it is not known i f a G V H D could be suppressed via bystander inhibition by conventional ap C D 4 or C D 8 T cells. Thus our study was the first to demonstrate that a G V H D could be controlled via bystander inhibition mediated by conventional ap C D 4 Tregs induced by a bacterial superantigen. Engraftment, activation and expansion o f allogeneic donor T cells by host A P C s are the critical steps i n the initiation o f a G V H D . Measures to block this  164  interaction have been shown to effectively inhibit a G V H D . F o r example, L a n et al showed that using a conditioning protocol that enriched the host N K T cells were able to prevent donor T cell engraftment and expansion (Lan et al., 2001). In addition, cotransfer o f ex vivo expanded allospecific natural C D 4 + C D 2 5 + Tregs was able to suppress the expansion o f donor T cells, resulting i n the suppression o f a G V H D (Trenado et al., 2006). The central role o f host A P C s i n initiating a G V H D was best demonstrated by the study showing that recipient mice, whose A P C s had been replaced by donor type A P C s , were resistant to lethal a G V H D induction (Shlomchik et al., 1999). However, our data did not reveal a role o f T S S T - l - i n d u c e d Tregs i n blocking these initiating steps o f lethal a G V H D , as no differences have been found i n terms o f expansion o f donor T cells and the host A P C s . This is i n line with our in vitro studies in Chapter 5 on the mechanisms o f T S S T - l - i n d u c e d Tregs, where it was found that T S S T - l - i n d u c e d Tregs did not inhibit the proliferation o f their target cells in vitro. Interestingly, a more vigorous expansion o f C D 8 + donor T cells was detected at day 7 in the recipients o f TSST-l-induced Tregs receiving two doses o f T S S T - 1 . A possible explanation is that TSST-l-induced Tregs were able to enhance the proliferation o f C D 8 + T cells because data in vitro i n Chapter 5 showed that naive splenocytes cocultured with T S S T - l - i n d u c e d Tregs showed enhanced proliferation! Further work is needed to elucidate the exact cause o f this observation. Despite the enhanced expansion o f donor C D 8 + T cells, a G V H D in this group was not deteriorated, suggesting a minor role o f C D 8 + T cells in mediating a G V H D in this model. Therefore, it was further examined i f T S S T - l - i n d u c e d Tregs could affect the production o f cytokines and endotoxin translocation that are important  in the  165  pathogenesis o f a G V H D (Reddy and Ferrara, 2003). Serum LFN-y was found to be down-regulated at day 7 post-transplant while serum levels o f T N F - a were not affected at any time points assayed. In vitro mixed lymphocyte reactions confirmed that the presence o f activated TSST-l-induced Tregs could down-regulate the proinflammatory cytokines induced by allo-responses, namely IL-2, IL-12 and LFN-y. The role of these proinflammatory  cytokines in a G V H D  is critical.  These  Thl  proinflammatory  cytokines have been found to potentiate a G V H D . For instance, IFN-y can prime monocytes to produce T N F - a , increased adhesion molecule expression by endothelial cells and promote C T L generation, thus leading to progression o f a G V H D . In addition, pro-inflammatory cytokines could also directly cause tissue injury o f the target organs (Reddy and Ferrara, 2003). Multiple methods which suppress a G V H D have been shown to be associated with the inhibition o f proinflammatory cytokine production (Edinger et al., 2003; H i l l et al., 1998). Our data also suggest that T S S T - l - i n d u c e d Tregs can suppress a G V H D by down-regulation o f T h l proinflammatory cytokines. Moreover, previous work by Ferrara's group demonstrated that i n a parent-to-Fl murine a G V H D model using B 6 and B 6 D 2 F 1 donor and recipient combination, a G V H D was mediated by CD4+ T cells (Teshima et al., 1999). Thus taken together, our data suggest that the suppressive effects o f T S S T - l - i n d u c e d Tregs were exerted on donor C D 4 + T cells. Endotoxin translocation is another important factor i n the progression o f a G V H D . Host conditioning or alloresponses can lead to gastrointestinal tract damage, which results in leakage o f endotoxin into the systemic circulation. The translocated endotoxin can greatly potentiate the alloresponses. This positive feedback eventually  166  results in the progression o f a G V H D ( H i l l et al., 1997; H i l l et al., 1999; H i l l and Ferrara, 2000). A m p l e evidence has been accumulated that interruption o f this vicious circle can attenuate a G V H D (Cooke et a l , 2001; H i l l et al., 1998; Krijanovski et al., 1999). Our data showing that recipient mice o f activated T S S T - l - i n d u c e d Tregs had both suppressed production o f proinflammatory cytokines and lower serum level o f endotoxin suggested that TSST-l-induced Tregs attenuated a G V H D by preventing the potentiative effects of both proinflammatory cytokines and translocation o f endotoxin. One interesting observation in this study is that the production o f T N F - a was not affected by the activated T S S T - l - i n d u c e d Tregs. T N F - a has been shown to be an important proinflammatory cytokine i n the pathogenesis o f a G V H D , as blockade with anti-TNF-a could prevent lethality o f a G V H D (Teshima et al., 2002). T N F - a derived from donor T cells (Schmaltz et al., 2007), C D 4 + T cells i n particular (Ewing et al., 2007),  plays an important role in enhancing a G V H D . One possibility for our  observation is that i n addition to T N F - a produced by alloreactive donor T cells, it was also produced by the non-alloreactive T S S T - l - i n d u c e d Tregs as data i n Chapter 4 showed that these Tregs expressed and secreted T N F - a when they were reactivated in vitro. It was also demonstrated that C D 8 + donor T cells could also be a source o f T N F - a i n a G V H D (Ewing et al., 2007). Since G V T effects, which is likely mediated by C D 8 + donor T cells (Hill et al., 1999), were not affected i n our system, it is highly probable that T S S T - l - i n d u c e d Tregs did not inhibit C D 8 + T cell function, and thus production o f T N F - a by C D 8 + T cells might not be suppressed. Even though the production o f T N F - a was not suppressed by TSST-1 induced Tregs, a G V H D was definitely attenuated, suggesting that other proinflammatory cytokines (such as IFN-y)  167  might be more important, either alone or i n combination with T N F - a , to mediate the pathology o f a G V H D in target organs (Reddy and Ferrara, 2003). Although T N F - a is a critical mediator o f a G V H D , it also plays a pivotal role i n G V T effects ( H i l l et al., 1999;  Schmaltz et al., 2007). Thus it w i l l be worthwhile to further determine the source  of T N F - a and its role in a G V H D and G V T effects i n our system i n further studies. It has been  well  documented  that IL-10-producing Tregs  can regulate  transplantation rejection. H i l l et al recently described IL-10-producing Tregs induced by dendritic cells harvested in donor administered with G - C S F . These Tregs were able to control a G V H D while sparing G V T effects (Morris et al.,2004). In our study, it was shown that T S S T - l - i n d u c e d Tregs produced high levels o f IL-10. Although serum I L 10 was not elevated, it could still function locally. Administration o f anti-IL-10 receptor antibody, however, did not reverse the suppression mediated b y activated T S S T - l - i n d u c e d Tregs on a G V H D , suggesting that these Tregs did not depend on I L 10 to exert their functions in vivo. Alternately, the systemic administration o f anti-IL10R antibody may not have been locally effective. These data were also i n line with results described i n chapter 5, where it was shown that blockade o f IL-10 did not reverse the suppressive activities o f T S S T - l - i n d u c e d Tregs in vitro. Although T S S T - l - i n d u c e d Tregs suppressed a G V H D , our data showed that the G V T response was preserved. One major reason might be the production o f T N F - a was not affected. H i l l et al has shown using the same combination o f donor and recipient mice that T N F - a played an essential role i n mediating G V T effects against P815 cells since neutralizing T N F - a abrogated G V T effects ( H i l l et al., 1999). Moreover, the data showing that the engraftment and expansion o f both C D 4 + and C D 8 + donor T cells  168  were not affected by T S S T - l - i n d u c e d Tregs suggested that they might still mediate alloresponses that resulted i n G V T effects, although the alloresponses were attenuated leading to suppression o f lethal a G V H D . This is another critical contributing factor to the G V T effects in this study. It was also shown by Ferrara's group in a G V H D using the same donor and recipient combination that G V T effects against P815 cells were dependent on both C D 4 + and CD8+ T cells, with C D 8 + T cells mediating more potent G V T effects and C D 4 + T cells providing help to achieve the optimal G V T effects (Teshima et al.,  1999). The preserved G V T effect i n our study was probably due to the  unaffected C D 8+ T cell response. Although the C D 4 + T cell response was attenuated by T S S T - l - i n d u c e d Tregs, these cells were still able to provide sufficient help to aid the G V T response while they were prevented from mediating lethal a G V H D . One drawback o f the use o f powerful immunosuppressants such as cyclosporin is their non-specific immunosuppressive activities. Although they are able to suppress G V H D , they may also blunt the G V T responses. In some clinical settings, intentional withdrawal o f immunosuppressive drug was actually used to allow G V T effects. However, this w i l l also cause the occurrence o f uncontrolled a G V H D . In our study, since T S S T - l - i n d u c e d Tregs were not alloreactive, it is possible that activation o f this population may lead to non-specific immunosuppression similar to that caused by immunosuppressive drugs. However, the presence  o f G V T effects,  despite  the  inhibition o f a G V H D , suggested that unlike these immunosuppressants, activation o f this population suppressed the pathogenic responses leading to a G V H D specifically while leaving favourable G V T responses intact. It remains to be determined whether  169  activation o f T S S T - l - i n d u c e d Tregs affects the ability o f immune systems to fight infections. This should clearly be further investigated i n the future. In summary, we demonstrated that C D 4 + Tregs induced b y TSST-1 were able to control lethality o f a G V H D v i a bystander inhibition while sparing G V T responses. This may have important clinical implications. Although human natural CD4+CD25+ Tregs are able to suppress allo-responses and murine studies have confirmed their potential use i n the clinical setting, the scarcity o f these cells in vivo mandates extremely laborious ex vivo expansion. In contrast, as superantigens  are potent  stimulators for T cells and can activate large numbers o f T cells, they may have the potential to generate large number o f Tregs in vitro with relative ease. Indeed, we did show previously that human Tregs can be generated in vitro b y repeated stimulation o f peripheral blood mononuclear cells during a 5-day period ( K u m W W et al., 2006). Our study w i l l further raise the possibility o f using donor-derived T S S T - l - i n d u c e d Tregs to prevent a G V H D . However, the requirement to administer cognate superantigen for the suppressive functions o f these cells w i l l make this therapy impossible i n the clinical settings, as superantigens themselves may cause potentially lethal effects. Nevertheless, it has been shown that mutants o f superantigens with T cell stimulation ability but without causing lethality can be generated (Dinges et a l , 2000). This may be a potential solution for the clinical use o f superantigen-induced Tregs i n a G V H D , and clearly warrants further study.  170  Chapter 7: General discussion, conclusions and future directions 7.1. General discussion Tregs have been one o f the major developments i n immunology i n recent years. Research i n this field has generated insightful knowledge o f the mechanisms o f immune tolerance, leading to a better understanding o f physiological processes and the pathogenesis o f many diseases as well as novel therapeutic strategies for these diseases. Thus, experimental systems to induce Tregs are important to advancing the field o f research in Tregs. One problem with inducing Tregs with nominal antigens is that they can only activate a very low number o f antigen-specific T cells, making it difficult to track and study these antigen-induced Tregs. In this case, use o f repeated injection o f superantigens  to  induce Tregs may provide a solution for this problem, as  superantigens have potent stimulatory effects on T cells and are able to activate a large number o f superantigen-reactive T cells, making it easier to generate antigen-reactive Tregs. Moreover, due to the VP specific effects, the VP segments can also serve as useful markers to detect T cell responses. However, due to the differences o f recognition o f superantigens by T cells from that o f nominal antigens, more studies are needed to elucidate i f the biologic properties o f superantigen-induced Tregs are different from those induced by nominal antigens. Furthermore, as this study has indicated, Tregs induced by different  superantigens  may also display different  biological properties. Results o f this study showed that repeated injection o f TSST-1 induced Tregs which share properties similar to T r l cells, such as the production o f high levels IL-10. However, many o f the characteristics were different from those o f T r l , for instance, the hyperproliferative profile and their independence on IL-10 to  171  exert the suppressive functions. This implied that the generation o f TSST-l-induced Tregs was associated with unique signals different from those i n the induction o f T r l cells. Thus the ability o f TSST-1 to induce Tregs with distinctive properties may offer an excellent opportunity to further investigate the unique signals that drive the development o f these Tregs. Although quite a lot o f research has been done i n the field o f Tregs, progress in elucidating the mechanisms o f Tregs is still slow. It has been well established that immunosuppressive cytokines, such as IL-10 and T G F - P , are main players in mediating suppression by Tregs. However, these mediators could not completely explain the mechanisms o f suppression in many cases, especially contact-dependent suppression. Although cytokine competition was responsible for the regulation o f I L - 2 , other molecules may be involved in modulating other cytokines, which were not identified i n this study. These data demonstrate that unlike T r l cells, the cytokine profile o f T S S T l-induced Tregs might not be causally related to their suppressive function. Identifying molecules mediating the suppressive function o f T S S T - l - i n d u c e d Tregs w i l l not be an easy task. Advances in molecular techniques may be helpful i n this regard. A s unique properties have been found i n the mechanisms o f T S S T - l - i n d u c e d Tregs, use o f these cells for proteomic analysis may be helpful i n identifying more novel molecules responsible for their suppressive functions. The promise o f potential therapeutic applications o f Tregs is one o f the main reasons that fuel research in this area. A s mentioned in chapter 1, a lot o f work has previously focused on natural Tregs. However, many problems have been identified, such as the difficulty to obtain sufficient number o f natural Tregs for study and the  172  existence o f functional defects in some situations. Antigen-induced Tregs can be a very useful alternative for study. In this thesis research, it was clearly shown that T S S T - l induced Tregs were able to exert bystander inhibition o f a G V H D in vivo while sparing G V T effects. These data are in line with other reports showing that bystander suppression mediated by antigen-induced Tregs has great potential for the treatment o f multiple diseases. Several important issues have to be addressed regarding the use o f superantigen-induced Tregs in the clinical setting. First, more work needs to be done to determine i f T S S T - l - i n d u c e d Tregs can be applied to treat other diseases, such as autoimmune diseases. Second, it remains to be determined i f other TSST-1 mutants are also able to generate Tregs with the same ability to suppress a G V H D or other diseases. Third, efforts should be made to generate non-toxic TSST-1 mutants for clinical application. A s TSST-1 has been implicated in the cause o f T S S , it may not be possible to use T S S T - l - i n d u c e d Tregs in humans as these Tregs need to be reactivated by their cognate antigens. However, it is possible to generate superantigen mutants that have T cell stimulatory effects but have lost the ability to cause lethal shock. Generation o f such mutant  superantigens  w i l l be a necessary  step i n promoting the  clinical  application o f T S S T - l - i n d u c e d Tregs. Taken together, knowledge gained i n this study o f T S S T - l - i n d u c e d Tregs not only extends our understanding o f superantigen-induced Tregs, but also sheds new light on potential therapeutic applications o f T S S T - l - i n d u c e d Tregs.  173  7.2. Conclusions Based on the results o f this study, the following conclusions can be drawn: •  Repeated subcutaneous injection o f TSST-1 to C 5 7 B L / 6 mice induced C D 4 + Tregs with potent suppressive activities. These Tregs could downregulate the TSST-1-triggered proinflammatory cytokine responses o f na'ive splenocytes in vitro and inhibit TSST-l-induced lethal shock in vivo.  •  Repeated injection o f TSST-1 induced C D 4 + Tregs with distinctive properties as compared to those induced by S E A . T S S T - l - i n d u c e d Tregs had broader suppressive activities in vitro and in vivo. They were hyperproliferative in vitro in the absence or presence o f TSST-1 restimulation in vitro and had more potent proliferative ability. They also produced T N F - a and IFN-y in response to TSST-1 stimulation and had enhanced expression o f C D 2 5 and C T L A - 4 . Moreover,  TSST-l-induced  Tregs  demonstrated  antigen-independent  proliferation and cytokine expression in vitro. •  T S S T - l - i n d u c e d Tregs depend on cell-contact to mediate their suppressive activities and cytokine competition for IL-2 appears to be a major mechanism for their suppressive function on target cells. Induction o f cell death, infectious tolerance and secretion o f soluble factors are not directly involved i n the suppression mediated by TSST-1 ^induced Tregs.  •  Transfer o f T S S T - l - i n d u c e d Tregs to allogeneic recipients followed by TSST-1 injection to activate them attenuated a G V H D . This suppression was not due to inhibition o f donor T cell engraftment and expansion, or enhanced elimination of host A P C s ,  but associated with decreased  proinflammatory cytokine  174  production. Furthermore, control o f a G V H D b y T S S T - l - i n d u c e d Tregs did not jeopardise G V T effects.  7.3. Future directions This study revealed distinctive properties o f T S S T - l - i n d u c e d Tregs. Further studies should be devoted to aspects concerning the development o f TSST-l-induced Tregs. First, the origin o f T S S T - l - i n d u c e d Tregs remains unclear. Data in this study showed that T S S T - l - i n d u c e d Tregs did not demonstrate enhanced expression o f Foxp3, a cellular marker for natural Tregs. However, C D 2 5 and C T L A - 4 expression were enhanced in T S S T - l - i n d u c e d Tregs. Theoretically, during in vivo chronic exposure to T S S T - 1 , any TSST-1-reactive C D 4 + T cells could develop into T S S T - l - i n d u c e d Tregs. Since C D 4 + T cells could be divided into natural Tregs and conventional T cells, both of which contain TSST-l-reactive C D 4 + T cells, it is possible that T S S T - l - i n d u c e d Tregs consist o f TSST-l-reactive C D 4 + T cells derived from both compartments. To address  the  origin  of TSST-l-induced  Tregs,  C D 2 5 + natural  Tregs-depleted  C D 4 + C D 2 5 - conventional T cells should be isolated, mixed with C D 2 5 + natural Tregs isolated from congenic mice bearing markers  for their identification and then  adoptively transferred to S C I D mice. These recipient mice then can be given repeated injection o f T S S T - 1 . The transferred T cells w i l l be re-isolated and separated according to the congenic marker and their suppressive functions reassayed. These experiments w i l l lead to better understanding o f the origin o f T S S T - l - i n d u c e d C D 4 + Tregs. Second, the role o f natural Tregs in the generation o f T S S T - l - i n d u c e d Tregs is another interesting issue for further investigation. Studies have shown that the presence  175  of natural Tregs are required for the generation o f antigen-induced Tregs (Taylor et al., 2001) , possibly due to the ability o f natural Tregs to limit the proinflammatory responses and thus to ensure a non-inflammatory environment that is more conducive for the generation o f Tregs. The role o f natural Tregs i n the generation o f T S S T - l induced Tregs may be studied by transferring natural Tregs-depleted conventional T cells to SCLD mice followed by repeated administration o f TSST-1 to these recipient mice. The transferred T cells can then be assayed for their regulatory functions. Another approach is to deplete the natural Tregs in vivo by the administration o f a depleting antibody against C D 2 5 combined with removal o f the thymus. Then these mice may be repeatedly injected with TSST-1 and the functions o f the C D 4 + T cells can be assayed to determine i f conventional T cells can develop into Tregs i n the absence o f natural Tregs. Third, the role o f certain cytokines should also be further clarified. In a previous study, it was demonstrated that repeated mucosal administration o f S E A resulted in tolerance to this superantigen which was IL-10 dependent (Collins et al., 2002). IL-10 was also required for the induction o f Tregs v i a mucosal tolerance (Massey et al., 2002) . Thus it is important to determine the role o f IL-10 i n the generation o f T S S T - l induced Tregs as these cells appear different from SEA-induced Tregs i n many aspects. LFN-y has been shown to play a role in the induction o f Foxp3+ Tregs (Hong et al., 2005) and data i n this study showed that T S S T - l - i n d u c e d Tregs secreted LFN-y, suggesting that it may play a role in the generation or suppressive function o f these Tregs.  176  Studies on the mechanisms o f T S S T - l - i n d u c e d Tregs revealed two main findings. T S S T - l - i n d u c e d Tregs depend upon cell contact to mediate their suppression and cytokine competition may be one important mechanism. The role o f IL-10 in the suppression mediated by T S S T - l - i n d u c e d Tregs should be further studied. Results i n this study did not show a role o f IL-10 for suppressive activities o f TSST-l-induced Tregs both in vitro and in vivo. It should be kept in mind that using antibody to block actions o f IL-10 may have intrinsic drawbacks. For example, the in vivo administered antibody may not reach sufficient concentrations locally to totally block the actions o f IL-10. Thus results o f this study certainly have not ruled out the possibility that T S S T l-induced Tregs may depend on IL-10 to exert their functions. It is also worthwhile to use more stringent methods to investigate the role o f IL-10. The IL-10 knockout mouse is an excellent candidate for such studies. Thus identifying the molecules responsible for their action should be a major priority in future studies. One o f the potential targets is C T L A - 4 as this molecule has been shown to mediate suppression by Tregs and its expression has been found to be up-regulated in T S S T - l - i n d u c e d Tregs, suggesting a potential causative role for this molecule. However, searching for the relevant molecules amongst the hundreds o f proteins expressed by suppressive T cells w i l l not be an easy task. Powerful molecular biology methods have been developed to screen for  gene expression or down-regulation, and using a genomics and proteomics  approach may greatly facilitate this process. Multiple molecules involved i n Tregs function were identified using this technique. For example, C T L A - 4 , G I T R , Foxp3, L A G - 3 were among the most typical ones (Hori et a l , 2003; Huang et al., 2004; Shimizu et a l , 2002; Takahashi et al., 2000). It must be kept in mind that the change o f  177  gene expression o f certain proteins must be confirmed b y functional assays to ascertain their causative role in the suppression mediated by Tregs. The data showing that T S S T - l - i n d u c e d Tregs suppressed a G V H D while sparing G V T effects clearly suggests the potential therapeutic application o f these cells i n the clinical setting. Further studies should be directed to some aspects that are o f clinical importance; for example, the feasibility to obtain sufficient numbers o f cells for study of their efficacy and effectiveness. In this setting, ex vivo repeated stimulation o f human P B M C to generate human T S S T - l - i n d u c e d Tregs is desirable and this has already been shown to be feasible i n our laboratory. The next step is to further develop an ex vivo method for the generation o f T S S T - l - i n d u c e d Tregs using murine cells and then determine i f these ex vivo generated T S S T - l - i n d u c e d Tregs could be expanded to obtain sufficient numbers for transplantation and i f the expanded Tregs could suppress a G V H D i n the murine model. Another important issue i n A H S C T is immune reconstitution (Porter and June, 2005). Due to the conditioning before A H S C T , the recipient's immune system is severely suppressed and may result i n an immune deficiency that can last 1 to 2 years after A H S C T . Life-threatening infections can develop prior to completion o f the immune reconstitution. a G V H D can further cause delayed immune reconstitution leading to more severe immune deficiency. Since T S S T - l - i n d u c e d Tregs are able to control a G V H D , it is important to investigate i f the co-transfer o f these Tregs is indeed able to enhance immune reconstitution after A H S C T . T o this end, a murine bone marrow transplantation model may be used, where donor hematopoietic stem cells and T cells are transplanted to recipients.  178  Bibliography  Acharya K R , Passalacqua E F , Jones E Y , Harlos K , Stuart D l , Brehm R D , Tranter H S . (1994). Structural basis o f superantigen action inferred from crystal structure o f toxicshock syndrome toxin-1. Nature 367:94-97. Achiron A , Lavie G , Kishner I, Stern Y , Sarova-Pinhas I, B e n Aharon T, Barak Y , Raz H , Lavie M , Barliya T. (2004). T cell vaccination in multiple sclerosis relapsingremitting nonresponders patients. C l i n Immunol 113:155-160. Adachi M , Suematsu S, Suda T, Watanabe D , Fukuyama H , Ogasawara J, Tanaka T, Yoshida N , Nagata S. (1996). Enhanced and accelerated lymphoproliferation in Fasnull mice. P N A S 93:2131-2136. Akasaki Y , L i u G , Chung N H C , Ehtesham M , Black K L , Y u JS. (2004). Induction o f a C D 4 + T Regulatory Type 1 Response by Cyclooxygenase-2-Overexpressing Glioma. J Immunol 173:4352-4359. Akbar A N , Taams L S , Salmon M , Vukmanovic-Stejic M . (2003). The peripheral generation o f C D 4 + CD25+ regulatory T cells. Immunology 109:319-325. Akbari O, D e K r u y f f R H , Umetsu D T . (2001). Pulmonary dendritic cells producing I L 10 mediate tolerance induced by respiratory exposure to antigen. Nat Immunol 2:725731. Akbari O, Freeman G J , Meyer E H , Greenfield E A , Chang T T , Sharpe A H , Berry G , D e K r u y f f R H , Umetsu D T . (2002). Antigen-specific regulatory T cells develop via the ICOS-ICOS-ligand pathway and inhibit allergen-induced airway hyperreactivity. Nat M e d 8:1024-1032. A l l a n S E , Passerini L , Bacchetta R , Crellin N , D a i M , Orban P C , Ziegler S F , Roncarolo M G , Levings M K . (2005). The role o f 2 F O X P 3 isoforms i n the generation o f human C D 4 + Tregs. J C l i n Invest 115:3276-3284. A l p a n O, Bachelder E , Isil E , Arnheiter H , Matzinger P . (2004). 'Educated' dendritic cells act as messengers from memory to naive T helper cells. Nat Immunol 5:615-622. Annacker O, Pimenta-Araujo R, Burlen-Defranoux O, Barbosa T C , Cumano A , Bandeira A . (2001). CD25+ C D 4 + T Cells Regulate the Expansion o f Peripheral C D 4 T Cells Through the Production o f IL-10. J Immunol 166:3008-3018. Antony P A , Piccirillo C A , Akpinarli A , Finkelstein S E , Speiss P J , Surman D R , Palmer D C , Chan C C , Klebanoff C A , Overwijk W W , Rosenberg S A , Restifo N P . (2005). CD8+ T C e l l Immunity Against a Tumor/Self-Antigen Is Augmented by C D 4 + T Helper Cells and Hindered by Naturally Occurring T Regulatory Cells. J Immunol 174:2591-2601.  179  Apostolou I, von Boehmer H . (2004). In V i v o Instruction o f Suppressor Commitment in Naive T Cells. J E x p M e d 199:1401-1408. Arad G , H i l l m a n D , L e v y R, Kaempfer R. (2001). Superantigen antagonist blocks T h l cytokine gene induction and lethal shock. J Leukoc B i o l 69:921-927. Arpinati M , Green C L , Heimfeld S, Heuser J E , Anasetti C . (2000). Granulocyte-colony stimulating factor mobilizes T helper 2-inducing dendritic cells. B l o o d 95:2484-2490. Asano M , Toda M , Sakaguchi N , Sakaguchi S. (1996). Autoimmune disease as a consequence o f developmental abnormality o f a T cell subpopulation. J E x p M e d 184:387-396. Asseman C , Mauze S, Leach M W , Coffman R L , Powrie F . (1999). A n Essential Role for Interleukin 10 in the Function o f Regulatory T Cells That Inhibit Intestinal Inflammation. J E x p M e d 190:995-1004. A z u m a T, Takahashi T, Kunisato A , Kitamura T, Hirai H . (2003). Human C D 4 + CD25+ Regulatory T Cells Suppress N K T C e l l Functions. Cancer Res 63:4516-4520. Bacchetta R, Passerini L , Gambineri E , D a i M , A l l a n S E , Perroni L , Dagna-Bricarelli F, Sartirana C , Matthes-Martin S, Lawitschka A , Azzari C , Ziegler S F , Levings M K , Roncarolo M G . (2006). Defective regulatory and effector T cell functions in patients with F O X P 3 mutations. J C l i n Invest 116:1713-1722. Baecher-Allan C , B r o w n J A , Freeman G J , Hafler D A . (2001). C D 4 + C D 2 5 h i g h Regulatory Cells in Human Peripheral Blood. J Immunol 167:1245-1253. Baecher-Allan C , W o l f E , Hafler D A . (2006). M H C Class II Expression Identifies Functionally Distinct Human Regulatory T Cells. J Immunol 176:4622-4631. Balandina A , Lecart S, Dartevelle P, Saoudi A , Berrih-Aknin S. (2005). Functional defect o f regulatory CD4+CD25+ T cells in the thymus o f patients with autoimmune myasthenia gravis. B l o o d 105:735-741. Banchereau J, Briere F, Caux C , Davoust J, Lebecque S, L i u Y J , Pulendran B , Palucka K . (2000). Immunobiology o f Dendritic Cells. A n n u Rev Immunol 18:767-811. Barrat F J , Cua D J , Boonstra A , Richards D F , Crain C , Savelkoul H F , Waal-Malefyt R, Coffman R L , Hawrylowicz C M , O'Garra A . (2002). In Vitro Generation o f Interleukin 10-producing Regulatory C D 4 + T Cells Is Induced by Immunosuppressive Drugs and Inhibited by T Helper Type 1 ( T h l ) - and Th2-inducing Cytokines. J E x p M e d 195:603616. Barthlott T, Moncrieffe H , Veldhoen M , Atkins C J , Christensen J, O'Garra A , Stockinger B . (2005). CD25+CD4+ T cells compete with naive C D 4 + T cells for IL-2 and exploit it for the induction o f IL-10 production. Int Immunol 17:279-288.  180  Battaglia M , Stabilini A , Roncarolo M G . (2005). Rapamycin selectively expands CD4+CD25+FoxP3+ regulatory T cells. B l o o d 105:4743-4748. Belkaid Y , Piccirillo C A , Mendez S, Shevach E M , Sacks D L . (2002). C D 4 + C D 2 5 + regulatory T cells control Leishmania major persistence and immunity. Nature 420:502-507. Belkaid Y , Sun C M , Bouladoux N . (2006). Parasites and immunoregulatory T cells. Curr Opin Immunol 18:406-412. Beyersdorf N , Balbach K , Hunig T, Kerkau T. (2006). Large-scale expansion o f rat C D 4 + CD25+ Treg cells i n the absence o f T-cell receptor stimulation. Immunology . Blank C , L u z A , Bendigs S, Erdmann A , Wagner H , Heeg K . (1997). Superantigen and endotoxin synergize i n the induction o f lethal shock. E u r J Immunol 27:825-833. Blazar B R , Taylor P A . (2005). Regulatory T cells. B i o l B l o o d M a r r o w Transplant 11:46-49. Blazar B R , Taylor P A , Noelle R J , Vallera D A . (1998). C D 4 + T Cells Tolerized E x V i v o to Host Alloantigen by A n t i - C D 4 0 Ligand ( C D 4 0 L : C D 154) Antibody Lose Their Graft-Versus-Host Disease Lethality Capacity but Retain N o m i n a l Antigen Responses. J C l i n Invest 102:473-482. Bluestone J A , Abbas A K . (2003). Natural versus adaptive regulatory T cells. Nat Rev Immunol 3:253-257. Bolacchi F, Sinistro A , Ciaprini C , Demin F , Capozzi M , Carducci F C , Drapeau C M J , Rocchi G , Bergamini A . (2006). Increased hepatitis C virus (HCV)-specific C D 4 + C D 2 5 + regulatory T lymphocytes and reduced H C V - s p e c i f i c C D 4 + T cell response in HCV-infected patients with normal versus abnormal alanine aminotransferase levels. C l i n E x p Immunol 144:188-196. Brocke S, Gaur A , Piercy C , Gautam A , Gijbels K , Fathman C G , Steinman L . (1993). Induction o f relapsing paralysis in experimental autoimmune encephalomyelitis by bacterial superantigen. Nature 365:642-644. Bruder D , Probst-Kepper M , Westendorf A M , Geffers R , Beissert S, Loser K , von Boehmer H , Buer J, Hansen W . (2004). Neuropilin-1: a surface marker o f regulatory T cells. Eur J Immunol 34:623-630. Callahan J E , Herman A , Kappler J W , Marrack P. (1990). Stimulation o f B 1 0 . B R T cells with superantigenic staphylococcal toxins. J Immunol 144:2473-2479. Cameron S B . (2004). Generation and partial characterization o f a regulatory T cell phenotype following repeated in vivo treatment with the staphylococcal superantigen toxic shock syndrome toxin-1.  181  Cantor H . (2004). Reviving suppression? Nat Immunol 5:347-349. Cederbom L , H a l l H , Ivars F. (2000). C D 4 + C D 2 5 + regulatory T cells down-regulate co-stimulatory molecules on antigen-presenting cells. Eur J Immunol 30:1538-1543. Chang C C , Ciubotariu R, Manavalan JS, Y u a n J, Colovai A I , Piazza F , Lederman S, Colonna M , Cortesini R, Dalla-Favera R, Suciu-Foca N . (2002). Tolerization o f dendritic cells by TS cells: the crucial role o f inhibitory receptors I L T 3 and ILT4. Nat Immunol 3:237-243. Charbonnier L M , van Duivenvoorde L M , Apparailly F , Cantos C , H a n W G H , N o e l D , Duperray C , Huizinga T W J , Toes R E M , Jorgensen C , Louis-Plence P. (2006). Immature Dendritic Cells Suppress Collagen-Induced Arthritis by In V i v o Expansion o f CD49b+ Regulatory T Cells. J Immunol 177:3806-3813. Chen T C , Waldmann H , Fairchild P J . (2004). Induction o f dominant transplantation tolerance by an altered peptide ligand o f the male antigen D b y . J C l i n Invest 113:17541762. Chen W , Jin W , Hardegen N , L e i K j , L i L , Marinos N , M c G r a d y G , W a h l S M . (2003a). Conversion o f Peripheral C D 4 + C D 2 5 - Naive T Cells to C D 4 + C D 2 5 + Regulatory T Cells by TGF-{beta} Induction o f Transcription Factor Foxp3. J E x p M e d 198:18751886. Chen Y , Kuchroo V K , Inobe J, Hafler D A , Weiner H L . (1994). Regulatory T cell clones induced by oral tolerance: suppression o f autoimmune encephalomyelitis. Science 265:1237-1240. Chen Y , Inobe J i , Kuchroo V K , Baron J L , Janeway C A , Jr., Weiner H L . (1996). Oral tolerance in myelin basic protein T-cell receptor transgenic mice: Suppression o f autoimmune encephalomyelitis and dose-dependent induction o f regulatory cells. P N A S 93:388-391. Chen Z m , O'Shaughnessy M J , Gramaglia I, Panoskaltsis-Mortari A , M u r p h y W J , Narula S, Roncarolo M G , Blazar B R . (2003b). IL-10 and TGF-{beta} induce alloreactive C D 4 + C D 2 5 - T cells to acquire regulatory cell function. B l o o d 101:50765083. Cher D J , Mosmann T R . (1987). T w o types o f murine helper T cell clone. II. Delayedtype hypersensitivity is mediated by T H 1 clones. J Immunol 138:3688-3694. Childs R , Chernoff A , Contentin N , Bahceci E , Schrump D , Leitman S, Read E J , Tisdale J, Dunbar C , Linehan W M , Y o u n g N S , Barrett A J . (2000). Regression o f Metastatic Renal-Cell Carcinoma after Nonmyeloablative Allogeneic Peripheral-Blood Stem-Cell Transplantation. N Engl J M e d 343:750-758. Cobbold S, Waldmann H . (1998). Infectious tolerance. Current Opinion i n Immunology 10:518-524. 182  Coenen J J A , Koenen H J P M , van Rijssen E , Hilbrands L B , Joosten I. (2006). Rapamycin, and not cyclosporin A , preserves the highly suppressive C D 2 7 + subset o f human CD4+CD25+ regulatory T cells. B l o o d 107:1018-1023. Cohen JL, Trenado A , Vasey D , Klatzmann D , Salomon B L . (2002). C D 4 + C D 2 5 + Immunoregulatory T Cells: N e w Therapeutics for Graft-Versus-Host Disease. J Exp M e d 196:401-406. Collins L V , Eriksson K , U l r i c h R G , Tarkowski A . (2002). Mucosal Tolerance to a Bacterial Superantigen Indicates a N o v e l Pathway To Prevent Toxic Shock. Infect Immun 70:2282-2287. Colovai A l , Ciubotariu R , L i u Z , Cortesini R, Suciu-Foca N . (2001). C D 8 + C D 2 8 - T suppressor cells represent a distinct subset in a heterogeneous population. Transplantation Proceedings 33:104-107. Conrad B , Weidmann E , Trucco G , Rudert W A , Behboo R, Ricordi C , Rodriquez-Rilo H , Finegold D , Trucco M . (1994). Evidence for superantigen involvement in insulindependent diabetes mellitus aetiology. Nature 371:351-355. Conrad B , Weissmahr R N , B o n i J, Arcari R, Schupbach J, M a c h B . (1997). A Human Endogenous Retroviral Superantigen as Candidate Autoimmune Gene i n Type I Diabetes. C e l l 90:303-313. Cooke K R , Gerbitz A , Crawford J M , Teshima T, H i l l G R , Tesolin A , Rossignol D P , Ferrara J L M . (2001). L P S antagonism reduces graft-versus-host disease and preserves graft-versus-leukemia activity after experimental bone marrow transplantation. J C l i n Invest 107:1581-1589. Correale J, Lund B , M c M i l l a n M , K o D Y , McCarthy K , Weiner L P . (2000). T cell vaccination in secondary progressive multiple sclerosis. J Neuroimmunol 107:130-139. Cottrez F , Hurst S D , Coffman R L , Groux H . (2000). T Regulatory Cells 1 Inhibit a Th2-Specific Response In V i v o . J Immunol 165:4848-4853. Cozzo C , Larkin J, III, Caton A J . (2003). Cutting Edge: Self-Peptides Drive the Peripheral Expansion o f CD4+CD25+ Regulatory T Cells. J Immunol 171:5678-5682. Damle N K , Engleman E G . (1983). Immunoregulatory T cell circuits i n man. Alloantigen-primed inducer T cells activate alloantigen-specific suppressor T cells in the absence o f the initial antigenic stimulus. J E x p M e d 158:159-173. Dauwalder O, Thomas D , Ferry T, Debard A L , Badiou C , Vandenesch F , Etienne J, L i n a G , Monneret G . (2006). Comparative inflammatory properties o f staphylococcal superantigenic enterotoxins S E A and S E G : implications for septic shock. J Leukoc B i o l 80:753-758.  183  de la Rosa M , Dorninger H , Scheffold A . (2004). Interleukin-2 is essential for CD4+CD25+ regulatory T cell function. Eur J Immunol 34:2480-2488. Deng S, Moore D J , Huang X , Mohiuddin M , Lee M K , I V , Velidedeoglu E , Lian M M , Chiaccio M , Sonawane S, Orlin A , Wang J, Chen H , Caton A , Zhong R , Markmann JF. (2006). Antibody-Induced Transplantation Tolerance That Is Dependent on ThymusDerived Regulatory T Cells. J Immunol 176:2799-2807. Devetten M P , Vose J M . (2004). Graft-versus-host disease: H o w to translate new insights into new therapeutic strategies. B i o l B l o o d Marrow Transplant 10:815-825. Dhodapkar M V , Steinman R M . (2002). Antigen-bearing immature dendritic cells induce peptide-specific C D 8 + regulatory T cells i n vivo i n humans. B l o o d 100:174-177. Diaz-Gallo C , Moscovitch-Lopatin M , Strom T B , K e l l e y V R . (1992). A n Anergic, Islet-Infiltrating T - C e l l Clone that Suppresses Murine Diabetes Secretes a Factor That Blocks Interleukin 2/Interleukin 4-Dependent Proliferation. P N A S 89:8656-8660. Dieckmann D , Bruett C H , Ploettner H , Lutz M B , Schuler G . (2002). Human CD4+CD25+ Regulatory, Contact-dependent T Cells Induce Interleukin 10-producing, Contact-independent Type 1-like Regulatory T Cells. J E x p M e d 196:247-253. Dieckmann D , Plottner H , Dotterweich S, Schuler G . (2005). Activated C D 4 + CD25+ T cells suppress antigen-specific C D 4 + and C D 8 + T cells but induce a suppressive phenotype only in C D 4 + T cells. Immunology 115:305-314. D i n g Q, L u L , Wang B , Zhou Y , Jiang Y , Zhou X , X i n L , Jiao Z , Chou K Y . (2006). B7H1-Ig Fusion Protein Activates the C D 4 + IFN-{gamma} Receptor+ Type 1 T Regulatory Subset through LFN-{gamma}-Secreting T h l Cells. J Immunol 177:36063614. Dinges M M , Orwin P M , Schlievert P M . (2000). Exotoxins o f Staphylococcus aureus. C l i n M i c r o b i o l Rev 13:16-34. Dinges M M , Schlievert P M . (2001). Role o f T Cells and Gamma Interferon during Induction o f Hypersensitivity to Lipopolysaccharide by T o x i c Shock Syndrome Toxin 1 i n M i c e . Infect Immun 69:1256-1264. Dudley M E , Wunderlich JR, Robbins P F , Y a n g J C , H w u P, Schwartzentruber D J , Topalian S L , Sherry R, Restifo N P , Hubicki A M , Robinson M R , Raffeld M , Duray P, Seipp C A , Rogers-Freezer L , Morton K E , Mavroukakis S A , White D E , Rosenberg S A . (2002). Cancer Regression and Autoimmunity i n Patients After Clonal Repopulation with Antitumor Lymphocytes. Science 298:850-854. Duthoit C T , Nguyen P, Geiger T L . (2004). Antigen Nonspecific Suppression o f T C e l l Responses by Activated Stimulation-Refractory C D 4 + T Cells. J Immunol 172:22382246.  184  Edinger M , Hoffmann P, Ermann J, Drago K , Fathman C G , Strober S, Negrin R S . (2003) . CD4+CD25+ regulatory T cells preserve graft-versus-tumor activity while inhibiting graft-versus-host disease after bone marrow transplantation. Nat M e d 9:1144-1150. Ehrenstein M R , Evans J G , Singh A , Moore S, Warnes G , Isenberg D A , M a u r i C . (2004) . Compromised Function o f Regulatory T Cells i n Rheumatoid Arthritis and Reversal by A n t i - T N F {alpha} Therapy. J E x p M e d 200:277-285. Ellison C A , Fischer J M M , HayGlass K T , Gartner J G . (1998). Murine Graft-VersusHost Disease in an F l - H y b r i d M o d e l Using IFN-{gamma} Gene Knockout Donors. J Immunol 161:631-640. Ermann J, Hoffmann P, Edinger M , Dutt S, Blankenberg F G , Higgins JP, Negrin R S , Fathman C G , Strober S. (2005). Only the C D 6 2 L + subpopulation o f CD4+CD25+ regulatory T cells protects from lethal acute G V H D . B l o o d 105:2220-2226. E w i n g P, M i k l o s S, Olkiewicsz K M , Muller G , Andreesen R , Holler E , Cook K R , Hildebrandt G C . (2007). Donor C D 4 + T-cell production o f tumor necrosis factor alpha significantly contributes to the early proinflammatory events o f graft-versus-host disease. E x p Hematol 35:155-163. Fahlen L , Read S, Gorelik L , Hurst S D , Coffman R L , Flavell R A , Powrie F. (2005). T cells that cannot respond to TGF-{beta} escape control by C D 4 + C D 2 5 + regulatory T cells. J E x p M e d 201:737-746. Faria A M C , Weiner H L . (2005). Oral tolerance. Immunol R e v 206:232-259. Faulkner L , Cooper A , Fantino C , Altmann D M , Sriskandan S. (2005). The Mechanism of Superantigen-Mediated Toxic Shock: Not a Simple T h l Cytokine Storm. J Immunol 175:6870-6877. Fehr T, Sykes M . (2004). Tolerance induction i n clinical transplantation. Transpl Immunol 13:117-130. Ferguson T A , Griffith T S . (2006). A vision o f cell death: Fas ligand and immune privilege 10 years later. Immunol Rev 213:228-238. Feunou P, Poulin L , Habran C , L e Moine A , Goldman M , Braun M Y . (2003). CD4+CD25+ and C D 4 + C D 2 5 - T Cells A c t Respectively as Inducer and Effector T Suppressor Cells in Superantigen-induced Tolerance. J Immunol 171:3475-3484. Fisson S, Darrasse-Jeze G , Litvinova E , Septier F , Klatzmann D , Liblau R, Salomon B L . (2003). Continuous Activation o f Autoreactive C D 4 + C D 2 5 + Regulatory T Cells in the Steady State. J E x p M e d 198:737-746. Foley J E , Jung U , M i e r a A , Borenstein T, Mariotti J, Eckhaus M , Bierer B E , Fowler D H . (2005). E x V i v o Rapamycin Generates Donor Th2 Cells That Potently Inhibit 185  Graft-versus-Host Disease and Graft-versus-Tumor Effects v i a an IL-4-Dependent Mechanism. J Immunol 175:5732-5743. Fontenot J D , Gavin M A , Rudensky A Y . (2003). Foxp3 programs the development and function o f CD4+CD25+ regulatory T cells. Nat Immunol 4:330-336. Ford M S , Y o u n g K J , Zhang Z , Ohashi P S , Zhang L . (2002). The Immune Regulatory Function o f Lymphoproliferative Double Negative T Cells In Vitro and In V i v o . J Exp M e d 196:261-267. Fowler D H . (2006). Shared biology o f G V H D and G V T effects: Potential methods o f separation. Crit Rev Oncol Hematol 57:225-244. Franzese O, Kennedy P T F , Gehring A J , Gotto J, Williams R , M a i n i M K , Bertoletti A . (2005). Modulation o f the CD8+-T-Cell Response by C D 4 + C D 2 5 + Regulatory T Cells i n Patients with Hepatitis B Virus Infection. J V i r o l 79:3322-3328. Fraser J, Arcus V , K o n g P, Baker E , Proft T. (2000). Superantigens - powerful modifiers o f the immune system. M o i M e d Today 6:125-132. Fritzsching B , Oberle N , Eberhardt N , Quick S, Haas J, Wildemann B , Krammer P H , Suri-Payer E . (2005). Cutting Edge: In Contrast to Effector T Cells, CD4+CD25+FoxP3+ Regulatory T Cells A r e Highly Susceptible to C D 9 5 Ligand- but Not to TCR-Mediated C e l l Death. J Immunol 175:32-36. Fukaura H , Kent S C , Pietrusewicz M J , Khoury SJ, Weiner H L , Hafler D A . (1996). Induction o f Circulating M y e l i n Basic Protein and Proteolipid Protein-specific Transforming Growth Factor-beta 1-secreting Th3 T Cells b y Oral Administration o f M y e l i n in Multiple Sclerosis Patients. J C l i n Invest 98:70-77. Furtado G C , de Lafaille M A C , Kutchukhidze N , Lafaille JJ. (2002). Interleukin 2 Signaling Is Required for C D 4 + Regulatory T C e l l Function. J E x p M e d 196:851-857. Gajewski T F , M e n g Y , Blank C , B r o w n I, Kacha A , K l i n e J, Harlin H . (2006). Immune resistance orchestrated by the tumor microenvironment. Immunol Rev 213:131-145. Gattinoni L , Powell D J , Rosenberg S A , Restifo N P . (2006). Adoptive immunotherapy for cancer: building on success. Nat Rev Immunol 6:383-393. Gavin M A , Clarke S R , Negrou E , Gallegos A , Rudensky A . (2002). Homeostasis and anergy o f CD4+CD25+ suppressor T cells in vivo. Nat Immunol 3:33-41. George T C , Bilsborough J, Viney J L , Norment A M . (2003). H i g h antigen dose and activated dendritic cells enable T h cells to escape regulatory T cell-mediated suppression in vitro. Eur J Immunol 33:502-511. Ghiringhelli F , Menard C , Terme M , Flament C , Taieb J, Chaput N , Puig P E , Novault S, Escudier B , V i v i e r E , Lecesne A , Robert C , B l a y J Y , Bernard J, Caillat-Zucman S, 186  Freitas A , Tursz T, Wagner-Ballon O, Capron C , Vainchencker W , M a r t i n F , Zitvogel L . (2005). CD4+CD25+ regulatory T cells inhibit natural killer cell functions in a transforming growth factor-{beta}-dependent manner. J E x p M e d 202:1075-1085. G i d l o f C , Dohlsten M , Lando P, Kalland T, Sundstrom C , Totterman T H . (1997). A Superantigen-Antibody Fusion Protein for T - C e l l Immunotherapy o f Human B Lineage Malignancies. B l o o d 89:2089-2097. Gilliet M , L i u Y J . (2002). Generation o f Human C D 8 T Regulatory Cells by C D 4 0 Ligand-activated Plasmacytoid Dendritic Cells. J E x p M e d 195:695-704. Goker H , Haznedaroglu IC, Chao N J . (2001). Acute graft-vs-host disease: Pathobiology and management. E x p Hematol 29:259-277. Goodyear C S , Silverman G J . (2003). Death by a B C e l l Superantigen: In V i v o V H targeted Apoptotic Supraclonal B C e l l Deletion by a Staphylococcal Toxin. J E x p M e d 197:1125-1139. Graca L , Chen T C , M o i n e A L , Cobbold SP, Howie D , Waldmann H . (2005). Dominant tolerance: activation thresholds for peripheral generation o f regulatory T cells. Trends Immunol 26:130-135. Graca L , Honey K , Adams E , Cobbold SP, Waldmann H . (2000). Cutting Edge: A n t i C D 154 Therapeutic Antibodies Induce Infectious Transplantation Tolerance. J Immunol 165:4783-4786. Graca L , Thompson S, L i n C Y , Adams E , Cobbold SP, Waldmann H . (2002). Both CD4+CD25+ and C D 4 + C D 2 5 - Regulatory Cells Mediate Dominant Transplantation Tolerance. J Immunol 168:5558-5565. Green E A , Gorelik L , McGregor C M , Tran E H , Flavell R A . (2003). CD4+CD25+ T regulatory cells control anti-islet C D 8 + T cells through TGF-{beta}-TGF-{beta} receptor interactions in type 1 diabetes. P N A S 100:10878-10883. Grossman W J , Verbsky J W , Barchet W , Colonna M , Atkinson JP, L e y T J . (2004). Human T Regulatory Cells Can Use the Perforin Pathway to Cause Autologous Target C e l l Death. Immunity 21:589-601. Groux H , O'Garra A , Bigler M , Rouleau M , Antonenko S, de Vries J E , Roncarolo M G . (1997). A CD4+T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 389:737-742. Grundstrom S, Cederbom L , Sundstedt A , Scheipers P , Ivars F . (2003). Superantigeninduced Regulatory T Cells Display Different Suppressive Functions i n the Presence or Absence o f Natural CD4+CD25+ Regulatory T Cells In V i v o . J Immunol 170:50085017.  187  Gunther J, Haas W , von Boehmer H . (1982). Suppression o f T cell responses through competition for T cell growth factor (interleukin 2). Eur J Immunol 12:247-249. Hashimoto D , Asakura S, M i y a k e S, Yamamura T, V a n Kaer L , L i u C , Tanimoto M , Teshima T. (2005). Stimulation o f Host N K T Cells by Synthetic G l y c o l i p i d Regulates Acute Graft-versus-Host Disease by Inducing Th2 Polarization o f Donor T Cells. J Immunol 174:551-556. H i l l G R , Cooke K R , Teshima T, Crawford J M , Keith J C , Jr., Brinson Y S , Bungard D , Ferrara J L M . (1998). Interleukin-11 Promotes T C e l l Polarization and Prevents Acute Graft-Versus-Host Disease after Allogeneic Bone Marrow Transplantation. J C l i n Invest 102:115-123. H i l l G R , Crawford J M , Cooke K R , Brinson Y S , Pan L , Ferrara J L M . (1997). Total B o d y Irradiation and Acute Graft-Versus-Host Disease: The Role o f Gastrointestinal Damage and Inflammatory Cytokines. B l o o d 90:3204-3213. H i l l G R , Ferrara J L M . (2000). The primacy o f the gastrointestinal tract as a target organ o f acute graft-versus-host disease: rationale for the use o f cytokine shields in allogeneic bone marrow transplantation. B l o o d 95:2754-2759. H i l l G R , Teshima T, Gerbitz A , Pan L , Cooke K R , Brinson Y S , Crawford J M , Ferrara J L M . (1999). Differential roles o f IL-1 and TNF-{alpha} on graft-versus-host disease and graft versus leukemia. J C l i n Invest 104:459-467. Hoffmann P, Eder R , Kunz-Schughart L A , Andreesen R , Edinger M . (2004). Largescale in vitro expansion o f polyclonal human CD4+CD25high regulatory T cells. B l o o d 104:895-903. Hoffmann P, Ermann J, Edinger M , Fathman C G , Strober S. (2002). Donor-type CD4+CD25+ Regulatory T Cells Suppress Lethal Acute Graft-Versus-Host Disease after Allogeneic Bone Marrow Transplantation. J E x p M e d 196:389-399. Honey K , Cobbold SP, Waldmann H . (1999). C D 4 0 Ligand Blockade Induces CD4+ T C e l l Tolerance and Linked Suppression. J Immunol 163:4805-4810. Hong J, L i N , Zhang X , Zheng B , Zhang JZ. (2005). Induction o f C D 4 + C D 2 5 + regulatory T cells by copolymer-I through activation o f transcription factor Foxp3 40. P N A S 102:6449-6454. Hori S, Nomura T, Sakaguchi S. (2003). Control o f Regulatory T C e l l Development by the Transcription Factor Foxp3. Science 299:1057-1061. Horowitz M M , Gale R P , Sondel P M , Goldman J M , Kersey J, K o l b H J , R i m m A A , Ringden O, Rozman C , Speck B . (1990). Graft-versus-leukemia reactions after bone marrow transplantation. B l o o d 75:555-562.  188  Houot R, Perrot I, Garcia E , Durand I, Lebecque S. (2006). Human C D 4 + C D 2 5 h i g h Regulatory T Cells Modulate M y e l o i d but Not Plasmacytoid Dendritic Cells Activation. J Immunol 176:5293-5298. Huang C T , Workman C J , Flies D , Pan X , Marson A L , Zhou G , Hipkiss E L , R a v i S, K o w a l s k i J, Levitsky H I . (2004). Role o f L A G - 3 i n Regulatory T Cells. Immunity 21:503-513. Huang R W Y . (2004). The regulation o f apoptosis by toxic shock syndrome toxin-1 and associated mutants. Huber S, Schramm C , Lehr H A , M a n n A , Schmitt S, Becker C , Protschka M , Galle P R , Neurath M F , Blessing M . (2004). Cutting Edge: TGF-{beta} Signaling Is Required for the In V i v o Expansion and Immunosuppressive Capacity o f Regulatory C D 4 + C D 2 5 + T Cells. J Immunol 173:6526-6531. Hudson K R , Tiedemann R E , Urban R G , Lowe S C , Strominger J L , Fraser J D . (1995). Staphylococcal enterotoxin A has two cooperative binding sites on major histocompatibility complex class II. J E x p M e d 182:711-720. Huehn J, Hamann A . (2005). H o m i n g to suppress: address codes for Treg migration. Trends in Immunology 26:632-636. Hurley J M , Shimonkevitz R, Hanagan A , Enney K , Boen E , Malmstrom S, K o t z i n B L , Matsumura M . (1995). Identification o f class II major histocompatibility complex and T cell receptor binding sites in the superantigen toxic shock syndrome toxin 1. J Exp M e d 181:2229-2235. Itoh M , Takahashi T, Sakaguchi N , Kuniyasu Y , Shimizu J, Otsuka F , Sakaguchi S. (1999). Thymus and Autoimmunity: Production o f C D 2 5 + C D 4 + Naturally Anergic and Suppressive T Cells as a K e y Function o f the Thymus i n Maintaining Immunologic Self-Tolerance. J Immunol 162:5317-5326. Jaksch M , Mattsson J. (2005). The Pathophysiology o f Acute Graft- Versus-Host Disease. Scand J Immunol 61:398-409. Janssens W , Carlier V , W u B , VanderElst L , Jacquemin M G , Saint-Remy J M . (2003). CD4+CD25+ T Cells Lyse Antigen-Presenting B Cells by Fas-Fas Ligand Interaction in an Epitope-Specific Manner. J Immunol 171:4604-4612. Jiang H , Curran S, Ruiz-Vazquez E , Liang B , Winchester R , Chess L . (2003). Regulatory C D 8 + T cells fine-tune the myelin basic protein-reactive T cell receptor V{beta} repertoire during experimental autoimmune encephalomyelitis. P N A S 100:8378-8383. Jiang H , Braunstein N S , Y u B , Winchester R, Chess L . (2001). C D 8 + T cells control the T H phenotype o f MBP-reactive C D 4 + T cells in E A E mice. P N A S 98:6301-6306.  189  Jiang H , Kashleva H , X u L X , Forman J, Flaherty L , Pernis B , Braunstein N S , Chess L . (1998). T cell vaccination induces T cell receptor Vbeta -specific Qa-1-restricted regulatory C D 8 + T cells. P N A S 95:4533-4537. Jiang H , Ware R, Stall A , Flaherty L , Chess L , Pernis B . (1995). Murine C D 8 + T cells that specifically delete autologous C D 4 + T cells expressing V[beta]8 T C R : a role o f the Qa-1 molecule. Immunity 2:185-194. Jiang H , Zhang SI, Pernis B . (1992). Role o f C D 8 + T Cells i n Murine Experimental Allergic Encephalomyelitis. Science 256:1213-1215. Jin H , Gong D , Adeegbe D , Bayer A L , Rolle C , Y u A , M a l e k T R . (2006). Quantitative assessment concerning the contribution o f IL-2R{beta} for superantigen-mediated T cell responses in vivo. Int Immunol 18:565-572. Jones S C , Murphy G F , Korngold R. (2003). Post-hematopoietic cell transplantation control o f graft-versus-host disease by donor CD4+25+ T cells to allow an effective graft-versus-leukemia response. B i o l B l o o d Marrow Transplant 9:243-256. Jonuleit H , Schmitt E , Kakirman H , Stassen M , K n o p J, Enk A H . (2002). Infectious Tolerance: Human CD25+ Regulatory T Cells Convey Suppressor A c t i v i t y to Conventional C D 4 + T Helper Cells. J E x p M e d 196:255-260. Jonuleit H , Schmitt E , Schuler G , K n o p J, Enk A H . (2000). Induction o f Interleukin 10-producing, Nonproliferating CD4+ T Cells with Regulatory Properties by Repetitive Stimulation with Allogeneic Immature Human Dendritic Cells. J E x p M e d 192:1213-1222. Jordan M S , Boesteanu A , Reed A J , Petrone A L , Holenbeck A E , Lerman M A , Naji A , Caton A J . (2001). Thymic selection o f CD4+CD25+ regulatory T cells induced by an agonist self-pep tide. Nat Immunol 2:301-306. June C H , Blazar B R . (2006). Clinical application o f expanded CD4+25+ cells. Semin Immunol 18:78-88. Kageyama Y , Koide Y , Nagata T, Uchijima M , Yoshida A , A r a i T, M i u r a T, Miyamoto C , Nagano A . (2001). Toxic Shock Syndrome Toxin-1 Accelerated Collagen-induced Arthritis in M i c e . J Autoimmun 16:125-131. Karakhanova S, Munder M , Schneider M , Bonyhadi M , H o A D , Goerner M . (2006). Highly Efficient Expansion o f Human CD4+CD25+ Regulatory T Cells for Cellular Immunotherapy in Patients with Graft-Versus-Host Disease. [Article]. J Immunother 29:336-349. K a r i m M , Feng G , W o o d K J , Bushell A R . (2005). C D 2 5 + C D 4 + regulatory T cells generated by exposure to a model protein antigen prevent allograft rejection: antigenspecific reactivation in vivo is critical for bystander regulation. B l o o d 105:4871-4877.  190  K a r i m M , Kingsley C I , Bushell A R , Sawitzki B S , W o o d K J . (2004). AlloantigenInduced CD25+CD4+ Regulatory T Cells Can Develop In V i v o from C D 2 5 - C D 4 + Precursors i n a Thymus-Independent Process. J Immunol 172:923-928. Kawabe Y , Ochi A . (1991). Programmed cell death and extrathymic reduction o f V[beta]8+CD4+ T cells in mice tolerant to Staphylococcus aureus enterotoxin B . Nature 349:245-248. Kawamura T, Nagata M , Utsugi T, Y o o n J W . (1993). Prevention o f autoimmune type I diabetes by C D 4 + suppressor T cells in superantigen-treated non-obese diabetic mice. J Immunol 151:4362-4370. Kemper C , Chan A C , Green J M , Brett K A , Murphy K M , Atkinson JP. (2003). Activation o f human C D 4 + cells with C D 3 and C D 4 6 induces a T-regulatory cell 1 phenotype. Nature 421:388-392. Khattri R , C o x T, Yasayko S A , Ramsdell F. (2003). A n essential role for Scurfin in CD4+CD25+ T regulatory cells. Nat Immunol 4:337-342. K i m C , Siminovitch K A , Ochi A . (1991). Reduction o f lupus nephritis i n M R L / l p r mice by a bacterial superantigen treatment. J E x p M e d 174:1431-1437. K i m H J , Hwang SJ, K i m B K , Jung K C , Chung D H . (2006). N K T cells play critical roles in the induction o f oral tolerance by inducing regulatory T cells producing IL-10 and transforming growth factor beta, and by clonally deleting antigen-specific T cells. Immunology 118:101-111. Kingsley C I , K a r i m M , Bushell A R , W o o d K J . (2002). C D 2 5 + C D 4 + Regulatory T Cells Prevent Graft Rejection: C T L A - 4 - and IL-10-Dependent Immunoregulation o f Alloresponse. J Immunol 168:1080-1086. K o h m A P , Carpentier P A , Anger H A , M i l l e r S D . (2002). Cutting Edge: CD4+CD25+ Regulatory T Cells Suppress Antigen-Specific Autoreactive Immune Responses and Central Nervous System Inflammation During Active Experimental Autoimmune Encephalomyelitis. J Immunol 169:4712-4716. K o l b H J , Mittermuller J, C l e m m C , Holler E , Ledderose G , Brehm G , H e i m M , Wilmanns W . (1990). Donor leukocyte transfusions for treatment o f recurrent chronic myelogenous leukemia in marrow transplant patients. B l o o d 76:2462-2465. K o l b H J , Schattenberg A , Goldman J M , Hertenstein B , Jacobsen N , Arcese W , Ljungman P, Ferrant A , Verdonck L , Niederwieser D . (1995). Graft-versus-leukemia effect o f donor lymphocyte transfusions in marrow grafted patients. European Group for Blood and Marrow Transplantation Working Party Chronic Leukemia [see comments]. B l o o d 86:2041-2050. K o l b H J , Schmid C , Barrett A J , Schendel D J . (2004). Graft-versus-leukemia reactions in allogeneic chimeras. B l o o d 103:767-776. 191  Korngold R, Friedman T M , Jones S C . (2005). A n i m a l models o f graft-vs-host disease. In: G R A F T - V S - H O S T D I S E A S E , 3rd ed, Ferrara J M , Cooke K R , Deeg H J (eds), Marcel Dekker, N e w York, pp 35-58. Kosiewicz M M , Alard P, Liang S, Clark S L . (2004). Mechanisms o f tolerance induced by transforming growth factor-{beta}-treated antigen-presenting cells: C D 8 regulatory T cells inhibit the effector phase o f the immune response in primed mice through a mechanism involving Fas ligand. Int Immunol 16:697-706. Krenger W , Snyder K M , B y o n J C , Falzarano G , Ferrara J L . (1995). Polarized type 2 alloreactive C D 4 + and C D 8 + donor T cells fail to induce experimental acute graftversus-host disease. J Immunol 155:585-593. Kriegel M A , Lohmann T, Gabler C , Blank N , Kalden JR, Lorenz H M . (2004). Defective Suppressor Function o f Human C D 4 + CD25+ Regulatory T Cells i n Autoimmune Polyglandular Syndrome Type II. J E x p M e d 199:1285-1291. Krijanovski OI, H i l l G R , Cooke K R , Teshima T, Crawford J M , Brinson Y S , Ferrara J L M . (1999). Keratinocyte Growth Factor Separates Graft-Versus-Leukemia Effects From Graft-Versus-Host Disease. B l o o d 94:825-831. Kudo-Saito C , Schlom J, Camphausen K , Coleman C N , Hodge J W . (2005). The Requirement o f Multimodal Therapy (Vaccine, L o c a l Tumor Radiation, and Reduction o f Suppressor Cells) to Eliminate Established Tumors. C l i n Cancer Res 11:4533-4544. Kullberg M C , H a y V , Cheever A W , Mamura M , Sher A , Letterio JJ, Shevach E M , Piccirillo C A . (2006). T G F - b e t a l production by C D 4 + C D 2 5 + regulatory T cells is not essential for suppression o f intestinal inflammation. Eur J Immunol 35:2886-2895. K u m W W , Laupland K B , Chow A W . (2000). Defining a novel domain o f staphylococcal toxic shock syndrome toxin-1 critical for major histocompatibility complex class II binding, superantigenic activity, and lethality. C a n J M i c r o b i o l 46:171-179. K u m W W , L i H W , Chow A W . (2006). In vitro generation o f proliferative human C D 4 + Regulatory T cells with suppressor function by the superantigen toxic shock toxin-1. Manuscript in preparation . K u m W W , W o o d J A , Chow A W . (1996). A mutation at glycine residue 31 o f toxic shock syndrome toxin-1 defines a functional site critical for major histocompatibility complex class II binding and superantigenic activity. J Infect D i s 174:1261-1270. K u m W W , Laupland K B , See R H , Chow A W . (1993). Improved purification and biologic activities o f staphylococcal toxic shock syndrome toxin 1. J C l i n M i c r o b i o l 31:2654-2660. K u m W W S , Cameron S B , Hung R W Y , Kalyan S, C h o w A W . (2001). Temporal Sequence and Kinetics o f Proinflammatory and Anti-Inflammatory Cytokine Secretion 192  Induced by Toxic Shock Syndrome T o x i n 1 i n Human Peripheral B l o o d Mononuclear Cells. Infect Immun 69:7544-7549. Kumanogoh A , Wang X , Lee I, Watanabe C , Kamanaka M , Shi W , Yoshida K , Sato T, Habu S, Itoh M , Sakaguchi N , Sakaguchi S, Kikutani H . (2001). Increased T Cell Autoreactivity in the Absence o f C D 4 0 - C D 4 0 Ligand Interactions: A Role o f C D 4 0 i n Regulatory T C e l l Development. J Immunol 166:353-360. Kumar V , Sercarz E E . (1993). The involvement o f T cell receptor peptide-specific regulatory C D 4 + T cells i n recovery from antigen-induced autoimmune disease. J E x p M e d 178:909-916. Kursar M , Bonhagen K , Fensterle J, Kohler A , Hurwitz R , Kamradt T, Kaufmann S H E , Mittrucker H W . (2002). Regulatory CD4+CD25+ T Cells Restrict M e m o r y C D 8 + T C e l l Responses. J E x p M e d 196:1585-1592. Kurtz J , Shaffer J, L i e A , Anosova N , Benichou G , Sykes M . (2004a). Mechanisms o f early peripheral C D 4 T-cell tolerance induction by anti-CD 154 monoclonal antibody and allogeneic bone marrow transplantation: evidence for anergy and deletion but not regulatory cells. B l o o d 103:4336-4343. Kurtz J, Wekerle T, Sykes M . (2004b). Tolerance in mixed chimerism - a role for regulatory cells? Trends Immunol 25:518-523. Lan F , Zeng D , Higuchi M , Higgins JP, Strober S. (2003). Host conditioning with total lymphoid irradiation and antithymocyte globulin prevents graft-versus-host disease: the role o f CDl-reactive natural killer T cells. B i o l B l o o d Marrow Transplant 9:355-363. Lan F , Zeng D , Higuchi M , Huie P, Higgins JP, Strober S. (2001). Predominance o f NK1.1+TCR{alpha} {beta}+ or DX5+TCR{alpha} {beta}+ T Cells i n M i c e Conditioned with Fractionated Lymphoid Irradiation Protects Against Graft-VersusHost Disease: "Natural Suppressor" Cells. J Immunol 167:2087-2096. Lavelle E C , M c N e e l a E , Armstrong M E , Leavy O, Higgins S C , M i l l s K H G . (2003). Cholera T o x i n Promotes the Induction o f Regulatory T Cells Specific for Bystander Antigens by Modulating Dendritic Cell Activation. J Immunol 171:2384-2392. Lawson C A , B r o w n A K , Bejarano V , Douglas S H , Burgoyne C H , Greenstein A S , Boylston A W , Emery P, Ponchel F , Isaacs J D . (2006). Early rheumatoid arthritis is associated with a deficit in the CD4+CD25high regulatory T cell population in peripheral blood. Rheumatology 45:1210-1217. Lechler RI, Sykes M , Thomson A W , Turka L A . (2005). Organ transplantation: how much o f the promise has been realized? Nat M e d 11:605-613. Lee W T , Pasos G , Cecchini L , Mittler J N . (2002). Continued Antigen Stimulation Is Not Required During C D 4 + T C e l l Clonal Expansion. J Immunol 168:1682-1689.  193  Leung D , Huber B , Schlievert P. (1997). Historical perspective o f superantigens and their biological activities. In: Superantigens. Molecular biology, immunology, and relevance to human disease, ed, Leung D , Huber B , Schlievert P (eds), Marcel Dekker, Inc, pp 1-13. Levings M K , Gregori S, Tresoldi E , Cazzaniga S, Bonini C , Roncarolo M G . (2005). Differentiation o f T r l cells by immature dendritic cells requires IL-10 but not CD25+CD4+ Tr cells. B l o o d 105:1162-1169. Levings M K , Sangregorio R, Roncarolo M G . (2001). Human C D 2 5 + C D 4 + T Regulatory Cells Suppress Naive and M e m o r y T C e l l Proliferation and Can B e Expanded In Vitro without Loss o f Function. J E x p M e d 193:1295-1302. Levings M K , Sangregorio R, Sartirana C , M o s c h i n A L , Battaglia M , Orban P C , Roncarolo M G . (2002). Human CD25+CD4+ T Suppressor C e l l Clones Produce Transforming Growth Factor {beta}, but not Interleukin 10, and A r e Distinct from Type 1 T Regulatory Cells. J E x p M e d 196:1335-1346. L i J, Bracht M , Shang X , Radewonuk J, E m m e l l E , Griswold D E , L i L . (2006). E x vivo activated O V A specific and non-specific CD4+CD25+ regulatory T cells exhibit comparable suppression to O V A mediated T cell responses. Cellular Immunology 241:75-84. Liang S, A l a r d P, Zhao Y , Parnell S, Clark S L , K o s i e w i c z M M . (2005). Conversion o f C D 4 + C D 2 5 - cells into CD4+ CD25+ regulatory T cells i n vivo requires B 7 costimulation, but not the thymus. J E x p M e d 201:127-137. L i m H W , Hillsamer P, Banham A H , K i m C H . (2005). Cutting Edge: Direct Suppression o f B Cells by CD4+CD25+ Regulatory T Cells. J Immunol 175:41804183. L i m H W , Hillsamer P, K i m C H . (2004). Regulatory T cells can migrate to follicles upon T cell activation and suppress G C - T h cells and G C - T h cell-driven B cell responses. J C l i n Invest 114:1640-1649. L i n C Y , Graca L , Cobbold SP, Waldmann H . (2002). Dominant transplantation tolerance impairs C D 8 + T cell function but not expansion. Nat Immunol 3:1208-1213. Lindley S, Dayan C M , Bishop A , Roep B O , Peakman M , Tree T I M . (2005). Defective Suppressor Function in CD4+CD25+ T-Cells F r o m Patients W i t h Type 1 Diabetes. Diabetes 54:92-99. Lombardi G , Sidhu S, Batchelor R, Lechler R . (1994). Anergic T cells as suppressor cells i n vitro. Science 264:1587-1589. L o w s k y R, Takahashi T, L i u Y P , Dejbakhsh-Jones S, Grumet F C , Shizuru J A , Laport G G , Stockerl-Goldstein K E , Johnston L J , Hoppe R T , B l o c h D A , B l u m e K G , Negrin  194  R S , Strober S. (2005). Protective Conditioning for Acute Graft-versus-Host Disease. N Engl J M e d 353:1321-1331. L u L F , L i n d E F , Gondek D C , Bennett K A , Gleeson M W , Pino-Lagos K , Scott Z A , Coyle A J , Reed J L , V a n Snick J, Strom T B , Zheng X X , Noelle R J . (2006). Mast cells are essential intermediaries in regulatory T-cell tolerance. Nature 442:997-1002. Lundqvist A , Palmborg A , Pavlenko M , Levitskaya J, Pisa P. (2005). Mature Dendritic Cells Induce Tumor-Specific Type 1 Regulatory T Cells. [Miscellaneous Article]. J Immunother 28:229-235. MacDonald R H , Baschieri S, Lees R K . (1991). Clonal expansion precedes anergy and death o f V beta 8+ peripheral T cells responding to staphylococcal enterotoxin B i n vivo. Eur J Immunol 21:1963-1966. Mahnke K , Qian Y , Knop J, Enk A H . (2003). Induction o f CD4+/CD25+ regulatory T cells by targeting o f antigens to immature dendritic cells. B l o o d 101:4862-4869. M a i l e R , Pop S M , Tisch R , Collins E J , Cairns B A , Frelinger J A . (2006). Low-avidity CD81o T cells induced by incomplete antigen stimulation in vivo regulate naive higher avidity C D 8 h i T cell responses to the same antigen. Eur J Immunol 36:397-410. M a k i d a R , Hofer M F , Takase K , Cambier J C , Leung D Y M . (1996). Bacterial superantigens induce V[beta]-specific T cell receptor internalization. Molecular Immunology 33:891-900. Malek T R , Y u A , Vincek V , Scibelli P, K o n g L . (2002). C D 4 Regulatory T Cells Prevent Lethal Autoimmunity in IL-2R[beta]-Deficient M i c e : Implications for the Nonredundant Function o f IL-2. Immunity 17:167-178. Mallat Z , Gojova A , Brun V , Esposito B , Fournier N , Cottrez F , Tedgui A , Groux H . (2003). Induction o f a Regulatory T C e l l Type 1 Response Reduces the Development o f Atherosclerosis in Apolipoprotein E-Knockout M i c e . Circulation 108:1232-1237. Marie J C , Letterio JJ, Gavin M , Rudensky A Y . (2005). TGF-{beta} 1 maintains suppressor function and Foxp3 expression i n C D 4 + C D 2 5 + regulatory T cells. J Exp M e d 201:1061-1067. Marrack P, Kappler J W . (1990). The staphylococcal enterotoxins and their relatives. Science 248:705-711. Marshall N A , Vickers M A , Barker R N . (2003). Regulatory T Cells Secreting IL-10 Dominate the Immune Response to E B V Latent Membrane Protein 1. J Immunol 170:6183-6189. Massey E J , Sundstedt A , D a y M J , Corfield G , Anderton S, Wraith D C . (2002). Intranasal peptide-induced peripheral tolerance: the role o f IL-10 i n regulatory T cell  195  function within the context o f experimental autoimmune encephalomyelitis. Vet Immunol Immunopathol 87:357-372. Masteller E L , Warner M R , Tang Q, Tarbell K V , McDevitt H , Bluestone J A . (2005). Expansion o f Functional Endogenous Antigen-Specific C D 4 + C D 2 5 + Regulatory T Cells from Nonobese Diabetic M i c e . J Immunol 175:3053-3059. Masuyama Ji, Kaga S, Kano S, Minota S. (2002). A N o v e l Costimulation Pathway V i a the 4C8 Antigen for the Induction o f C D 4 + Regulatory T Cells. J Immunol 169:37103716. Matthys P, Mitera T, Heremans H , V a n Damme J, B i l l i a u A . (1995). Anti-gamma interferon and anti-interleukin-6 antibodies affect staphylococcal enterotoxin B induced weight loss, hypoglycemia, and cytokine release i n D-galactosaminesensitized and unsensitized mice. Infect Immun 63:1158-1164. M c C o r m i c k J K , Yarwood J M , Schlievert P M . (2.001). Toxic shock syndrome and bacterial superantigens: an update. A n n u Rev M i c r o b i o l 55:77-104. M c G u i r k P, M c C a n n C , M i l l s K H G . (2002). Pathogen-specific T Regulatory 1 Cells Induced i n the Respiratory Tract by a Bacterial Molecule that Stimulates Interleukin 10 Production by Dendritic Cells: A N o v e l Strategy for Evasion o f Protective T Helper Type 1 Responses by Bordetella pertussis. J E x p M e d 195:221-231. M c H u g h R S , Shevach E M . (2002). Cutting Edge: Depletion o f C D 4 + C D 2 5 + Regulatory T Cells Is Necessary, But Not Sufficient, for Induction o f Organ-Specific Autoimmune Disease. J Immunol 168:5979-5983. M c H u g h R S , Whitters M J , Piccirillo C A , Y o u n g D A , Shevach E M , Collins M , Byrne M C . (2002). C D 4 + C D 2 5 + Immunoregulatory T Cells: Gene Expression Analysis Reveals a Functional Role for the Glucocorticoid-Induced T N F Receptor. Immunity 16:311-323. M e m p e l T R , Pittet M J , Khazaie K , Weninger W , Weissleder R , von Boehmer H , von Andrian U H . (2006). Regulatory T Cells Reversibly Suppress Cytotoxic T C e l l Function Independent o f Effector Differentiation. Immunity 25:129-141. Mendez S, Reckling S K , Piccirillo C A , Sacks D , Belkaid Y . (2004). Role for CD4+ C D 2 5 + Regulatory T Cells in Reactivation o f Persistent Leishmaniasis and Control o f Concomitant Immunity. J E x p M e d 200:201-210. Miethke T, Duschek K , Wahl C , Heeg K , Wagner H . (1993). Pathogenesis o f the toxic shock syndrome: T cell mediated lethal shock caused by the superantigen T S S T - 1 . Eur J Immunol 23:1494-1500. M i l l e r C , Ragheb J A , Schwartz R H . (1999). Anergy and Cytokine-mediated Suppression as Distinct Superantigen-induced Tolerance Mechanisms In V i v o . J Exp M e d 190:53-64. 196  M i s r a N , Bayry J, Lacroix-Desmazes S, Kazatchkine M D , Kaveri S V . (2004). Cutting Edge: Human CD4+CD25+ T Cells Restrain the Maturation and Antigen-Presenting Function o f Dendritic Cells. J Immunol 172:4676-4680. M o l l e r G . (1988). D o suppressor T cells exist? Scand J Immunol 27:247-250. Moore A C , Gallimore A , Draper SJ, Watkins K R , Gilbert S C , H i l l A V S . (2005). A n t i C D 2 5 Antibody Enhancement o f Vaccine-Induced Immunogenicity: Increased Durable Cellular Immunity with Reduced Immunodominance. J Immunol 175:7264-7273. Moore K W , Waal Malefyt R, Coffman R L , O'Garra A . (2001). Interleukin-10 and the interleukin-10 receptor. A n n u Rev Immunol 19:683-765. Morris E S , M a c D o n a l d K P A , Rowe V , Johnson D H , Banovic T, Clouston A D , H i l l G R . (2004). Donor treatment with pegylated G - C S F augments the generation o f IL-10producing regulatory T cells and promotes transplantation tolerance. B l o o d 103:35733581. Mosmann T R , Cherwinski H , Bond M W , Giedlin M A , Coffman R L . (1986). T w o types o f murine helper T cell clone. I. Definition according to profiles o f lymphokine activities and secreted proteins. J Immunol 136:2348-2357. Mosmann T R , Sad S. (1996). The expanding universe o f T-cell subsets: T h l , Th2 and more. Immunology Today 17:138-146. Mottet C , U h l i g H H , Powrie F. (2003). Cutting Edge: Cure o f Colitis by CD4+CD25+ Regulatory T Cells. J Immunol 170:3939-3943. Mountz J D , Baker T J , Borcherding D R , Bluethmann H , Zhou T, Edwards C K , III. (1995). Increased susceptibility o f fas mutant M R L - l p r / l p r mice to staphylococcal enterotoxin B-induced septic shock. J Immunol 155:4829-4837. M u n n D H , M e l l o r A L . (2006). The tumor-draining lymph node as an immuneprivileged site. Immunological Reviews 213:146-158. Nakamura K , Kitani A , Fuss I, Pedersen A , Harada N , Nawata H , Strober W . (2004). TGF-{beta} 1 Plays an Important Role in the Mechanism o f C D 4 + C D 2 5 + Regulatory T C e l l Activity i n Both Humans and M i c e . J Immunol 172:834-842. Nishimura E , Sakihama T, Setoguchi R, Tanaka K , Sakaguchi S. (2004). Induction o f antigen-specific immunologic tolerance by i n vivo and i n vitro antigen-specific expansion o f naturally arising Foxp3+CD25+CD4+ regulatory T cells. Int Immunol 16:1189-1201. N o e l C , Florquin S, Goldman M , Braun M Y . (2001). Chronic exposure to superantigen induces regulatory C D 4 + T cells with IL-10-mediated suppressive activity. Int Immunol 13:431-439.  197  O'Farrell A M , L i u Y , Moore K W , M u i A L . (1998). IL-10 inhibits macrophage activation and proliferation by distinct signaling mechanisms: evidence for Stat3dependent and -independent pathways. E M B O J 17:1006-1018. Ochi H , Abraham M , Ishikawa H , Frenkel D , Y a n g K , Basso A S , W u H , Chen M L , Gandhi R, M i l l e r A , Maron R, Weiner H L . (2006). Oral CD3-specific antibody suppresses autoimmune encephalomyelitis by inducing C D 4 + C D 2 5 - L A P + T cells. Nat M e d 12:627-635. Oida T, Zhang X , Goto M , Hachimura S, Totsuka M , Kaminogawa S, Weiner H L . (2003). C D 4 + C D 2 5 - T Cells That Express Latency-Associated Peptide on the Surface Suppress CD4+CD45RBhigh-Induced Colitis by a TGF-{beta}-Dependent Mechanism. J Immunol 170:2516-2522. Ono M , Shimizu J, M i y a c h i Y , Sakaguchi S. (2006). Control o f Autoimmune Myocarditis and Multiorgan Inflammation by Glucocorticoid-Induced T N F Receptor Family-Related Proteinhigh, Foxp3-Expressing CD25+ and C D 2 5 - Regulatory T Cells. J Immunol 176:4748-4756. Overwijk W W , Theoret M R , Finkelstein S E , Surman D R , de Jong L A , Vyth-Dreese F A , Dellemijn T A , Antony P A , Spiess P J , Palmer D C , Heimann D M , Klebanoff C A , Y u Z , Hwang L N , Feigenbaum L , Kruisbeek A M , Rosenberg S A , Restifo N P . (2003). Tumor Regression and Autoimmunity after Reversal o f a Functionally Tolerant State of Self-reactive C D 8 + T Cells. J E x p M e d 198:569-580. Owens A H , Jr., Santos G W . (1968). The induction o f graft versus host diease i n mice treated with cyclophosphamide. J Exp M e d 128:277-291. Paliard X , West S G , Lafferty J A , Clements JR, Kappler J W , Marrack P, K o t z i n B L . (1991). Evidence for the effects o f a superantigen i n rheumatoid arthritis. Science 253:325-329. Pan Z , Chen Y , Zhang W , Jie Y , L i N , W u Y . (2003). Rat Corneal Allograft Survival Prolonged by the Superantigen Staphylococcal Enterotoxin B . Invest Ophthalmol V i s Sci 44:3346-3351. Papiernik M , de Moraes M L , Pontoux C , Vasseur F , Penit C . (1998). Regulatory C D 4 T cells: expression o f I L - 2 R alpha chain, resistance to clonal deletion and IL-2 dependency. Int Immunol 10:371-378. Park H B , Paik D J , Jang E , Hong S, Y o u n J. (2004). Acquisition o f anergic and suppressive activities i n transforming growth factor-{beta}-costimulated C D 4 + C D 2 5 T cells. Int Immunol 16:1203-1213. Patterson A E , Korngold R. (2001). Infusion o f select leukemia-reactive T C R Vbeta+ T cells provides graft-versus-leukemia responses with minimization o f graft-versus-host disease following murine hematopoietic stem cell transplantation. B i o l o g y o f B l o o d and M a r r o w Transplantation 7:187-196. 198  Piccirillo C A , Letterio JJ, Thornton A M , M c H u g h R S , M a m u r a M , Mizuhara H , Shevach E M . (2002). CD4+CD25+ Regulatory T Cells Can Mediate Suppressor Function in the Absence o f Transforming Growth Factor {beta} 1 Production and Responsiveness. J Exp M e d 196:237-246. Piccirillo C A , Shevach E M . (2001). Cutting Edge: Control o f C D 8 + T C e l l Activation by CD4+CD25+ Immunoregulatory Cells. J Immunol 167:1137-1140. Pless D D , Ruthel G , Reinke E K , U l r i c h R G , Bavari S. (2005). Persistence o f ZincBinding Bacterial Superantigens at the Surface o f Antigen-Presenting Cells Contributes to the Extreme Potency o f These Superantigens as T - C e l l Activators. Infect Immun 73:5358-5366. Ploix C , Bergerot I, Durand A , Czerkinsky C , Holmgren J, Thivolet C . (1999). Oral administration o f cholera toxin B-insulin conjugates protects N O D mice from autoimmune diabetes by inducing CD4+ regulatory T-cells. Diabetes 48:2150-2156. Porter D L , June C H . (2005). T-cell reconstitution and expansion after hematopoietic stem cell transplantation: A T / ' it up! 35:935-942. Prabhu Das M R , Cohen A , Z a m v i l SS, Offner H , Kuchroo V K . (1996). Prior exposure to superantigen can inhibit or exacerbate autoimmune encephalomyelitis: T-cell repertoire engaged by the autoantigen determines clinical outcome. J Neuroimmunol 71:3-10. Prasad G S , Earhart C A , Murray D L , N o v i c k R P , Schlievert P M , Ohlendorf D H . (1993). Structure o f toxic shock syndrome toxin 1. Biochemistry 32:13761-13766. Prasad G S , R A D H A K R I S H N A N R , Mitchell D T , Earhart C A , Dinges M M , Cook W J , Schlievert P M , Ohlendorf D H . (1997). Refined structures o f three crystal forms o f toxic shock syndrome toxin-1 and o f a tetramutant with reduced activity. Protein Sci 6:1220-1227. Pro ft T, Fraser J D . (2003). Bacterial superantigens. C l i n E x p Immunol 133:299-306. Pulaski B A , Terman D S , K h a n S, Muller E , Ostrand-Rosenberg S. (2000). Cooperativity o f Staphylococcal aureus Enterotoxin B Superantigen, Major Histocompatibility Complex Class II, and C D 8 0 for Immunotherapy o f Advanced Spontaneous Metastases in a Clinically Relevant Postoperative Mouse Breast Cancer M o d e l . Cancer Res 60:2710-2715. Q i n H Y , Mukherjee R, Lee-Chan E , E w e n C , Bleackley R C , Singh B . (2006). A novel mechanism o f regulatory T cell-mediated down-regulation o f autoimmunity. Int Immunol 18:1001-1015. Q i n S, Cobbold SP, Pope H , Elliott J, Kioussis D , Davies J, Waldmann H . (1993). "Infectious" transplantation tolerance. Science 259:974-977.  199  Rad R , Brenner L , Bauer S, Schwendy S, Layland L , da Costa C P , Reindl W , Dossumbekova A , Friedrich M , Saur D . (2006). CD25+/Foxp3+ T Cells Regulate Gastric Inflammation and Helicobacter pylori Colonization In V i v o . Gastroenterology 131:525-537. Read S, Malmstrom V , Powrie F. (2000). Cytotoxic T Lymphocyte-associated Antigen 4 Plays an Essential Role in the Function o f C D 2 5 + C D 4 + Regulatory Cells that Control Intestinal Inflammation. J Exp M e d 192:295-302. Reddy P, Ferrara J L M . (2003). Immunobiology o f acute graft-versus-host disease. B l o o d Rev 17:187-194. Rellahan B L , Jones L A , Kruisbeek A M , F r y A M , Matis L A . (1990). In vivo induction of anergy i n peripheral V beta 8+ T cells by staphylococcal enterotoxin B . J E x p M e d 172:1091-1100. Ren S, Terman D S , Bohach G , Silvers A , Hansen C , Colt H , Sahn S A . (2004). Intrapleural Staphylococcal Superantigen Induces Resolution o f Malignant Pleural Effusions and a Survival Benefit i n Non-Small C e l l Lung Cancer. Chest 126:15291539. Riemekasten G , Langnickel D , Enghard P, Undeutsch R , H u m r i c h J, Ebling F M , Hocher B , Humaljoki T, Neumayer H , Burmester G R , Hahn B H , Radbruch A , Hiepe F. (2004). Intravenous Injection o f a D l Protein o f the Smith Proteins Postpones Murine Lupus and Induces Type 1 Regulatory T Cells. J Immunol 173:5835-5842. R i n k L , L u h m J, Koester M , Kirchner H . (1996). Induction o f a cytokine network by superantigens with parallel T H 1 and T H 2 stimulation. J Interferon Cytokine Res 16:4147. Robertson SJ, Messer R J , Carmody A B , Hasenkrug K J . (2006). In Vitro Suppression of C D 8 + T C e l l Function by Friend Virus-Induced Regulatory T Cells. J Immunol 176:3342-3349. Romagnoli P, Hudrisier D , van Meerwijk J P M . (2005). Molecular Signature o f Recent Thymic Selection Events on Effector and Regulatory C D 4 + T Lymphocytes. J Immunol 175:5751-5758. Rouse B T , Sarangi P P , Suvas S. (2006). Regulatory T cells i n virus infections. Immunol Rev 212:272-286. Rushbrook S M , Ward S M , Unitt E , Vowler S L , Lucas M , Klenerman P, Alexander G J M . (2005). Regulatory T Cells Suppress In Vitro Proliferation o f Virus-Specific C D 8 + T Cells during Persistent Hepatitis C Virus Infection. J V i r o l 79:7852-7859. Rutella S, Pierelli L , Bonanno G , Sica S, Ameglio F , Capoluongo E , Mariotti A , Scambia G , d'Onofrio G , Leone G . (2002). Role for granulocyte colony-stimulating factor in the generation o f human T regulatory type 1 cells. B l o o d 100:2562-2571. 200  Rutella S, Zavala F , Danese S, Kared H , Leone G . (2005). Granulocyte ColonyStimulating Factor: A N o v e l Mediator o f T C e l l Tolerance. J Immunol 175:7085-7091. Rybka W B . (1994). Amplification o f the graft-versus-host reaction by cyclophosphamide: dependence on timing o f drug administration. Bone M a r r o w Transplant 13:473-477. Sabapathy T, Hwang L A , H u i K M . (1994). Differential Effect o f Staphylococcal Enterotoxin B upon the Induction o f Tolerance on Peripheral CD4+V[beta]8+ and CD8+V[beta]8+ T Cells. C e l l Immunol 158:83-95. Sadlack B , Lohler J, Schorle H , Klebb G , Haber H , Sickel E , Noelle R J , Horak I. (1995). Generalized autoimmune disease in interleukin-2-deficient mice is triggered by an uncontrolled activation and proliferation o f C D 4 + T cells. Eur J Immunol 25:30533059. Sadlack B , M e r z H , Schorle H , Schimpl A , Feller A C , Horak I. (1993). Ulcerative colitis-like disease i n mice with a disrupted interleukin-2 gene. C e l l 75:253-261. Saha B , Harlan D M , Lee K P , June C H , A b e R. (1996a). Protection against lethal toxic shock by targeted disruption o f the C D 2 8 gene. J E x p M e d 183:2675-2680. Saha B , Jaklic B , Harlan D M , Gray G S , June C H , A b e R . (1996b). Toxic shock syndrome toxin-1-induced death is prevented by C T L A 4 I g . J Immunol 157:3869-3875. Sakaguchi S, Sakaguchi N , Asano M , Itoh M , Toda M . (1995). Immunologic selftolerance maintained by activated T cells expressing I L - 2 receptor alpha-chains (CD25). Breakdown o f a single mechanism o f self-tolerance causes various autoimmune diseases. J Immunol 155:1151-1164. Sakaguchi S. (2004). Naturally Arising C D 4 + Regulatory T Cells for Immunologic Self-Tolerance and Negative Control o f Immune Responses. A n n u R e v Immunol 22:531-562. Sakaguchi S. (2005). Naturally arising Foxp3-expressing C D 2 5 + C D 4 + regulatory T cells i n immunological tolerance to self and non-self. Nat Immunol 6:345-352. Sakaguchi S, Ono M , Setoguchi R, Y a g i H , H o r i S, Fehervari Z , Shimizu J, Takahashi T, Nomura T. (2006). Foxp3+CD25+CD4+ natural regulatory T cells i n dominant selftolerance and autoimmune disease. Immunol Rev 212:8-27. Salomon B , Lenschow D J , Rhee L , Ashourian N , Singh B , Sharpe A , Bluestone J A . (2000). B 7 / C D 2 8 Costimulation Is Essential for the Homeostasis o f the CD4+CD25+ Immunoregulatory T Cells that Control Autoimmune Diabetes. Immunity 12:431-440. Sawitzki B , Kingsley C I , Oliveira V , K a r i m M , Herber M , W o o d K J . (2005). I F N {gamma} production by alloantigen-reactive regulatory T cells is important for their regulatory function i n vivo. J Exp M e d 201:1925-1935. 201  Schad E , Zaitseva I, Zaitsev V N , Dohlsten M , Kalland T, Schlievert P M , Ohlendorf D H , Svensson L A . (1995). Crystal structure o f the superantigen staphylococcal enterotoxin type A . E M B O J 14:3292-3301. Scheffold A , Huhn J, Hofer T. (2005). Regulation o f C D 4 + C D 2 5 + regulatory T cell activity: it takes (IL-)two to tango. Eur J Immunol 35:1336-1341. Schiffenbauer J, Johnson H M , Butfiloski E J , Wegrzyn L , Soos J M . (1993). Staphylococcal Enterotoxins can Reactivate Experimental Allergic Encephalomyelitis. P N A S 90:8543-8546. Schluns K S , Lefrancois L . (2003). Cytokine control o f memory T-cell development and survival. Nat Rev Immunol 3:269-279. Schmaltz C , Alpdogan Onder, Muriglan SJ, Kappel B J , Rotolo J A , Ricchetti E T , Greenberg A S , W i l l i s J M , Murphy G F , Crawford J M , van den B r i n k M R . (2007). Donor T cell-drived T N F is required for graft-versus-host disease and graft-versus-host tumor activity after bone marrow transplantation. B l o o d 101:2440-2445. Schwab J H , B r o w n R R , Anderle S K , Schlievert P M . (1993). Superantigen can reactivate bacterial cell wall-induced arthritis. Journal O f Immunology (Baltimore, M d 1950) 150:4151-4159. Seddiki N , Santner-Nanan B , Martinson J, Zaunders J, Sasson S, Landay A , Solomon M , Selby W , Alexander SI, Nanan R, Kelleher A , de St.Groth B F . (2006). Expression o f interleukin (IL)-2 and IL-7 receptors discriminates between human regulatory and activated T cells. J E x p M e d 203:1693-1700. Seddon B , Mason D . (2000). The third function o f the thymus. Immunology Today 21:95-99. Segundo D S , R u i z J C , Izquierdo M , Fernandez-Fresnedo G , Gomez-Alamillo C , Merino R , Benito M J , Cacho E , Rodrigo E , Palomar R , Lopez-Hoyos M , Arias M . (2006). Calcineurin Inhibitors, but not Rapamycin, Reduce Percentages o f C D 4 + C D 2 5 + F O X P 3 + Regulatory T Cells i n Renal Transplant Recipients. [Article]. Transplantation 82:550-557. Seo Sj, Fields M L , Buckler J L , Reed A J , Mandik-Nayak L , N i s h S A , Noelle R J , Turka L A , Finkelman F D , Caton A J , Erikson J. (2002). The Impact o f T Helper and T Regulatory Cells on the Regulation o f Anti-Double-Stranded D N A B Cells. Immunity 16:535-546. Sharif S, Arreaza G A , Zucker P E , Delovitch T L . (2002). Regulatory Natural K i l l e r T Cells Protect against Spontaneous and Recurrent Type 1 Diabetes. A n n N Y A c a d Sci 958:77-88. Sharif S, Arreaza G A , Zucker P, M i Q S , Sondhi J, Naidenko O V , Kronenberg M , Koezuka Y , Delovitch T L , Gombert J M , Leite-de-Moraes M , Gouarin C , Z h u R, 202  Hameg A , Nakayama T, Taniguchi M , Lepault F , Lehuen A , B a c h JF, Herbelin A . (2001). Activation o f natural killer T cells by [alpha]-galactosylceramide treatment prevents the onset and recurrence o f autoimmune Type 1 diabetes. Nat M e d 7:10571062. Shevach E M . (2000). Regulatory T Cells in Autoimmmunity. Immunol Rev 18:423449. Shimizu J, Yamazaki S, Sakaguchi S. (1999). Induction o f Tumor Immunity by Removing CD25+CD4+ T Cells: A Common Basis Between Tumor Immunity and Autoimmunity. J Immunol 163:5211-5218. Shimizu J, Yamazaki S, Takahashi T, Ishida Y , Sakaguchi S. (2002). Stimulation o f CD25+CD4+ regulatory T cells through G I T R breaks immunological self-tolerance. Nat Immunol 3:135-142. Shlomchik W D , Couzens M S , Tang C B , M c N i f f J, Robert M E , L i u J, Shlomchik M J , Emerson S G . (1999). Prevention o f Graft Versus Host Disease by Inactivation o f Host Antigen-Presenting Cells. Science 285:412-415. Shu S, Krinock R A , Matsumura T, Sussman JJ, F o x B A , Chang A E , Terman D S . (1994). Stimulation o f tumor-draining lymph node cells with superantigenic staphylococcal toxins leads to the generation o f tumor-specific effector T cells. J Immunol 152:1277-1288. Shull M M , Ormsby I, K i e r A B , Pawlowski S, Diebold R J , Y i n M , A l l e n R, Sidman C , Proetzel G , Calvin D , Annunziata N , Doetschman T. (1992). Targeted disruption o f the mouse transforming growth factor-[beta] 1 gene results in multifocal inflammatory disease. Nature 359:693-699. Siegmund K , Feuerer M , Siewert C , Ghani S, Haubold U , D a n k o f A , Krenn V , Schon M P , Scheffold A , Lowe J B , Hamann A , Syrbe U , Huehn J. (2005). Migration matters: regulatory T-cell compartmentalization determines suppressive activity in vivo. B l o o d 106:3097-3104. Slavin S, Nagler A , Shapira M Y , Aker M , Gabriel C , Or R. (2002). Treatment o f leukemia by alloreactive lymphocytes and nonmyeloablative stem cell transplantation. J C l i n Immunol 22:64-69. Smyth M J , Teng M W L , Swann J, Kyparissoudis K , Godfrey D l , Hayakawa Y . (2006). CD4+CD25+ T Regulatory Cells Suppress N K Cell-Mediated Immunotherapy o f Cancer. J Immunol 176:1582-1587. Soos J M , Hobeika A C , Butfiloski E J , Schiffenbauer J, Johnson H M . (1995). Accelerated Induction o f Experimental Allergic Encephalomyelitis i n P L / J M i c e by a Non-V{beta}8-Specific Superantigen. P N A S 92:6082-6086.  203  Soos J M , Schiffenbauer J, Johnson H M . (1993). Treatment o f P L / J mice with the superantigen, staphylococcal enterotoxin B , prevents development o f experimental allergic encephalomyelitis. J Neuroimmunol 43:39-43. Steinbrink K , Graulich E , Kubsch S, Knop J, Enk A H . (2002). C D 4 + and C D 8 + anergic T cells induced by interleukin-10-treated human dendritic cells display antigenspecific suppressor activity. B l o o d 99:2468-2476. Stephens L A , Mason D . (2000). C D 2 5 Is a Marker for C D 4 + Thymocytes That Prevent Autoimmune Diabetes in Rats, But Peripheral T Cells with This Function A r e Found in Both CD25+ and C D 2 5 - Subpopulations. J Immunol 165:3105-3110. Stober C B , Lange U G , Roberts M T M , A l c a m i A , B l a c k w e l l J M . (2005). IL-10 from Regulatory T Cells Determines Vaccine Efficacy i n Murine Leishmania major Infection. J Immunol 175:2517-2524. Stock P, Akbari O, Berry G , Freeman G J , D e K r u y f f R H , Umetsu D T . (2004). Induction o f T helper type 1-like regulatory cells that express Foxp3 and protect against airway hyper-reactivity. Nat Immunol 5:1149-1156. Stohlman S A , Pei L , Cua D J , L i Z , Hinton D R . (1999). Activation o f Regulatory Cells Suppresses Experimental Allergic Encephalomyelitis V i a Secretion o f IL-10. J Immunol 163:6338-6344. Strober S, Palathumpat V , Schwadron R, Hertel-Wulff B . (1987). Cloned natural suppressor cells prevent lethal graft-vs-host disease. J Immunol 138:699-703. Sundstedt A , Hoiden I, Rosendahl A , Kalland T, van Rooijen N , Dohlsten M . (1997). Immunoregulatory role o f IL-10 during superantigen-induced hyporesponsiveness in vivo. J Immunol 158:180-186. Suri-Payer E , A m a r A Z , Thornton A M , Shevach E M . (1998). C D 4 + C D 2 5 + T Cells Inhibit Both the Induction and Effector Function o f Autoreactive T Cells and Represent a Unique Lineage o f Immunoregulatory Cells. J Immunol 160:1212-1218. Taams L S , Smith J, Rustin M H , Salmon M , Poulter L W , A k b a r A N . (2001). Human anergic/suppressive CD4(+)CD25(+) T cells: a highly differentiated and apoptosisprone population. Eur J Immunol 31:1122-1131. Tabbara K S , Peters N C , Afrin F, Mendez S, Bertholet S, Belkaid Y , Sacks D L . (2005). Conditions Influencing the Efficacy o f Vaccination with L i v e Organisms against Leishmania major Infection. Infect Immun 73:4714-4722. Tadokoro C E , Shakhar G , Shen S, D i n g Y , Lino A C , Maraver A , Lafaille JJ, Dustin M L . (2006). Regulatory T cells inhibit stable contacts between C D 4 + T cells and dendritic cells i n vivo. J E x p M e d 203:505-511.  204  Takahashi T, Kuniyasu Y , Toda M , Sakaguchi N , Itoh M , Iwata M , Shimizu J, Sakaguchi S. (1998). Immunologic self-tolerance maintained by C D 2 5 + C D 4 + naturally anergic and suppressive T cells: induction o f autoimmune disease by breaking their anergic/suppressive state. Int Immunol 10:1969-1980. Takahashi T, Tagami T, Yamazaki S, Uede T, Shimizu J, Sakaguchi N , M a k T W , Sakaguchi S. (2000). Immunologic Self-Tolerance Maintained by C D 2 5 + C D 4 + Regulatory T Cells Constitutively Expressing Cytotoxic T Lymphocyte-associated Antigen 4. J E x p M e d 192:303-310. Takenaka K , Fujiyama Y , A n d o h A , Sasaki T, Amakata Y , Matsubara H , Hodohara K , Bamba T. (2001). Prevention o f murine acute graft-versus-host disease by staphylococcal enterotoxin B treatment. C l i n E x p Immunol 123:155-161. Talmadge J E . (2003). Hematopoietic stem cell graft manipulation as a mechanism o f immunotherapy. Int Immunopharmacol 3:1121-1143. Tang Q, Adams J Y , Tooley A J , B i M , Fife B T , Serra P , Santamaria P , Locksley R M , Krummel M F , Bluestone J A . (2006). Visualizing regulatory T cell control o f autoimmune responses in nonobese diabetic mice. Nat Immunol 7:83-92. Tang Q, Henriksen K J , B i M , Finger E B , Szot G , Y e J, Masteller E L , McDevitt H , Bonyhadi M , Bluestone J A . (2004). In Vitro-expanded Antigen-specific Regulatory T Cells Suppress Autoimmune Diabetes. J E x p M e d 199:1455-1465. Taylor C E , Stashak P W , Caldes G , Prescott B , Chused T E , Brooks A , Baker P J . (1983). Activation o f antigen-specific suppressor T cells by B cells from mice immunized with type III pneumococcal polysaccharide. J E x p M e d 158:703-717. Taylor M D , L e G o f f L , Harris A , Malone E , A l l e n J E , Maizels R M . (2005). Removal o f Regulatory T C e l l Activity Reverses Hyporesponsiveness and Leads to Filarial Parasite Clearance In V i v o . J Immunol 174:4924-4933. Taylor P A , Friedman T M , Korngold R, Noelle R J , Blazar B R . (2002a). Tolerance induction o f alloreactive T cells v i a ex vivo blockade o f the C D 4 0 : C D 4 0 L costimulatory pathway results in the generation o f a potent immune regulatory cell. Blood 99:4601-4609. Taylor P A , Lees C J , Blazar B R . (2002b). The infusion o f ex vivo activated and expanded CD4+CD25+ immune regulatory cells inhibits graft-versus-host disease lethality. B l o o d 99:3493-3499. Taylor P A , Noelle R J , Blazar B R . (2001). C D 4 + C D 2 5 + Immune Regulatory Cells A r e Required for Induction o f Tolerance to Alloantigen v i a Costimulatory Blockade. J E x p M e d 193:1311-1318. Taylor P A , Panoskaltsis-Mortari A , Swedin J M , Lucas P J , Gress R E , Levine B L , June C H , Serody JS, Blazar B R . (2004). L-Selectinhi but not the L-selectinlo CD4+25+ T205  regulatory cells are potent inhibitors o f G V H D and B M graft rejection. B l o o d 104:3804-3812. Teshima T, H i l l G R , Pan L , Brinson Y S , van den Brink M R M , Cooke K R , Ferrara J L M . (1999). IL-11 separates graft-versus-leukemia effects from graft-versus-host disease after bone marrow transplantation. J C l i n Invest 104:317-325. Teshima T, Ordemann R , Reddy P, Gagin S, L i u C , Cooke K R , Ferrara J L M . (2002). Acute graft-versus-host disease does not require alloantigen expression on host epithelium. Nat M e d 8:575-581. Thibodeau J, Cloutier I, Lavoie P M , Labrecque N , Mourad W , Jardetzky T, Sekaly R P . (1994). Subsets o f H L A - D R 1 molecules defined by S E B and TSST-1 binding. Science 266:1874-1878. Thornton A M , Shevach E M . (1998). CD4+CD25+ Immunoregulatory T Cells Suppress Polyclonal T Cell Activation In Vitro by Inhibiting Interleukin 2 Production. J E x p M e d 188:287-296. Tischner D , Weishaupt A , Brandt Jvd, M u l l e r N , Beyersdorf N , Ip C W , Toyka K V , Hunig T, G o l d R , Kerkau T, Reichardt H M . (2006). Polyclonal expansion o f regulatory T cells interferes with effector cell migration in a model o f multiple sclerosis. Brain 129:2635-2647. Toka F N , Suvas S, Rouse B T . (2004). CD4+ C D 2 5 + T Cells Regulate VaccineGenerated Primary and Memory C D 8 + T - C e l l Responses against Herpes Simplex Virus Type 1. J V i r o l 78:13082-13089. Torres B A , K o m i n s k y S, Perrin G Q , Hobeika A C , Johnson H M . (2001). Superantigens: The Good, the Bad, and the Ugly. E x p B i o l M e d 226:164-176. Trenado A , Charlotte F , Fisson S, Yagello M , Klatzmann D , Salomon B L , Cohen J L . (2003). Recipient-type specific CD4+CD25+ regulatory T cells favor immune reconstitution and control graft-versus-host disease while maintaining graft-versusleukemia. J C l i n invest 112:1688-1696. Trenado A , Sudres M , Tang Q, M a u r y S, Charlotte F , Gregoire S, Bonyhadi M , Klatzmann D , Salomon B L , Cohen J L . (2006). E x Vivo-Expanded C D 4 + C D 2 5 + Immunoregulatory T Cells Prevent Graft-versus-Host-Disease by Inhibiting Activation/Differentiation o f Pathogenic T Cells. J Immunol 176:1266-1273. Truitt R L . (2004). The Mortimer M . Bortin lecture: to destroy by the reaction o f immunity: the search for separation o f graft-versus-leukemia and graft-versus-host. B i o l B l o o d Marrow Transplant 10:505-523. Trzonkowski P, Szmit E , M y s l i w s k a J, M y s l i w s k i A . (2006). C D 4 + C D 2 5 + T regulatory cells inhibit cytotoxic activity o f C T L and N K cells i n humans—impact o f immunosenescence. C l i n Immunol 119:307-316. 206  Tsurufuji M , Benacerraf B , Sy M S . (1983). A n antigen-specific signal is required for the activation o f second- order suppressor T cells i n the regulation o f delayed-type hypersensitivity to 2,4,6-trinitrobenzene sulfonic acid. J E x p M e d 158:932-945. Turk M J , Guevara-Patino J A , Rizzuto G A , Engelhorn M E , Houghton A N . (2004). Concomitant Tumor Immunity to a Poorly Immunogenic Melanoma Is Prevented by Regulatory T Cells. J Exp M e d 200:771-782. Unger W W J , Jansen W , Wolvers D A W , van Halteren A G S , Kraal G , Samsom J N . (2003). Nasal tolerance induces antigen-specific C D 4 + C D 2 5 - regulatory T cells that can transfer their regulatory capacity to naive C D 4 + T cells. Int Immunol 15:731-739. Uraushihara K , Kanai T, K o K , Totsuka T, Makita S, Iiyama R , Nakamura T, Watanabe M . (2003). Regulation o f Murine Inflammatory B o w e l Disease by CD25+ and C D 2 5 - C D 4 + Glucocorticoid-Induced T N F Receptor Family-Related Gene+ Regulatory T Cells. J Immunol 171:708-716. V i a u M , Zouali M . (2005). Effect o f the B cell superantigen protein A from S. aureus on the early lupus disease o f ( N Z B x N Z W ) F l mice. M o i Immunol 42:849-855. V i e h l C T , Moore T T , Liyanage U K , Frey D M , Ehlers JP, Eberlein T J , Goedegebuure P S , Linehan D C . (2006). Depletion o f CD4+CD25+ Regulatory T Cells Promotes a Tumor-Specific Immune Response in Pancreas Cancer-Bearing M i c e . A n n Surg Oncol 13:1252-1258. Vieira P L , Christensen JR, Minaee S, O'Neill E J , Barrat F J , Boonstra A , Barthlott T, Stockinger B , Wraith D C , O'Garra A . (2004). IL-10-secreting regulatory T cells do not express Foxp3 but have comparable regulatory function to naturally occurring CD4+CD25+ regulatory T cells. J Immunol 172:5986-5993. Viglietta V , Baecher-Allan C , Weiner H L , Hafler D A . (2004). Loss o f Functional Suppression by CD4+CD25+ Regulatory T Cells in Patients with Multiple Sclerosis. J Exp M e d 199:971-979. Vigouroux S, Y v o n E , Biagi E , Brenner M K . (2004). Antigen-induced regulatory T cells. B l o o d 104:26-33. Visvanathan K , Charles A , Barman J, Pugach P, Kashfi K , Zabriskie J B . (2001). Inhibition o f Bacterial Superantigens by Peptides and Antibodies. Infect Immun 69:875-884. Waaga A M , Gasser M , Kist-van Holthe J E , Najafian N , M u l l e r A , V e l l a JP, Womer K L , Chandraker A , Khoury SJ, Sayegh M H . (2001). Regulatory functions o f selfrestricted M H C class II allopeptide-specific Th2 clones in vivo. J C l i n Invest 107:909916: Wade T, B i l l J, Marrack P C , Palmer E , Kappler J W . (1988). Molecular basis for the nonexpression o f V beta 17 in some strains o f mice. J Immunol 141:2165-2167. 207  Wahlsten J L , M i l l s C D , Ramakrishrian S. (1998). Antitumor Response Elicited by a Superantigen- Transmembrane Sequence Fusion Protein Anchored onto Tumor Cells. J Immunol 161:6761-6767. Walker L S K , Chodos A , Eggena M , Dooms H , Abbas A K . (2003). Antigen-dependent Proliferation o f C D 4 + CD25+ Regulatory T Cells In V i v o . J E x p M e d 198:249-258. Watanabe T, Masuyama J i , Sohma Y , Inazawa H , Horie K , K o j i m a K , Uemura Y , A o k i Y , K a g a S, Minota S. (2006). C D 5 2 is a novel costimulatory molecule for induction o f C D 4 + regulatory T cells. C l i n Immunol 120:247-259. Watson J L , Vicario M , Wang A , Moreto M , M c K a y D M . (2005). Immune cell activation and subsequent epithelial dysfunction by Staphylococcus enterotoxin B is attenuated by the green tea polyphenol (-)-epigallocatechin gallate. C l i n Immunol 237:7-16. W e i S, Kryczek I, Z o u W . (2006). Regulatory T-cell compartmentalization and trafficking. B l o o d 108:426-431. Weiner H L . (2001). Induction and mechanism o f action o f transforming growth factorbeta-secreting Th3 regulatory cells. Immunol Rev 182:207-214. White J, Herman A , Pullen A M , K u b o R, Kappler J W , Marrack P. (1989). The V betaspecific superantigen staphylococcal enterotoxin B : stimulation o f mature T cells and clonal deletion i n neonatal mice. Cell 56:27-35. White J, Herman Andrew, Pullen A n n M , K u b o Ralph, Kappler John W , Marrack Philippa. (1989). The V beta-specific superantigen staphylococcal enterotoxin B : stimulation o f mature T cells and clonal deletion i n neonatal mice. C e l l 56:27-35. Willerford D M , Chen J, Ferry J A , Davidson L , M a A , A l t F W . (1995). Interleukin-2 receptor [alpha] chain regulates the size and content o f the peripheral lymphoid compartment. Immunity 3:521-530. W o l f M , Schimpl A , Hunig T. (2001). Control o f T cell hyperactivation i n IL-2deficient mice by CD4(+)CD25(-) and CD4(+)CD25(+) T cells: evidence for two distinct regulatory mechanisms. Eur J Immunol 31:1637-1645. Wong P, Pamer E G . (2003). C D 8 T Cell responses to infectious pathogens. Annual Review o f Immunology 21:29-70. Wysocki C A , Jiang Q, Panoskaltsis-Mortari A , Taylor P A , M c K i n n o n K P , Su L , Blazar B R , Serody JS. (2005). Critical role for C C R 5 in the function o f donor CD4+CD25+ regulatory T cells during acute graft-versus-host disease. B l o o d 106:3300-3307. X u D , F u J, Jin L , Zhang H , Zhou C , Z o u Z , Zhao J M , Zhang B , Shi M , D i n g X , Tang Z , F u Y X , Wang F S . (2006). Circulating and Liver Resident C D 4 + C D 2 5 + Regulatory 208  T Cells Actively Influence the Antiviral Immune Response and Disease Progression in Patients with Hepatitis B . J Immunol 177:739-747. Yoshida M , K e z u k a T, Streilein J W . (2000). Participation o f Pigment Epithelium o f Iris and Ciliary B o d y in Ocular Immune Privilege. 2. Generation o f TGF-{beta} Producing Regulatory T Cells. Invest Ophthalmol V i s Sci 41:3862-3870. Young K J , DuTemple B , Phillips M J , Zhang L . (2003). Inhibition o f Graft-VersusHost Disease by Double-Negative Regulatory T Cells. J Immunol 171:134-141. Yurchenko E , Tritt M , H a y V , Shevach E M , Belkaid Y , Piccirillo C A . (2006). C C R 5 dependent homing o f naturally occurring C D 4 + regulatory T cells to sites o f Leishmania major infection favors pathogen persistence. J E x p M e d :jem. Zeiser R , Nguyen V H , Beilhack A , Buess M , Schulz S, Baker J, Contag C H , Negrin R S . (2006). Inhibition o f CD4+CD25+ regulatory T-cell function by calcineurindependent interleukin-2 production. B l o o d 108:390-399. Zeller J C , Panoskaltsis-Mortari A , M u r p h y W J , Ruscetti F W , Narula S, Roncarolo M G , Blazar B R . (1999). Induction o f C D 4 + T C e l l Alloantigen-Specific Hyporesponsiveness by IL-10 and TGF-{beta}. J Immunol 163:3684-3691. Zeng D , Lewis D , Dejbakhsh-Jones S, L a n F , Garcia-Ojeda M , Sibley R , Strober S. (1999). Bone Marrow N K 1 . 1 - and N K 1 . 1 + T Cells Reciprocally Regulate Acute Graft versus Host Disease. J E x p M e d 189:1073-1081. Zeng D , L i u Y , Sidobre S, Kronenberg M , Strober S. (2003). Activation o f natural killer T cells in N Z B / W mice induces Thl-type immune responses exacerbating lupus. J C l i n Invest 112:1211-1222. Zhang X , Izikson L , L i u L , Weiner H L . (2001). Activation o f C D 2 5 + C D 4 + Regulatory T Cells by Oral Antigen Administration. J Immunol 167:4245-4253. Zhang Y , Joe G , Hexner E , Z h u J, Emerson S G . (2005a). Alloreactive M e m o r y T Cells Are Responsible for the Persistence o f Graft-versus-Host Disease. J Immunol 174:3051-3058. Zhang Y , Joe G , Hexner E , Z h u J, Emerson S G . (2005b). Host-reactive C D 8 + memory stem cells in graft-versus-host disease. Nat M e d 11:1299-1305. Zhang Z X , Y a n g L , Y o u n g K J , DuTemple B , Zhang L . (2000). Identification o f a previously unknown antigen-specific regulatory T cell and its mechanism o f suppression. Nat M e d 6:782-789. Zhao D M , Thornton A M , DiPaolo R J , Shevach E M . (2006). Activated CD4+CD25+ T cells selectively k i l l B lymphocytes. B l o o d 107:3925-3932.  209  Zheng S G , Gray J D , Ohtsuka K , Yamagiwa S, Horwitz D A . (2002). Generation E x V i v o o f TGF-{beta}-Producing Regulatory T Cells from C D 4 + C D 2 5 - Precursors. J Immunol 169:4183-4189. Zheng S G , Wang J H , Koss M N , Quismorio F , Jr., Gray J D , Horwitz D A . (2004). C D 4 + and C D 8 + Regulatory T Cells Generated E x V i v o with I L - 2 and TGF-{beta} Suppress a Stimulatory Graft-versus-Host Disease with a Lupus-Like Syndrome. J Immunol 172:1531-1539. Z o u W . (2005). Immunosuppressive networks i n the tumour environment and their therapeutic relevance. Nat Rev Cancer 5:263-274.  210  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100485/manifest

Comment

Related Items