Dissecting The Role Of TNFα In Kawasaki Disease: Alteration Of Cell Fate By TNFα After Superantigen

Activation

by

Aaron Wong

A thesis submitted in conformity with the requirements for the degree of Master of Science

Department of Immunology University of Toronto

© Copyright by Aaron Wong 2011

ii

Dissecting The Role Of TNFα In Kawasaki Disease: Alteration Of

Cell Fate By TNFα After Superantigen Activation

Aaron Wong

Master of Science

Department of Immunology University of Toronto

2011

Abstract

Kawasaki disease (KD) is an acute inflammatory disease characterized by persistent

inflammation of the coronary arteries. KD is characterized by the release of cytokines such as

tumor necrosis factor alpha (TNFα) and is thought to be initiated by a superantigen (SAg). The

Lactobacillus casei cell wall extract model of KD demonstrates a critical requirement for TNFα

and its receptor during pathogenesis, although the precise effect of TNFα is unknown. A

persistent T cell infiltrate in the coronary artery disagrees with established fates of SAg activated

cells, which undergo apoptosis. In this work, TNFα was found to promote the survival of SAg-

reactive T cells. The results demonstrate that TNFα regulates B7.2 molecule expression on

antigen presenting cells, and that TNFα indirectly promotes the survival of SEB-stimulated T

cells by driving costimulation. These observations demonstrate how TNFα prevents T cell

apoptosis and lend support to KD therapies which target TNFα and B7.

iii

Acknowledgments

I would like to thank my supervisor Dr. Rae Yeung and my committee members Dr. Tania Watts

and Dr. Eleanor Fish. I would also like to thank Dr. Pamela Ohashi for generously offering her

reagents and mice. Last, I would like to thank the members of the Yeung Lab.

iv

Table of Contents

Page

Abstract ii

Acknowledgements iii

Table of Contents iv

List of Figures viii

List of Abbreviations ix

v

1 Introduction

1.1 Kawasaki Disease

1.1.1 Overview of Kawasaki Disease 1

1.1.2 Clinical Kawasaki Disease 1

1.1.3 Etiology 3

1.1.4 Immunological Findings 4

1.2 Superantigens

1.2.1 Superantigens and SEB 7

1.2.2 Cell fate after SEB stimulation 9

1.2.3 Superantigens in Kawasaki Disease 11

1.3 Animal models of KD 12

1.3.1 The LCWE mouse model 13

1.3.2 Superantigens in LCWE mouse model 15

1.3.3 TNFα in the LCWE mouse model 15

1.4 TNFα

1.4.1 Brief overview 16

1.4.2 TNFα and cell survival 20

1.4.3 TNFα during superantigen responses and autoimmunity 22

vi

1.5 T cell Activation and Costimulation

1.5.1 T cell receptor signaling with conventional antigens 23

1.5.2 Costimulation with CD28, B7 ligands, CTLA-4, and Cbl-b 26

1.5.3 T cell activation and costimulation with superantigens 30

1.5.4 Costimulation and apoptotic molecules 31

2 RATIONALE 32

3 HYPOTHESIS 32

4 EXPERIMENTAL AIMS 33

5 METHODS AND MATERIALS

5.1 Experimental Mice 34

5.2 Cell Culture 34

5.3 Purification of T cells and Antigen Presenting Cells 34

5.4 Antibodies and Flow Cytometry 35

5.5 Measurement of Proliferation 36

5.5.1 [3H]-thymidine incorporation assay 36

5.5.2 BrdU incorporation assay 36

5.6 Statistical Analysis 36

vii

6 RESULTS

6.1 SEB Upregulates B7.2 Ligand on B cells but not B7.1 37

6.2 TNFα and its receptors regulate B7.2 expression on B cells 39

6.3 Enhanced costimulation rescues SEB-reactive cells from apoptosis 42

6.4 TNFα enhances cell survival in a dose dependent manner 47

6.5 TNFα receptors on APC regulate superantigen-induced death of T cells 50

6.6 TNFα promotes survival of T cells via TNFR on APC 52

6.7 TNFR does not regulate SEB-induced proliferation 54

6.8 Contribution of B7.2 on APC and T cells to survival of SEB-stimulated T cells 55

6.9 Enhanced costimulation increases Bcl-XL and Bcl-2 expression 57

6.10 TNFR promotes Bcl-2 and Bcl-XL expression in SEB-reactive T cells 59

6.11 TNFR contributes to the activity of surviving Vβ8+ T cells 59

7 DISCUSSION

7.1 TNFα drives survival of SAg-reactive T cells by costimulation 62

7.2 TNFα and autoimmunity 69

7.3 Future Directions 70

7.4 Conclusion 71

8 REFERENCES 73

viii

List of Figures

Fig 1 Superantigens bind to the variable beta chain of the T cell receptor

8

Fig 2 Overview of the LCWE model of Kawasaki Disease 14 Fig 3 TNFR1 and TNFR2 drive apoptosis, survival, differentiation,

and proliferation through different cell signaling pathways 19

Fig 4 TNFα drives apoptotic and survival signaling pathways. 21 Fig 5 The T cell receptor complex is made up of antigen-

recognition proteins and invariant signaling proteins. 23

Fig 6 Different intracellular pathways result from TCR activation. 25 Fig 7 CD28 Signaling Enhances TCR Signaling via NF-κB, NFAT, and

AP-1 Transcription factors. 28

Fig 8 Stimulation with SEB increases B7.2 expression on B cells 38 Fig 9 TNFα upregulates B7.2 expression on B cells in a dose

dependent manner 40

Fig 10 TNF receptors on B cells regulate the expression of B7.2 during SEB stimulation

41

Fig 11 Enhanced costimulation rescues SEB-stimulated T cells from apoptosis

43

Fig 12 Blocking costimulation with CTLA4-Ig increases apoptosis of Vβ8+ T cells

45

Fig 13 Deletion of Cbl-b enhances cell survival 46 Fig 14 TNFα enhances survival of T cells in a dose dependent

manner 48

Fig 15 Neutralization of endogenous TNFα decreases survival of SEB-reactive T cells

49

Fig 16 TNFR on APC regulate survival of SEB-reactive T cells 51 Fig 17 TNFR on APC direct T cell survival 53 Fig 18 Lack of TNFRs does not affect SEB-induced proliferation 54 Fig 19 Contribution of B7.2 on APC and T cells to survival of SEB-

reactive T cells 56

Fig 20 Enhanced costimulation increases expression of anti-apoptotic proteins

58

Fig 21 TNFR promotes the expression of survival molecules in SEB-reactive T cells

60

Fig 22 TNFR contributes to the activity of surviving Vβ8+ T cells 61 Fig 23 Model for TNFα-driven survival of SAg activated T cells 72

ix

List of Abbreviations

7AAD 7-Aminoactinomycin D AP-1 Activator Protein 1 APC Antigen Presenting Cell Bcl B cell lymphoma BrdU Bromodeoxyuridine Ca Calcium CAD Caspase activated DNAse CAWS Candida Albicans Water Soluble Cbl-b Casitas B-lineage Lymphoma CD Cluster of Differentation cDNA Complentary Deoxyribonucleic acid CFSE Carboxyfluorescein succinimidyl ester c-IAP Cellular Inhibitor of Apoptosis CTLA-4 Cytotoxic T lymphocyte-associated molecule-4 DPM Disintegrations per minute ERK Extracellular regulated MAP-Kinase EVG Elastic Van Gieson FADD Fas-Associated protein with Death Domain FoxP3 Forkhead box P3 GEF Guanine nucleotide exchange factor GITR Glucocorticoid-induced TNFR-related protein GWAS Genome Wide Association Study H&E Hematoxylin and eosin IFN Interferon Ig Immunoglobulin IKK IkB Kinase IL Interleukin IP3 Inositol trisphosphate ITAM Immunoreceptor tyrosine-based activation motif ITPKC Inositol 1,4,5-trisphosphate 3-kinase IVIG Intervenous immunoglobulin JNK c-Jun N-terminal kinase KD Kawasaki Disease LCWE Lactobacillus casei cell wall extract LPS Lipopolysaccharide MCP-1 Monocyte chemotactic protein-1 MFI Mean fluorescence intensity MHC II Major histocompatibility complex II

x

MMP Matrix metalloproteinase MMTV Mouse mammory tumor virus NFAT Nuclear factor of activated T-cells NFKB Nuclear factor kappa B NOD Non obese diabetic mice PCC Pigeon cytochrome C PDK1 Pyruvate dehydrogenase PIP2 Phosphatidylinositol 4,5 bisphosphate PLAD Pre-ligand-binding assembly domain PLCγ Phospholipase C gamma RA Rheumatoid arthritis RIP Receptor-interacting protein SAg Superantigen SEA Staphylococcal enterotoxin A SEB Staphylococcal enterotoxin B SEM Standard error of measurement SLP-76 Lymphocyte cytosolic protein 2 SPE A Streptococcal pyrogenic exotoxin A T1DM Type 1 Diabetes Mellitus TCR T cell receptor Th1 T helper cell 1 Th17 T helper cell 17 Th2 T helper cell 2 TNFα Tumor necrosis factor alpha TNFR TNFα receptor TRADD TNFα-associated DEATH domain protein TRAF TNFα receptor associated factor Treg T regulatory cell TSST-1 Toxic shock syndrome toxin -1 Vb Variable beta VEGF Vascular endothelial growth factor ZAP-70 Zeta-chain-associated protein kinase 70

1

1 Introduction

1.1 Kawasaki Disease

1.1.1 Overview of Kawasaki Disease

Kawasaki Disease (KD) was first characterized by Dr. Tomasaku Kawasaki in

1967, as a puzzling inflammatory condition in children, which had lymphoid involvement and

specific desquamation of the fingers and toes [1]. Also known as “mucocutaneous lymph node

syndrome”, it has since been characterized as an acute, pediatric self-limiting, multi-systemic

inflammatory condition, where children have the potential to develop life-threatening cardiac

abnormalities, such as coronary artery aneurysms or ectasia, as well as ischemic heart disease

and sudden death [2]. The persistent inflammation of the coronary arteries has been shown to

lead to vessel wall breakdown and cardiac complications are seen in 20-25% of untreated

children [3]. As a result, KD is the most common cause of multisystem vasculitis and acquired

heart disease in the US and Japan [4]. Currently, the etiology of KD is unknown, although many

epidemiological studies suggest that an infectious trigger may be involved [2]. Treatment of KD

begins with careful diagnosis of KD by clincal criteria which include prolonged fever of > 5

days, and 4 of 5 other principal features: cervical adenopathy, polymorphous skin rash, extremity

changes, nonpurulant conjectivitis, and oral mucosal changes [5]. After KD is diagnosed,

children are treated within 10 days of the fever onset with high dose aspirin and intravenous

immunoglobulin therapy [2]. Despite ongoing research in KD, many aspects of the pathogenesis

remain unknown. The etiology of the disease is still unclear and mechanisms governing how

autoimmunity arises in this condition remain elusive.

1.1.2 Clinical Kawasaki Disease

Kawasaki Disease occurs mostly in children from 6 months to 5 years of age,

although it may occur in both younger and older individuals too [6]. The disease is broken down

into three phases: the acute, sub-acute, and convalescent phases. The acute phase lasts 7 to 14

days and is characterized by fever, conjunctival injection, mouth and lip changes, swelling and

erythema of the hands and feet, rash and cervical lymphadenopathy [7]. The subacute phase

2

starts after the fever resolves and lasts until the resolution of inflammation. During this phase,

patients may have desquamation of the fingers and toes. The convalescent phase normally occurs

between 6-8 weeks after the onset and during this phase clinical signs are absent and the

erythrocyte sedimentation rate (ESR) returns to normal.

Diagnosis of KD remains difficult and based entirely on clinical features because

there is no specific test or biomarker available for the disease. Aside from the clinical signs and

symptoms of KD, there are also significant immunological laboratory findings. Leukocytosis is

common during the acute phase, with approximately 50% of patients having white blood cell

counts above 15 000/mm3 [2]. Classic markers of inflammation are also elevated, including acute

phase reactants like ESR and C-reactive protein (CRP). During the subacute phase,

thrombocytosis occurs with elevated platelet counts often reaching greater than 1 000 000/mm3.

Although these laboratory findings are non specific, they add additional information to help

diagnose patients who are suspected, but do not fulfill, the complete criteria for KD [2].

After a diagnosis of KD is made, therapy begins with the aim to prevent cardiac

disease by decreasing inflammation. In North America, the American Heart Association’s

recommendations include a single dose of intravenous immunoglobulin (IVIG) given during the

acute phase of KD. This has been found to significantly reduce the prevalence of coronary artery

abnormalities if given during the acute phase, typically before the 10th day of fever [8-11]. The

mechanism of action of IVIG in KD is unknown, but it appears to have a generalized anti-

inflammatory effect. Potential mechanisms of IVIG action include: modulation of cytokine

production, neutralization of bacterial or viral proteins, augmentation of suppressor T cell

activity, or inhibition of antibody synthesis [2]. In addition to IVIG, high-dose aspirin is also

used as an important anti-inflammatory. Although treatment with aspirin alone does not seem to

lower the frequency of coronary abnormalities [8], when used in combination with IVIG, there

seems to be a synergistic effect [2].

3

Treatment with aspirin and IVIG results in subsidence of inflammatory symptoms

within 48 hours after IVIG infusion [12]. However, approximately 5-15% of patients will fail to

respond to this treatment, and have persistent fever and elevated CRP levels. In these cases, a

second course of IVIG therapy is often given and 70-80% of patients will typically respond to

treatment. In cases of IVIG-refractory KD, corticosteroids and other therapies that target tumor

necrosis factor alpha (TNFα) have also been used. Corticosteroids have been shown to reduce

fever [13, 14] but the effects on coronary artery abnormalities are still unclear. TNFα-based

therapies have also been effective in salvage therapy and include pentoxifylline, which specially

inhibits TNFα messenger TNF transcription [15], and Infliximab, a humanized monoclonal

antibody against TNFα which neutralizes it [16, 17].

1.1.3 Etiology

The etiology of KD is unknown although most investigators agree that it involves

an infectious trigger. Epidemiological studies have shown that KD is endemic, with seasonal

fluctuations [18, 19]. In North America, KD is more common during the winter and early spring

months [20], and appears to occur in geographical outbreaks. Furthermore, many different

viruses and bacteria have been isolated from KD patients and include: coronavirus [21],

cytomegalovirus [22-25], Epstein-Barr virus [24, 26-28], parvovirus [29], and members of the

streptococci [30, 31], staphylococci [30], and mycoplasma [32, 33] families of bacteria. Despite

these findings, no single bacterium or virus has been consistently isolated from KD patients.

It has also become clear that genetics play a strong role in the susceptibility of

children to KD. Asian children of Japanese, Korean or Chinese decent have the highest

incidence, with the disease being most prevalent in Japan. A recent epidemiological study has

reported an average annual incidence rate of 184.6 per 100000 Japanese children aged 0-4 years

[34]. Other ethnic groups are affected as well, but Japanese children have incidence rates

approximately tenfold higher than those of Caucasians [20, 34]. This increased incidence seems

to be independent of geography as Japanese children living in a Western environment are still

more prone to develop KD [20]. These data are further supported by studies which have looked

4

at the incidences of KD within families. The risk of occurrence in twins is 13% [35, 36], and

within 1 year after onset of the first case in a family, the rate in a sibling is 2.1% (which is a

relative risk of approximately ten-fold) [37, 38].

Although the precise etiology of KD remains unknown, clinical and

epidemiological studies strongly suggest an infectious agent which causes disease in young

susceptible children, who likely share a genetic predisposition to an inflammatory response to

infection. The clinical signs of a self-limited, inflammatory disease, with fever, rash, enanthem,

conjunctival injection, and cervical adenopathy correspond nicely with an infectious etiology.

The epidemiological features such as the seasonality, punctuated outbreaks, and geographical

spread also align closely with this hypothesis.

1.1.4 Immunological Findings

Research on the immune system in KD patients has not been completely

straightforward. There are numerous ethical and practical considerations that need to be made

when studying a pediatric population. Over the last 40 years, studies on the immune system have

come from the analysis of patient plasma, immunohistochemical analysis of damaged vessels on

autopsy and, more recently, the use of molecular genetics. The findings can be categorized in

terms of: cells types, cytokines, growth factors, and enzymes implicated in KD.

Although neutrophils make up the dominant inflammatory cell type in the

peripheral blood of KD patients, at the coronary artery, they are not the predominant cell type

[39]. Instead, histological studies have shown that the inflammatory infiltrate in KD tissues

contains large mononuclear cells, T cells, macrophages, dendritic cells, plasma cells, and a

limited number of neutrophils [39-42]. In the early stages of KD, an influx of neutrophils is seen,

transitioning after 7-9 days to large mononuclear cells with a high proportion of T cells [40].

Although other cell types have been studied in KD including B cells [41, 43] and macrophages

[44, 45], T cells appear to play a key role in disease as they are seen to co-localize and

5

accumulate with myeloid dendritic cells in coronary artery lesions [46] and also appear at the

other sites of KD inflammation [47]. The functional state of T cells in KD has also been

investigated. Studies demonstrated that mRNA expression levels of T-bet and GATA-3 (which

play critical roles in Th1 and Th2 polarization, respectively), are significantly decreased in the

blood mononuclear cells from KD patients [48]. Proportions of regulatory CD4+ CD25+ T cells

were significantly lower in patients with acute-phase KD and mRNA expression levels of Foxp3,

CTLA4, and GITH were also found to be decreased [49]. Supporting these results, current

investigations have focused on the more recently discovered Th17 subset of helper cells [50]

which are posited to be the reciprocal arm to regulatory T cells [51]. It has been shown that Th17

cell proportions and frequencies of CD4+ IL-17+ T cells are increased in patients with KD, and

had a tendency to decrease after treatment of IVIG [52]. This group also confirmed that

regulatory T cell frequencies were lower among KD patients. Taken together, these findings

suggest that T cells play a role in disease pathogenesis. These observations are also in keeping

with the view that KD is an inflammatory disease which results from a dyregulated immune

response to an infectious agent.

Results from a genome wide association study (GWAS) have implicated

particular genes involved in T cell activation in the pathogenesis of KD [53]. Different signaling

molecules have since been reported to play important roles in predisposing individuals to KD

[54]. For example, a single nucleotide polymorphism in inositol 1,4,5-triphosphate 3-kinase

(ITPKC) was found to associate with disease susceptibility and coronary artery lesion formation

[55]. ITPKC is an important kinase that regulates TCR signaling and T-cell activation by

phosphorylating inositol 1,4,5-triphosphate (IP3) [56, 57]. During T cell activation, stimulation

of the TCR leads to activation of phopholipase C-gamma (PLC-γ), which then acts to produce

IP3 by cleavage of phosphatidylinositol bisphosphate (PIP2). IP3 binds to IP3 receptors present

on the endoplasmic reticulum and calcium is released, leading to the activation of the

phosphatase, calcineurin. Calcineurin activates the transcription factor nuclear factor of activated

T cells (NFAT), which then drives specific gene transcription leading to cell proliferation and

differentiation. ITPKC is thus a negative regulator of T cell activation because it downregulates

the calcium/NFAT signaling [55]. Impairment in the activity of ITPKC, as in the KD associated

polymorphism identified, would contribute to a hyperactive T cell phenotype.

6

Other T cell associated genes encoding costimulatory receptors were identified in

studies of children with KD. In particular, members of the TNF-superfamily of costimulatory

receptors such as CD40 ligand (CD40L) and programmed death-1 (PD-1) were associated with

susceptibility and predisposition to poor coronary outcomes [58-60]. CD40L and PD-1 have

been shown to be expressed on the surface of activated T cells stimulated with anti-CD3

antibodies [61, 62] and have both been demonstrated to regulate the activation of T cells [63,

64]. During the acute phase of KD, it has also been reported that CD40L expression is higher on

CD4+ T cells and that expression correlated to coronary artery lesions and disease progression

[65]. Platelet expressed CD40L has been found to trigger thrombotic events and inflammatory

reactions on endothelial cells [66, 67]. These observations fit nicely with clinical observations as

thrombocytosis is one key pathognomic feature of KD.

The characteristic cytokines which accompany KD symptoms match those of an

overwhelming infection. Proinflammatory cytokines such as TNFα [68, 69], TNFβ [68], IFNγ

[70, 71], IL-6 [68, 72], IL-8 [72], IL-1[73], and IL-17 [52, 74] have all been reported to increase

during the acute phase of disease. This cytokine profile may be reflective of Th1 and Th17

responses, as TNFα, IFNγ, and TNFβ are produced by Th1 cells [75] and IL-1 and IL-6 promote

the differentiation of IL-17-producing Th17 cells [76]. In particular, increases in TNFα [68] and

also its shed receptor (p60) [77] were found in the sera of KD patients, and the extent of ligand

and receptor levels positively correlated with the severity of coronary artery lesions. These

findings are important because they support the use of biologics that interfere with TNFα as a

therapy for the treatment of KD [2]. Interestingly, plasma levels of IL-4 and IL-10 were also

found to be significantly higher in the acute phase, but their levels correlated inversely with each

other [70]. Since this study, genetic approaches have shown that IL-10 polymorphism are

associated with coronary artery lesions in acute KD [78], however, conflicting studies have also

reported both positive and negative roles of IL-4 [79, 80].

Other enzymes and soluble factors such as MMP9 and MMP2 have also been

reported to have roles in KD [81] and have previously been reported to be produced by T cells

and macrophages [82, 83]. Coronary artery aneurysms had high levels of MMP-9 expression,

7

while MMP-2 expression was most prominent in the thickened neointima and in endothelial cells

[84]. Although circulating levels of MMP-9 were reported to be increased in sera from patients

with acute KD [85], systemic MMP-9 concentrations and enzymatic activity did not correlate

with coronary outcomes [81]. Instead, it is thought that MMP-9 plays a localized role in

aneurysm formation, by degrading elastin in an animal model of KD [86, 87].

1.2 Superantigens

1.2.1 Superantigens and SEB

Superantigens (SAg) are bacterial and viral proteins that are potent activators of T

cells. In contrast to conventional antigens, which are first processed within antigen presenting

cells, loaded into the peptide binding groove of MHC, and presented to T cells bearing an

antigen specific TCR, SAg are able to circumvent these processes. Superantigens bind directly to

MHC, outside the peptide groove and away from the TCR, enabling them to bypass MHC

restriction and TCR specificity (Figure 1). This SAg binding mechanism allows for activation of

a large proportion of T cells (up to 25% of all T cells) in comparison to conventional antigens

(0.0001% of all T cells) [88].

There are two classifications of SAg: endogenous and exogenous. Endogenous

SAg include the minor lymphocyte stimulating (mls) determinants, which are the result of

chromosome integrated mouse mammary tumor virus (MMTV) [89, 90]. The existence of these

determinants within the mouse genome results in their expression in the thymus and subsequent

negative selection of lymphocytes which react with MMTV SAg. This results in resistance to

exogenous infection by MMTV [88].

8

Figure 1: Superantigens bind to the variable beta chain of the T cell receptor

Activation of naive T cells with conventional antigens proceeds when peptide is presented on MHC class II molecules to T cells bearing a specific TCR. This process activates ~0.0001% of all T cells and requires antigen processing. In contrast to conventional antigens, superantigens bind to the TCR outside of the peptide groove; bypassing antigen processing and peptide specificity. This binding mechanism results in activation of large proportion of T cells (up to ~25% of all T cells).

9

Exogenous SAg are produced by many types of bacteria and viruses. The most

well studied groups of SAg are those produced by Staphylococcus and Streptococcus bacteria.

Crystallization studies of bacterial SAg have determined that SAg bind to specific variable beta

(Vβ) chains of the TCR [91-93]. Importantly, they can also bind to multiple Vβ chains, inducing

specific activation and proliferation of T cells which bear the target Vβ TCRs. This binding leads

to a characteristic Vβ “signature”, where specific Vβ families are expanded. For example,

Staphylococcal Enterotoxin B (SEB) is known to expand Vβ 3, 7, 8.1, 8.2, 8.3, and 17 in mice

[94], while staphylococcal Enterotoxin A (SEA) expands Vβ 1, 5.3, 6.3, 6.4, 6.9, 7.4, 9.1, and 23

[95].

Teleologically, SAg are thought to function as virulence factor bacteria allowing

them to evade the immune system. Expression of SAg has been shown to prevent the

development of a normal immune response [96]. Antigen-specific T cells are noted to fail to

proliferate in response to antigens that are presented normally by MHC class II because they are

rendered “angeric” (see 1.2.2 Cell fates). In addition, SAg stimulation of T cells has also been

shown to induce IL-2 production initially, but result in inhibition of IL-2 release thereafter [97].

Thus, SAg might act to assist microbes in evading the immune system by depleting IL-2,

interfering with T cell help, and by rendering T cells anergic [88].

1.2.2 Cell Fate after SEB Stimulation

Similar to in vitro studies [94, 98-100], after mice are injected with SEB, Vβ8

reactive CD4 and CD8+ T cells undergo multiple rounds of proliferation [101]. Reactive Vβ8+ T

cells are seen to undergo increased cell cycling and expand for about 4 days following injection.

However, after this phase of rapid cell expansion, the reactive cells undergo apoptosis or are

rendered “anergic” [102], and the number of SEB-reactive cells eventually reach lower levels

than those found in untreated mice. The decrease is mainly in CD4+ Vβ8+ T cells and represents

apoptotic processes [103], although the specific mechanism for inducing death is still unclear

[101]. Many investigators have suggested that survival of SEB-stimulated cells is heavily

dependent on the growth factor IL-2, but initial reports were controversial [104, 105]. In

10

addition, it has become clear that other molecules, such as Fas [106], TNF and its receptors

[107], and the lymphoid environment (antigen presenting cells) [108] might also have potent

roles in SEB-induced apoptosis.

Although large proportions of SEB-stimulated cells undergo apoptosis, another

population of “anergic” Vβ8+ cells remain afterwards. They have been called anergic because

they fail to proliferate after subsequent exposure to SEB in vitro, while their response to other

SAg remains intact [103, 109, 110]. This lack of proliferation corresponds to a defect of the

Vβ8+ T cells to secrete IL-2 in vitro upon restimulation with SEB. The cells have lower

transcription of IL-2 and decreased amounts of steady state IL-2 mRNA [111]. Furthermore,

SAg-induced anergy has been shown to be reversed if IL-2 is given in vivo [112]. This led to the

idea that anergy results from a simple withdrawl of costimulatory cytokines from an IL-2 poor

environment (due to high catabolism of IL-2 by T cells) [113]. However, later reports have

demonstrated that SAg-induced anergy results mainly from cell autonomous defects [114], which

likely include: impaired TCR signaling [115] and inhibited expression of transcription factors

[116]. Some investigators were reluctant to acknowledge that anergic cells could truly exist,

because of the response/anergy paradox: that multiple cell fates (death and anergy) can result

from the same initial stimulation [101]. Instead, they proposed that unresponsiveness was due to

the failure of some SAg-reactive T cells to be sufficiently activated in the first place [117], or

that anergy was simply ‘slow deletion’ since Vβ8+ cells would eventually disappear upon

repeated stimulation of SAg [118].

More recently, a few elegant studies have demonstrated that T cell anergy is

regulated and maintained by persistent exposure to antigen. In the first study, Rocha and

colleagues adoptively transferred transgenic CD8+ T cells specific for the H-Y antigen into nude

male mice which were thymectomized. Much like exposure to SAg, the CD8+ T cells rapidly

expanded then declined in numbers and did not proliferate upon restimulation in vitro with male

spleen cells or anti-TCR antibodies. When these anergic cells were harvested and re-transferred

into a second female nude recipient, the anergized cells regained their ability to proliferate. The

authors concluded that antigen persistence was required to maintain the unresponsive state [119].

11

In addition, in another study, CD4+ TCR transgenic T cells were adoptively transferred into

irradiated syngenic mice, then the mice were challenged with i.v. injections of the soluble

peptide antigen. In this study, the T cells similarly expanded and a residual subset of cells were

found to be unresponsive to antigen restimulation in vivo and in vitro. Although this anergy did

reverse over the course of 7 weeks, if peptide was reinjected every week the anergy was

maintained [120]. In support of both of these studies, a third report by Tanchot et al. used a

double transgenic model to investigate anergy. The authors transferred CD4+ TCR transgenic T

cells on a RAG2 -/- background into a CD3ε -/- mouse expressing the antigen pigeon cytochrome

c (PCC), under an MHC class I promoter and an Ig enhancer. Since the CD4+ T cells underwent

a enormous expansion (50-100 fold), a large residual population of anergic cells was observed

(10-40 million) which could be followed for as long as 5 months. In keeping with previous

reports, when these cells were adoptively transferred into CD3ε -/- mice without PCC, the cells

reversed their tolerance to PCC [121]. This confirmed that the anergic state is regulated by

persistent exposure to antigen and can be maintained.

1.2.3 Superantigens in Kawasaki Disease

There is currently a debate over the nature of the immune response in KD, and

whether it is driven by a SAg or conventional antigen. Multiple lines of evidence support a role

for SAg in KD [30]. Early clinical symptoms of KD appear similar to other SAg-mediated

diseases such as toxic shock syndrome (TSS), scarlet fever, and streptococcal toxic shock

syndrome (STTS). These diseases are all characterized by massive T cell activation and also

share the same profile of proinflammatory cytokines: IL-2, TNFα, IL-1, and IL-6 [122].

Investigators have also isolated SAg producing bacteria from children with acute KD [123-125].

In one study, 11 of 13 SAg positive cultures from KD patients contained TSST-1 secreting S.

aureus [39], which has led to many investigators focusing their efforts on TSST-1. Although a

variety of bacterial and viral species have been isolated from KD patients, most share the

capacity to produce SAg (for example: Coronavirus [21], Cytomegalovirus [22-25], Epstein-Barr

virus [24, 26-28], Parvovirus [29], and Streptococci [30, 31], Staphylococci [30], and

Mycoplasma [32, 33] family of bacteria). Furthermore, the populations of T cells of children

with KD are also characteristic of SAg responses (see 1.2.1). Characteristic ‘V beta skewing’ of

12

Vβ2+ and Vβ8.1+ T cells was first reported in KD patients by Leung and colleagues [126].

Subsequent studies have confirmed the role of a SAg in skewing the T cell repertoire [126-132]

and demonstrated infiltration of affected cardiac tissues by the polyclonal SAg-reactive TCR Vβ

families [133]. Serological evidence also points to the presence of SAg in KD infections.

Significant elevation of IgM antibodies against staphylococcal Enterotoxin A and B, TSST-1,

and streptococcal exotoxin A (SPEA) were found in a group of 65 KD patients and also 120

control subjects [134]. Thus, investigators believe that KD may not result from a single

pathogen, but rather a group of microbes which produce SAg, strongly activate the immune

system, and cause disease in susceptible hosts [135].

1.3 Animal Models of Kawasaki Disease

Investigators have taken advantage of animal models to investigate the

pathogenesis of KD. Numerous animal models currently exist, which include a rabbit [136, 137],

canine [138], and swine model [139], as well as three different mouse models in which yeast

extract from candida albicans [140, 141], peptidoglycan from group A streptococcus [142], or

lactobacillus casei cell wall extract (LCWE) [143, 144] is injected to induce disease. All the

animal models aim to model the characteristic arteritis of KD, although the swine and rabbit

models fail to display specific inflammation of the coronary arteries.

In the Candida–induced model of KD, young mice are injected i.p. with the water

soluble fraction of Candida albicans extract (CAWS) [140]. Inflammation can be seen within

hours after the first injection and, after approximately 8 weeks, the mice develop proliferative

and granulomatous inflammation of the heart with macrophages, lymphocytes, plasma cells and

neutrophil infiltration that closely mimics KD [145]. Characteristic vessel damage also occurs

with intimal thickening and destruction of the internal elastic lamina and media. Although this

model is useful for studying KD, repeated daily injections of CAWS are required to sustain the

inflammation, as well the arteritis is granulomatous and more widespread affecting the

abdominal aorta as well.

13

Few studies have been published on the streptococcal-induced model of KD. In

this model, sonicated peptidoglycan fragments from Streptococcus pyogenes are injected

intravenously every week for 4 weeks [142]. Mice develop diffuse cellular infiltration in the

vascular wall and display marked hyperplasia of the endothelial cells, necrosis of the coronary

artery medial smooth muscle, and degradation of elastic fibers in the artery.

1.3.1 The Lactobacillus casei cell wall extract (LCWE) model of Kawasaki Disease

The first paper published on the LCWE-induced model was in 1985 by Lehman et

al. [143]. After a single i.p injection of the extract into mice, the group reported a specific

inflammation localized specifically to the coronary arteries. The arteritis displayed disruption of

the arterial intima and media. Various inbred mice strains such as C57BL/6, A/J, Balb/c and

C3H/FeJ were tested and found to be susceptible to the disease.

Further characterization of the LCWE-model demonstrated that it closely

mimicked KD with respect to susceptibility of young mice, time course, histopathology, and

response to IVIG therapy [146] (Figure 2). The disease can only be induced in young mice from

4 to 5 weeks of age. Proinflammatory cytokines induced with LCWE match those implicated in

human KD: TNFα [147], IFNγ [148], IL-1 and IL-6 [149]. Localized inflammation of the

coronary arteries can be seen at day 3 and persists until day 60 [146]. Maximum inflammation

occurs at day 28 and elastin breakdown and aneurysm formation (detected by eosin von gieson

(EVG) staining) occurs at day 42. These disease kinetics approximate to those of human KD as

aneurysms typically occur at weeks 4 to 8. Histopathologically, the inflammatory cell types

match those found in human tissues; with specific staining of F4/80+ macrophages, activated

MIDC-8+ myeloid DCs, plasmacytoid DCs, and specific colocalization of CD3+ T cells with

myeloid DCs detected within coronary artery lesions [150]. Interestingly, however, the group did

not detect B cells in the lesions and B cell null mice did not develop disease. These results

14

contrast from other histopathological findings that have demonstrated that B cells infiltrate the

coronary artery of children with KD [41]. Lastly, LCWE-treated mice respond to IVIG treatment

[146], which lends additional support to the use of IVIG as first line treatment in children with

KD [2].

Figure 2: Overview of the LCWE model of Kawasaki Disease

Lactobacillus casei cell wall extract (LCWE) induces coronary arteritis in mice and mimics human KD with respect to its histopathology, time course, susceptibility in young-aged mice, and response to IVIG therapy. Disease is induced in young inbred mice (4 to 5 weeks of age) with 1mg of LCWE injected i.p. at day 0. Systemic immune activation results with the production of TNFα, IFNγ, IL-6, as well as activation of TCR Vß 2, 4, and 6 family T cells.

15

1.3.2 Superantigens in the LCWE model

Our laboratory has discovered a novel SAg in LCWE [151], and has demonstrated

that SAg activity correlates well to its ability to induce disease in vivo. Stimulation with LCWE

replicates many of the hallmarks of a SAg response: marked proliferation of naive T cells, non-

classical MHC restriction with a hierarchy in the efficiency of different class II molecules to

present SAg, a requirement for antigen presentation but not processing, and polyclonal expansion

o f T cells in a TCR Vβ ch ain-dependent fashion. It was found that TCR Vß 2, 4, 6 and 14

specific T cells were activated.

1.3.3 Tumor Necrosis Factor alpha in the LCWE model

Stimulation with SAg is known to elicit the release of TNFα with biphasic

kinetics. Activated T lymphocytes produce the ‘first wave’ of TNFα, which then drives the

‘second wave’ of TNFα production from monocytes and macrophages within 24 hours [152].

Since TNFα plays a large role in patients during the acute phase of KD [153], our lab has also

characterized its role in the LCWE model [147]. TNFα was rapidly produced in the spleen 6

hours post-LCWE injection. Peak levels of TNFα mRNA were detected in murine hearts at days

3 and 28, correlating with the immune cell infiltrate on day 3 and maximal inflammation (day

28) of the heart observed by histological examination. Most importantly, TNFα activity was

required for disease induction. Mice given etanercept (soluble TNF receptor fusion protein) to

neutralize TNFα did not develop inflammation at any point. Additionally, TNF receptor I

(TNFRI) KO mice were also completely resistant to disease. These findings are significant

because they support the idea of using TNF-based pharmaceuticals as a therapy for the treatment

of KD. However, further work should be done to investigate the precise mechanisms by which

TNFα drives disease in this model.

Given the multitude of effects TNFα regulates, there were many potential

mechanisms by which TNFα could mediate disease in the LCWE model. The absence of

infiltrate in the coronary vessels of TNFRI KO mice, suggested that one function of TNFα might

16

be to affect local migratory signals. In agreement with this hypothesis, mRNA expression of

MIP-1α, RANTES, ICAM-1, VCAM-1, and E-selectin was significantly lower in mice lacking

TNFRI.

1.4 Tumor Necrosis Factor

1.4.1 Brief Overview

The history of TNFα begins from the observations of a German physician, P.

Bruns, who noted the regression of tumours in humans after bacterial infection [154]. It was

hypothesized then that bacterial extracts could thus be used to treat human cancers. Later work

demonstrated that LPS purified from bacterial extracts was responsible for tumour regression

[155], and O’Malley et al. eventually demonstrated that LPS did not act directly on tumours, but

through the production of a serum factor; which they called tumour-necrotizing factor [156].

Later in 1984, the cDNA of TNFα was cloned and sequenced [157], which led to intense

research of this pleiotropic cytokine over the next couple decades.

TNFα is a homotrimeric protein of 157 amino acids which is primarily produced

by macrophages, but is also produced by natural killer cells, T cells, B cells, monocytes, and

dendritic cells [158]. TNFα signals through two distinct cell surface receptors, TNFR1 (55kDa)

and TNFR2 (75kDa) which are expressed on a variety of different tissues. TNFR1 has been

shown to be widely expressed by all tissues tested so far [158] including: the brain [159, 160],

heart [161], lungs [162], kidney [163], pancreas [164], immune cells [165], reproductive organs

[166], and the gut [167]. This is in contrast to TNFR2, which is mainly expressed by immune

cells and endothelial cells [158]. Ubiquitous expression of TNFRs has been suggested as the

reason for the diversity of functions attributed to TNFα. Studies have suggested that TNFR1

mediates the majority of TNF’s inflammatory functions such as proliferation, differentiation, and

regulation of apoptsis [168], although this has been debated as other groups suggest that the both

receptors cooperate to transduce their signals [169].

17

The functions of TNFα include the regulation of apoptosis, survival,

differentiation, and proliferation [158] (Figure 3). The diversity of roles stem from the different

cell signaling pathway initiated. Binding of TNFα to its receptors leads to trimerization via their

intracellular pre-ligand-binding assembly domain (PLAD) [170]. TNFR1 contains an

intracellular death domain (DD) region which can associate with TNR receptor-associated death

domain (TRADD) [171], a cytosolic adaptor protein that acts to recruit the downstream signaling

adaptor molecule Fas-associated death domain (FADD) [172]. FADD is able to bind and activate

caspase-8, which can then initiate apoptosis.

Both TNFR1 and TNFR2 recruit TNF receptor-associating factors (TRAFs)

which serve as adaptors to drive dramatic changes in cell function [168]. One TRAF family

member, TRAF2, which can bind both TNF receptors, leads to activation of JUN N-terminal

kinase (JNK)/activator protein 1 (AP1), p38 mitogen-activated protein kinase (p38MAPK), and

nuclear factor kappa B (NF-κB) signaling pathways [158] (Figure 3). Activation of these cellular

pathways contribute to TNFα’s function as a stimulator of survival, differentiation, and

proliferation.

TNFα represents a double edged sword; having both positive and negative effects.

Stemming from its history, one positive effect was its ability to destroy tumor cells. Recent

studies demonstrated successful treatment for locally advanced limb soft tissue sarcomas and

other large tumours [173]. However, even with success, TNFα’s anticancer utility is limited by

its vasculotoxic effects which include constitutional symptoms such as fever, chills, headache,

and fatigue [174]. TNFα also has regulatory functions in the immune system. It has been shown

to provide key signals for the morphogenesis and microarchitecture of secondary lymphoid

organs. For example, mice deficient in TNFα fail to develop peyers patches [175] and

TNFα/Lymphotoxin double knockouts lacked follicular dendritic cell networks within mucosal

lymph nodes [176]. In addition, TNFα also contributes to the function of cytotoxic effector cells

in the detection and destruction of virally infected cells [177, 178]. Owing to its production by

almost all haematopoietic cells, TNF has also been implicated in the proliferation of lymphocytes

[179, 180]. In addition, genetic deletion studies have also shown its importance in protection

18

against microbial infections from Listeria monocytogenes and resistance to LPS-mediated septic

shock [181]. However, despite these positive aspects of TNFα function, dysregulation of TNFα

can lead to harmful effects such as autoimmunity (see 1.4.2), and other disease such as chronic

heart failure [182, 183] and bone destruction [184].

19

Figure 3: TNFR1 and TNFR2 drive apoptosis, survival, differentiation, and proliferation through different cell signaling pathways. Adapted from [158]. TNFα activates its receptors, TNFR1 and TNFR2, leading to the recruitment of cytosolic adaptor proteins TNF receptor-associated death domain (TRADD) and TNF receptor-associating factor 2 (TRAF2). These adaptors link to multiple signaling pathways controlling different cell functions. Signaling from TRADD leads to the recruitment and activation of Fas-associated death domain (FADD), which can lead to the initiation of apoptosis via caspases 8 and 3. TNF signaling via TRAF 2 results in the activation of activator protein 1 (AP1), p38 MAP kinase, and NF-κB signaling pathways.

20

1.4.2 TNFα and cell survival

TNFα has a prominent role in the regulation of apoptosis (reviewed in [168]).

Reports have shown both pro-apoptotic and anti-apoptotic functions for TNFα. Signaling by

TNFR1 is thought to occur by the formation of two complexes [185]. When TNF binds TNFR1,

the adaptor protein TNF receptor-associated death domain (TRADD), receptor interacting

protein 1 (RIP1), and TNF receptor associated factor 2 (TRAF2) are recruited to form complex I.

Complex I binds the I kappa B kinase (IKK) complex, and activates NFκB, resulting in the

production of anti-apoptotic factors such as FLIPL. The second complex (complex II) is found in

the cytoplasm and forms when TRADD and RIP1 associate with FADD and caspase-8 to drive

apoptosis. In cases where complex I activates NFκB, FLIPL is able to bind complex II and inhibit

caspase-8 to promote cell survival. In instances where the initial signal from complex I fails, cell

death via complex II is thought to occur. In addition to this pathway, signals mediated through

TRAF2 can also promote programmed cell death. TRAF2 can stimulate the MEKK pathway,

resulting in the activation of JNK signaling. Induction of JNK is strongly linked with the

induction of apoptosis, although the precise mechanism of action is unclear [186, 187].

TNFR1 also activates additional survival pathways through the activation of other

NF-κB –induced factors. For example, cellular inhibitor of apoptosis protein 2 (c-IAP2) is under

the transcriptional control of NF-κB and can bind to TRAF2 to block caspase-8 activation and

apoptosis [168]. A20 is another NF-κB -induced protein which can bind TRAF2 to block

apoptosis [188]. Finally, another induced protein named gadd45b strongly inhibits apoptosis via

downregulating the JNK cascade [189]. It was shown to bind directly to MKK7 to block its

activation of JNK [190]. Thus, there are multiple pathways of survival and apoptosis which stem

from TNFα signaling. However, how one pathway is chosen over another is not currently clear.

21

Figure 4: TNFα drives apoptotic and survival signaling pathways. Adapted from [158, 168, 185, 189]. Binding of TNF to TNFR1 is thought to result in two sequential signaling complexes which mediate apoptosis and survival. After TNF binds TNFR1, the adaptor protein TNF receptor-associated death domain (TRADD), receptor interacting protein 1 (RIP1), and TNF receptor associated factor 2 (TRAF2) are recruited to form complex I. Complex I activates the NF-κB signaling pathway via recruitment of the I kappa B kinase (IKK) complex which results in the production of NF-κB –induced factors such as FLIPL. The second signaling complex mediates apoptosis and is formed when TRADD and RIP1 associate with FADD and caspase-8 in the cytoplasm (complex II). When complex I activates NF-κB, FLIPL associates with complex II and inhibits caspase-8 to promote cell survival. In cases where complex I fails to activate NF-κB, cell death occurs. TNFα signaling can also result in survival with other NF-κB -induced factors such as A20 and cellular inhibitor of apoptosis protein 2 (c-IAP2). Both proteins can bind to TRAF2 to block caspase 8 activation. Another induced protein named gadd45b is also induced by NF-κB and inhibits apoptosis by downregulating the JNK signaling cascade.

22

1.4.3 TNFα during superantigen responses and autoimmunity

In KD, the immune response is thought to be initiated by a SAg-producing

infectious agent, which activates T cells and drives their expansion and subsequent infiltration

into the heart. However, the persistence of these T cells in the heart does not agree with the

established fate of SAg-activated T cells, as the SAg-reactive cells typically undergo apoptosis or

are rendered anergic (see 1.2.2). Thus, a paradox arises as it becomes unclear how autoimmunity

can result from SAg responses.

A few studies have shown that the fate of T cells after SAg activation can be

altered with concurrent stimulation. For example, it has been shown that dually reactive T cells

from mice immunized with both SAg and conventional peptide-antigen are able to escape

apoptosis [191]. Our lab has shown that this ‘rescue’ from apoptosis is mediated by enhanced

costimulation [192] as murine splenocytes cultured with SEB and agonistic anti-CD28 antibodies

displayed a reduction in annexin V+ staining cells within SEB-reactive Vβ8+ T cell population.

In addition to this, interesting work by Vella et al [193, 194], demonstrated that LPS could also

interefere with the induction of T cell death in response to the SAg SEA. In particular, it was

found that TNFα was critical for this ‘rescue,’ as neutralizing TNFα antibodies prevented the

survival of SEA-reactive T cells with LPS. In view of these results, it is demonstrated that T cell

fate following SAg stimulation is mutable, and that the apoptosis of peripheral T cells is not yet

completely understood.

Because TNFα is implicated as a pathogenic cytokine in KD (see 1.3.3), we

sought to understand how TNFα might modulate T cell fate after SAg stimulation. In this work,

we aim to demonstrate that TNFα is important in the regulation of costimulatory molecule

expression during SEB responses, effectively controlling the apoptosis of SEB-responsive T

cells.

23

1.5 T cell Activation and Costimulation

1.5.1 T cell receptor signaling with conventional antigens

During T cell activation with conventional peptides, naive T cells are stimulated

by their T cell receptor (TCR), which interacts with peptide presented on MHC molecules by

antigen presenting cells (APC). The T cell receptor is composed of the functional receptor

complex (made up of α and β chains) and the additional CD3 signaling complex composed of

four signaling chains (two ε, one δ, and one γ) (Figure 5). The ζ chains of CD3 are particularly

important because they contain immunoreceptor tyrosine-based activation motifs (ITAM[s])

which are phosphorylated to recruit proteins containing src homology (SH) domains.

Figure 5: Adapted from Immunobiology 7th Ed. by Janeway C. (Figure 6.10) The T cell receptor complex is made up of antigen-recognition proteins and invariant signaling proteins. The T cell receptor is made of α and β chains with variable regions (v) which allow for antigen recognition, and constant (c) regions. The CD3 complex ( ε , δ , and ζ chains) also associates with the TCR and helps mediate cell signaling via their ITAMs.

24

Interaction between the MHC:peptide complex and the TCR specific for the

peptide activates the T cell and induces multiple cell signaling cascades. After ligation of MHC

and TCR, the CD4 co-receptor present on the surface of T cells binds to MHC class II. The

intracellular domain of CD4 associates with a src-family kinase called Lck, which, now localized

to the TCR, can phosphorylate the ITAMs present on CD3’s ζ chains (Figure 6). Phosphorylated

ITAMs are able to recruit the syk-family kinase ZAP-70 via its SH2 domain, which then recruits

and phosphorylates linker of activation in T cells (LAT) and another protein named SLP-76

(Figure 6). LAT serves as an adaptor protein to recruit Tec kinases, such as Itk, which is then

activated by SLP-76, and initiates the phospholipase Cγ (PLC-γ) and guanine exchange factor

(GEF) pathways.

25

Figure 6: adapted from Immunobiology 7th Ed. by Janeway C. (Figure 6.18) Different intracellular pathways result from TCR activation. Interaction between the MHC:peptide complex and TCR results in phosphoylation of the ζ-chain ITAMs by CD4-associated Lck. ZAP70 then binds these ITAM and is activated by Lck. ZAP70 phosphorylates LAT and SLP-76, which recruit and stimulate Tec kinases, which activate PLC-γ and SOS pathways. PLC-γ converts PIP2 to IP3 and DAG, which leads to NFAT, NF-κB, and RasGRP signaling. SOS activates a MAPK signaling cascade involving Ras, Raf, Mek, Erk, and Elk. NFAT, NF-κB, and AP-1 transcription factors transcribe specific genes involved in cell proliferation and differentiation.

26

The PLC-γ pathway involves the cleavage of phosphatidylinositol bisphosphate

(PIP2) to yield diacylglycerol (DAG) and inositol trisphosphate (IP3) (Figure 6). Production of

these second messengers results in the release of Ca2+ from the endoplasmic reticulum (via IP3

receptors) and activation of protein kinase C (PKC) and calcineurin which activate transcription

factors NF-κB and NFAT respectively. Translocation of NF-κB and NFAT into the nucleus

drives specific transcription of genes involved in T cell activation.

The tec kinases associated with LAT also activate mitogen-activated protein

kinase (MAPK) cascades by activating the GEFs called SOS and Vav (Figure 6). Activation of

SOS leads to the activation of Ras, a small G-protein, which activates the MAPK cascade [Raf

(MAPKKK), MEK 1/2 (MAPKK), ERK 1/2 (MAPK)] which results in the phosphorylation and

activation of transcription factor Elk, which translocates to the nucleus to transcribe genes

involved in T cell activation. A similar pathway is initiated by Vav and will be expanded upon in

1.5.2 because it is associated with costimulatory CD28 signaling.

1.5.2 Costimulation with CD28, B7 ligands, CTLA4, and Cbl-b

During productive T cell responses, naive T cells must also receive a “second

signal” in addition to peptide stimulation of the TCR. Ligation of the TCR alone does not induce

proliferation and differentiation into effector T cells; instead, the T cells are rendered

unresponsive or anergic. Thus, for proper T cell expansion and function, a second costimulatory

signal, must be provided by the APC.

The most well studied costimulatory molecules are the glycoproteins B7.1 and

B7.2. They are homodimeric members of the immunoglobulin superfamily that are found

exclusively on the surfaces of cells that stimulate T cell proliferation. The differences between

B7 ligands appear to be slight. Although B7.2 is constitutively expressed at lower levels on APC

than B7.1, it is induced quickly after activation and is thought to mediate earlier events in T cell

27

activation. After exposure to an inflammatory stimulus (such as a foreign antigen), APC undergo

a maturation process whereby they upregulate B7.1 and B7.2 on their cell surface. In this way,

matured APC presenting a specific peptide also deliver a second costimulatory signal to antigen

specific T cells. B7 molecules on the APC interact with the CD28 receptor found on T cells.

Interaction between B7 and CD28 is required for T cell clonal expansion and has been shown to

dramatically lower the threshold of T cell activation.

CD28 is a glycoprotein with a single ligand-binding domain expressed on almost

all murine T cells, with CD4 T cells expressing higher levels [195]. Numerous studies have

shown that CD28 promotes T cell proliferation, cytokine production (via gene transcription and

mRNA stability), and cell survival [196]. Binding of B7 to CD28 results in the recruitment and

activation of phosphoinositide 3-kinase (PI3K) [197], which catalyzes the conversion of PIP2 to

phosphatidylinositol trisphosphate (PIP3) (Figure 7). Localized to the cell membrane, PIP3

serves as a docking site for 3-phosphoinositide-dependent protein kinase 1 (PDK1), which can

then activate its target kinase called Akt. Multiple signaling events are initiated downstream of

Akt’s kinase activity. Activation of Akt enhances nuclear translocation of NF-κB, which

increases the expression of prosurvival genes such as Bcl-xL. In addition, Akt inhibits

transcription factors that promote cell cycle arrest [198]. Akt accomplishes this by inactivating

glycogen-synthase kinase 3 (GSK3), which promotes nuclear export of NFAT [199], allowing

for optimal transcription of NFAT-genes such as IL-2.

28

Figure 7: adapted from Immunobiology 7th Ed. by Janeway C. (Figure 6.18) CD28 Signaling Enhances TCR Signaling via NF-κB, NFAT, and AP-1 Transcription factors. Elements from TCR signaling are labelled in grey, unless they are in common with CD28 signaling. Binding of B7 ligands to CD28 results in the recruitment of PI3K which activates PDK-1, then Akt1. Akt1is a kinase which phosphorylates many cytosolic proteins resulting in the increased NF-κB activity. One such protein is GSK3, which acts to increase NFAT activity by preventing its export. The activated CD28 receptor can also bind adaptor protein Grb2, which recruits SLP-76. Similar to TCR signaling, SLP-76 can also bind another GEF called Vav, which initiates another distinct MAPK cascade (Rac, Mekk, Jnkk, Jnk), resulting in enhanced AP-1 signaling. Costimulation is regulated by CTLA-4 and Cbl-b. CTLA-4 competes with CD28 for binding to B7 ligands. Cbl-b is a E3 ubiquitin ligase which negatively regulates PI3K and Vav.

29

In addition to recruiting PI3K, activated CD28 can also bind the adaptor protein

growth factor receptor-bound protein 2 (Grb2), which is able to recruit SLP-76 to activate

another GEF called Vav. Activation of Vav leads to the activation of Rac, a small g-protein,

which activates the another MAPK cascade [Mekk (MAPKKK), Jnkk (MAPKK), Jnk (MAPK)]

and results in the phosphorylation and activation of transcription factor Jun. Thus CD28

signaling can been seen to synergize with common elements from the TCR signaling pathway

(NFAT, NF-κB, AP-1) (Figure 7).

Other costimulatory proteins act to taper T cell activation. Cytotoxic T-

Lymphocyte Antigen 4 (CTLA4) is another costimulatory receptor which opposes the activities

of CD28 on T cells. CTLA-4 shares sequence homology with CD28 and is able to bind B7

ligands. CTLA-4 antagonizes CD28 signaling because, once induced on activated T cells, it

binds B7 ligands with 20-fold greater avidity; effectively preventing CD28 signaling by

competitive binding (Figure 7). In addition to competing with CD28 for B7, other studies have

suggested that CTLA-4 might interact with Lck, Fyn and ZAP-70 through SHP1 phosphatase to

prevent constitutive expression of downstream signaling pathways [200]. Mice lacking CTLA4

develop a strong lymphoproliferative phenotype and die prematurely [201]. Inhibition of T cell

activation by CLTA-4 has been suggested as a viable therapy to treat hyperactive immune

disorders [202, 203]. After its cloning in 1987 [204], a recombinant protein of CTLA-4 fused to

a human heavy Ig chain was developed [203], and was found to greatly suppress T cell

responses. In 2005, Abatacept (humanized CTLA4-Ig fusion protein) was approved for the

treatment for severe rheumatoid arthritis that cannot be controlled by traditional anti-rheumatic

drugs such as methotrexate.

Another mechanism by which T cell activation can be suppressed is by the

activity of the regulatory protein Cbl-b (Figure 7). Named after Casitas B-lineage Lymphoma,

Cbl-b is an E3 ubiquitin-protein ligase which acts seems to regulate costimulatory signaling in T

cells [205]. By ubiquitinating target proteins such as PI3K and Vav, Cbl-b negatively regulates T

cell activation. Notably, Cbl-b deficiency results in spontaneous autoimmunity as mice age

[206].

30

1.5.3 T cell activation and costimulation with superantigens

Activation with SAg has been assumed to be similar to TCR signaling with

conventional antigens [207]. Many studies have shown that particular features of conventional T

cell activation remain intact after stimulation with SAg. These features include similarities in

NFAT activity and calcium mobilization [208], formation of mature immunological synapses

[209], as well as expression of typical activation markers CD25 and CD69 [208]. Despite these

similarities, SAg stimulation seems to override the requirement for CD4 and Lck signaling. It is

known that CD4 binds MHC class II to stabilize interactions with the TCR and to bring its

associated Lck in proximity to the ζ chains. However, it has been shown that responses to some

SAg (SEB and SEC) are not affected by blockade of CD4 with monoclonal antibodies [210], and

T cell lines deficient in CD4 respond similarly to SAg [211]. Interestingly, Lck seems to be

dispensable for T cell activation with SAg [208], indicating the possibility of additional signaling

pathways emanating from the TCR used by bacterial toxins.

A recent study has elucidated an Lck-independent pathway of T cell activation

with SAg that converges with conventional signaling at the level of ERK1/2 activation [212]. In

the canonical pathway, PLC-γ is responsible for initiating the MAPK cascade leading to ERK1/2

activation. Yet, after stimulation with SAg, PLC-γ was not activated in the absence of Lck. Thus,

the investigators hypothesized that another PLC was involved to allow for ERK signaling. Their

study suggests that SAg bypass Lck-dependent TCR signaling by activating PLC-β via TCR-

mediated activation of a Gα11 (a G protein). The activity of PLC-β leads to IP3 and DAG

production, which result in activation of the MAPK cascade involving Ras, Raf, Mekk, and Erk.

Signaling differences between conventional antigens and SAg are interesting because they may

begin to explain how SAg-stimulation leads to apoptosis and anergy of T cells.

31

1.5.4 Costimulation and anti-apoptotic proteins

Signaling from CD28 results in enhanced expression of prosurvival genes such as

Bcl-XL [213]. Bcl-XL is a member of the Bcl-2 family of proteins which include both pro-

apoptotic (Bax and Bad) and anti-apoptotic (Bcl-XL and Bcl-2) members [214]. These proteins

are integral membrane proteins located in the outer membrane of the mitochondrion [215], which

regulate apoptosis by controlling the integrity of the membrane. Anti-apoptotic molecules

prevent the influx of ions through the mitochondrial membrane, thereby inhibiting the release of

cytochrome c, which triggers apoptosis if released into the cytosol. In contrast, pro-apoptotic

molecules such as Bax have opposite functions. Elevated levels of Bax promote the influx of

ions into the mitochondria and the release of cytochrome c while Bad initiates apoptosis by

inhibiting the activity of Bcl-2/Bcl-XL in the membrane by binding them. Costimulation provided

by CD28 also results in the activation of Akt kinase, which can also prevent the activities of pro-

apoptotic molecules. Akt phosphorylates Bad, which in phosphorylated form, cannot bind to Bcl-

2/Bcl-XL in the membrane and promote apoptosis. Thus, costimulation promotes the survival of T

cells by regulating the expression of Bcl-2 family proteins.

In addition to regulation of apoptotic molecules, CD28 signaling has also been

shown to induce high levels of IL-2 transcription [216] and stability [217]. Apoptosis is known

to be prevented by the addition of IL-2 [218, 219] and thought to play a major role in the survival

of activated T cells [220]. Therefore, costimulation promotes the viability of activated T cells

through multiple pathways, including increased production of IL-2 and enhancement of Bcl-XL

expression. Interestingly, these two mechanisms operate independently of each other [221], as it

was shown that CD28 mediates intrinsic survival via Bcl-XL and that IL-2 provides an extrinsic

survival signal to activated T cells [213].

32

2 RATIONALE Kawasaki disease (KD) is an acute, multi-systemic inflammatory disease

characterized by persistent inflammation in the coronary arteries. Acute KD is characterized by

the release of proinflammatory cytokines such as tumor necrosis factor alpha (TNFα) and is

thought to be initiated by a superantigen (SAg). Lactobacillus casei cell wall extract (LCWE)

induces coronary arteritis in mice and closely models human KD. Studies using this model have

shown that TNFα and its receptor are critically required for pathogenesis- although the precise

action of TNFα is/are unknown. A persistent T cell infiltrate in the coronary artery is not in

keeping with established cell fates for SAg stimulated cells, which typically undergo apoptosis or

are rendered anergic. TNF receptors are ubiquitously expressed in mouse tissues and are known

to regulate the apoptosis of T cells. Costimulation with CD28 and B7 ligands induce signaling

pathways which also prevent the apoptosis of activated T cells.

3 HYPOTHESIS TNFα plays a key role in the survival of SAg-activated T cells.

33

4 EXPERIMENTAL AIMS To evaluate the role of TNFα in KD, we aimed to investigate the direct and

indirect effects of TNFα on the apoptosis of SAg-activated T cells in vitro. Previous work had

shown that enhanced costimulation could rescue SAg-reactive T cells from apoptosis [192]; thus

we developed 4 specific aims:

1. Determine whether TNFα contributes to the survival of SAg-reactive T cells indirectly by

regulating costimulation

2. Determine whether TNFα regulates activated T cell survival directly via its own TNF

receptors

3. Identify the precise mechanism(s) and cellular players that regulate TNFα-driven survival

of SAg-reactive T cells

4. Characterize TNFα’s effects on T cell activity and the expression of survival molecules.

34

5 MATERIALS AND METHODS

5.1 Experimental Mice

Wild-type C57BL/6, TNFR1-/-R2-/-, and CD86 -/- were purchased from The

Jackson Laboratory (Bay Harbor, ME) and housed under specific pathogen-free conditions at the

Center for Phenogenomics and the Hospital for Sick Children in Toronto under institutional

guidelines. Cbl-b -/- spleens were a generous gift from Dr. Pamela Ohashi (University of

Toronto). All animal procedures were approved by the animal care committee at The Hospital

for Sick Children Research Institute.

5.2 Cell Culture Splenocytes from mice between 6-8 weeks of age were homogenized and treated

with Gey’s solution to lyse red blood cells. Cells were cultured in 24-, 12-, and 6-well plates at

2x106 cells/ml in Iscoves minimum essential medium supplemented with 10% heat-inactivated

fetal bovine serum, sodium pyruvate, non-essential amino acids, 50μM 2-ME, 2mM L-

glutamine, and 10mM HEPES. Cells were cultured with various combinations of stimulatory and

inhibitory agents including: the prototypical superantigen Staphylococcal Enterotoxin B (SEB)

(Toxin Technology (Specific doses are labelled figures), CD28 stimulatory antibody (37.51,

Biolegend) (0.5 μg/ml), CTLA4-Ig (Bristol-Myers Squibb) (50 μg/ml), TNFα neutralizing

antibody (MP6-XT22, eBioscience and BD Biosciences) (0.1 μg/ml), and recombinant mouse

TNFα (eBioscience) (see doses in figure). During incubations, cells were placed in a humidified

chamber containing 5% CO2 at 37°C.

5.3 Purification and Culture of T cells and Antigen Presenting Cells

T cells and antigen presenting cells (APC) were purified by magnetic bead

separation using the pan T isolation kit from Miltenyi Biotec as per the manufacturer’s

instructions. Specifically, the kit contains biotin-conjugated antibodies against CD11b, CD11c,

35

CD19, CD45R (B220), CD49b (DX5), CD105, Anti-MHC-class II, and Ter-119 and the

isolation of T cells is achieved by depletion of magnetically labeled cells. Isolation with this kit

does not disturb or activate T cells. Purity of the T cells and APC was assessed by flow

cytometry using fluorescent anti-CD3 antibodies or isotype control staining of isolated cells. In

all experiments, the purity of T cells was >90%. In mixing experiments involving co-culture of

APC and T cells, the APC fraction was further enriched using a second immunomagnetic

column. The purity of APC after this enrichment was >90%. For these mixing experiments,

purified T cells were added to APC at a ratio of 1:4 and were cultured at a total concentration of

2x106 cells/ml in 24-well plates.

5.4 Antibodies and Flow Cytometry

Cultured single cell suspensions were stained with fluorochrome-conjugated

antibodies and secondary reagents, and the immunofluorescence analyzed on a BD Bioscience

flow cytometers (initial experiments on BD FACScan and BD FACSCalibur, later experiments

on BD LSR II and BD Canto II) using standard labelling techniques. Briefly, 1x106 cells were

aliquoted into 4 ml round bottom tubes (Falcon 2052) and were washed with 1ml of staining

media (SM) (2% FBS, 25mM EDTA in PBS, pH 7.2). Cells were stained in 100ul of SM

containing optimal concentrations of fluorescent antibodies for 20-30 minutes at 4°C. Flow

cytometry analysis was performed on live cells by staining with 7AAD, propidium iodide, or

fixable viability dye eFluor® 450 (eBioscience). Gating of populations were determined using

isotype controls and fluorescence minus one (FMO) controls. Cells were stained extracellularly

with CD19-PE (1D3), CD11c-PE (HL3), F4/80-APC (BM8), CD80-FITC (16-10A1), CD86-

FITC and –PECy7 (GL-1), CD3-FITC (145-2C11), and V-beta 8 TCR-FITC (F23.1) antibodies.

Cells were stained intracellularly with TNF-APC (MP6-XT22), Bcl-2-PE (3F11), Bcl-XL (54H6)

antibodies. Biotinylated goat anti-rabbit antibodies were used as secondary antibodies for Bcl-XL

staining and were detected by strepavidin conjugates. Purified antibodies were purchased from

BD Biosciences (San Diego, CA), eBiosciences (San Diego, CA), Cedarlane (Hornby, ON,

Canada), and Cell Signaling Technology (Danvers, MA). Apoptosis was measured with the

Annexin V-PE Apoptosis Detection kit (BD biosciences) following the manufacturer’s protocol.

Flow cytometry data was analyzed using FlowJo (Tree Star, Inc., Ashland, OR).

36

5.5 Measurement of Proliferation

5.5.1 3H-Thymidine Incorporation

Murine splenocytes were cultured in 96-well plates at 5x105 cells/ 200ul of

complete media containing different stimuli. A minimum of 3 replicate wells were plated and

cultured for various durations (see figure legends). Before scintillation counting, cells were and

pulsed with 1µCi/well of [3H]-thymidine (GE Healthcare Bio-Sciences Inc, Baie d’Urfe, QC) for

the last 18-24 hours of incubation. Quantification was performed on a Beckman Coulter LS6500

scintillation counter. Disintegrations per minute (dpm) measurements from quadruplicate wells

were averaged and a stimulation index was calculated by dividing dpm counts from stimulated

wells by unstimulated controls.

5.5.2 BrdU Incorporation

For measurement of specific SAg-reactive T cell proliferation a BrdU

incorporation assay (BrdU-FITC Kit, BD Biosciences) was performed. Splenocytes were

cultured in 12-well plates at 2x106 cells/ml with complete media containing various stimuli.

BrdU was added during the last 18-24 hours of incubation, and the cells were stained for T cell

markers and BrdU as per the manufacturer’s instructions.

5.6 Statistical Analysis and Graphics Statistical significance was determined by either unpaired or paired two-tailed

student’s t-tests where appropriate using GraphPad Prism 4 (GraphPad Software Inc, La Jolla,

CA). A p-value of P<0.05 was considered significant. In certain experiments, values were

reported as the mean + standard error of measurement.

37

6 RESULTS

6.1 SEB upregulates B7.2 ligand on B cells but not B7.1

Superantigenic stimulation of T cells results in their activation and expansion,

followed by their deletion. Since SAgs can bind to MHC class II directly and do not require the

help of professional antigen presenting cells (APC), SAgs are thought to activate T cells without

costimulation [193]. Previous studies have investigated the role of costimulatory molecules

CD28 and B7 during SAg responses, but they have mainly focused on the initial activation and

proliferation of T cells [222, 223]. Furthermore, the role of endogenous B7 ligands affecting

survival in vitro and in vivo is still controversial [193, 194, 224-226].

To demonstrate whether B7 molecules have an endogenous role, we first

characterized the costimulatory ligands B7.1 and B7.2 in response to the SAg SEB. Splenocytes

from wildtype mice were cultured with SEB for 2-4 days and the expression of B7.1 and B7.2

was determined by flow cytometry. The expression of B7.2 was markedly upregulated after

stimulation with SEB at days 2 and 3, then returned back to control levels at day 4 (Figure 8A).

No upregulation of B7.1 was seen between days 2 – 4. On day 2, the mean increase in MFI after

SEB stimulation was plotted for B7.1 (0.366 +2.118, p=0.8785) and B7.2 (286.5 +39.03,

p=0.0101) (Figure 8C, left and middle). These findings support other data which demonstrate a

larger role for B7.2 ligand than B7.1 in response to SEB [226]. Further characterization of the

cells expressing B7.2 was performed. After 48hr culture with SEB, B cells expressed high levels

of B7.2 (Figure 8B), while T cells also upregulated B7.2 but at lower levels. The mean increase

in MFI after SEB stimulation was plotted for B cells (3489.0 + 915.5, p=0.0189) (Figure 8C,

right).

38

Figure 8: Stimulation with SEB increases B7.2 expression on B cells

(A) Splenocytes were cultured in media in the presence or absence of 0.3μg/ml SEB for 2-4 days. Staining was performed with antibodies specific for B7.1 and B7.2. Samples were analyzed by flow cytometry using a gate for viable cells. (B) Identical experiment as (A), except cells were analyzed on day 2 only and stained with antibodies to CD19 (B cells) and CD3 (T cells). Isotype control staining was performed on SEB stimulated cells and is shown in grey. Each histogram presented in this figure are representative of a minimum of n=3 independent experiments. (C) Changes in MFI on day 2 were plotted for the data shown in (A) (C left and middle) and (B) (C right). Statistical significance was determined by the student’s paired t-test.

39

6.2 TNFα and its receptors regulate B7.2 expression on B cells Since superantigens can stimulate the production of TNFα [107], we sought to

determine whether TNFα might contribute to the expression of costimulatory molecules directly.

Splenocytes were cultured in the presence or absence of recombinant TNFα and the expression

of B7.1 and B7.2 was determined by flow cytometry. Analysis was performed at day 2, because

previous experiments with SEB demonstrated sustained upregulation of B7.2 on days 2 and 3

(Figure 8). Stimulation with recombinant TNFα resulted in increased levels of B7.2 (Figure 9A)

but not B7.1 (data not shown). Characterization of the immune populations with specific staining

for CD19 and CD3, revealed that both B cells and T cells were able to upregulate B7.2

expression after TNFα stimulation (Figure 9A). Changes in MFI after stimulation with TNFα

were plotted for B cells (5055 + 450.3, p<0.0001) and T cells (2318 + 129.7, p<0.0001) (Figure

9B). This upregulation was also found to be dose dependent, as increasing concentrations of

recombinant TNFα (2ng/ml, 12ng/ml, 62ng/ml) induced higher expression of B7.2 (Figure 9C).

This effect was also mostly absent in TNFR1-/-R2-/- cultures although slight upregulation could

be seen.

Although TNFα was found to directly upregulate B7.2 expression on B cells, it

remained unclear whether SEB directly upregulated B7.2, or whether SEB indirectly upregulated

B7.2 through production of TNFα. To address this question, splenocytes from wildtype or

TNFR1-/-R2-/- mice were stimulated with SEB and the expression of B7.2 was determined.

Wildtype B cells expressed high levels of B7.2, while TNFR1-/-R2-/- cells did not display any

upregulation (Figure 10A). In addition, TNFR1-/-R2-/- cells had lower expression of B7.2 than

wildtype cells (Figure 10B) after stimulation with SEB. These findings support the idea that

endogenously produced TNFα regulates costimulatory molecules on B cells.

40

Figure 9: TNFα upregulates B7.2 expression on B cells in a dose dependent manner

(A) Splenocytes from wildtype mice were stimulated in the presence or absence of recombinant TNFα (62ng/ml). Specific staining for B7.1 and B7.2 was performed on day 2 and analyzed by flow cytometry with a gate on viable cells. (B) Changes in MFI were plotted for the data shown in (A). Statistical significance was determined by the student’s paired t-test. (C) Different doses of recombinant TNFα were added to wildtype and TNFR1-/-R2-/- splenocyte cultures, and the expression of B7.2 on B cells was determined on day 2. Fluorescence minus one control staining was performed on SEB stimulated cells and is shown in grey. Each presented histogram in this figure is representative of a minimum of n=3 independent experiments.

41

Figure 10: TNF receptors on B cells regulate the expression of B7.2 during SEB stimulation

(A) Wildtype and TNFR1-/-R2-/- splenocytes were cultured in the presence or absence of SEB (0.03μg/ml) for 2 days. Staining was performed with antibodies specific to CD19 and B7.2 and the samples were analyzed by flow cytometry using a gate for viable cells. (B) The data in (A) were replotted to show the relative expression of B7.2 between wildtype and TNFR1-/-R2-/- in media or SEB-stimulated cultures. Fluorescence minus one control staining was performed on SEB stimulated cells and is shown in grey. Each histogram is representative of n=2 independent experiments.

42

6.3 Enhanced costimulation rescues SEB-reactive cells from apoptosis

Initial reports demonstrated that B7 ligands are critical for the activation and

proliferation of responsive T cells, however, whether costimulation plays a role in cell survival

has been somewhat controversial. Some reports have shown that disrupting B7 ligands with

CTLA4-Ig antibodies prevents optimal activation and expansion of cells [194, 226] while other

studies using CTLA4-Ig transgenic mouse failed to find such differences [225]. Furthermore,

because it has been reported that only the proliferating SEB-reactive cells undergo death [102], it

remains unclear whether survival, or the lack thereof, was due to a lack of costimulation. It may

be the case that SEB-reactive cells were not effectively expanded in the first place.

To determine whether costimulation affects the survival of SEB-reactive T cells,

SEB-stimulated splenocytes were cultured with or without agonistic anti-CD28 antibody. After

the initial proliferative phase (days 1-3), the apoptosis of Vβ8+ T cells was detected with

Annexin V and 7AAD staining at day 5. The addition of anti-CD28 decreased the total frequency

of apoptotic Vβ8+ T cells from 27.1% to 19.7% in the experiment shown (Figure 11A). In

addition, enhanced costimulation also decreased the percentage of apoptotic Vβ8+ T cells

relative to the total Vβ8+ T cells from 72.6% to 65.6% (Figure 11B). These findings were

significant (p = 0.03) and reproducible (data are shown from 4 independent experiments).

43

Figure 11: Enhanced costimulation rescues SEB-stimulated T cells from apoptosis

(A) Wildtype splenocytes were cultured with SEB (0.3μg/ml) in the presence or absence of costimulatory anti-CD28 antibody (0.5μg/ml) for 5 days. Apoptosis of SEB-reactive T cells was determined by annexin V and Vβ8-TCR staining. Dead cells (7AAD+) were excluded from analysis and gating was determined by an isotype control stain. (B) The percentage of apoptotic Vβ8+ T cells within the total Vβ8+ population was plotted. Data from n=4 independent experiments are shown. Statistical significance was determined by the student’s paired t-test.

44

To further test whether costimulation regulates the survival of T cells after SAg

stimulation, additional costimulatory blockade experiments were performed using CTLA4-Ig.

When CTLA4-Ig was added to SEB-stimulated splenocytes, increased apoptosis was observed in

Vβ8+ T cell population (Figure 12A). In the particular experiment shown, the addition of

CTLA4-Ig increased the total frequency of apoptotic Vβ8+ T cells from 17.3% to 23.7%. In

addition, costimulatory blockade also increased the percentage of apoptotic Vβ8+ T cells relative

to the total Vβ8+ T cells (Figure 12B). These findings were significant (p = 0.03) and

reproducible (data are shown from 4 independent experiments).

Finally, experiments were also performed using Cbl-b -/- mice. These mice lack

the E3 ubiquitin ligase Cbl-b, which is a negative regulator of Vav, a critical guanine-nucleotide

exchange factor involved in CD28 signaling. Thus, T cells deficient in Cbl-b have enhanced

costimulatory signaling [227]. Less apoptosis was seen in Cbl-b -/- Vβ8+ T cells than wildtype

(Figure 13A) and adding anti-CD28 to cells lacking Cbl-b -/- could not further increase the

survival of Vβ8+ T cells (data not shown). Although Vβ8+ T cells from Cbl-b -/- mice displayed

a trend towards enhanced survival during 5 independent experiments (Figure 13B), statistical

significance was not achieved.

45

Figure 12: Blocking costimulation with CTLA4-Ig increases apoptosis of Vβ8+ T cells

(A) Wildtype splenocytes were cultured with SEB (0.3μg/ml) in the presence or absence of CTLA4-Ig (50μg/ml) for 5 days. Apoptosis of SEB-reactive T cells was determined by annexin V and Vβ8-TCR staining. Dead cells (7AAD+) were excluded from analysis and gating was determined by an isotype control stain. (B) The percentage of apoptotic Vβ8+ T cells within the total Vβ8+ population was plotted. Data from n=3 independent experiments are shown. Statistical significance was determined by the student’s paired t-test.

46

Figure 13: Deletion of Cbl-b enhances cell survival

(A) Wildtype and Cbl-b -/- splenocytes were cultured with SEB (0.3μg/ml) for 5 days. Apoptosis of SEB-reactive T cells was determined by annexin V and Vβ8-TCR staining. Dead cells (7AAD+) were excluded from analysis and gating was determined by an isotype control stain. (B) The percentage of apoptotic Vβ8+ T cells within the total Vβ8+ population was plotted. Data from n=5 independent experiments are shown. Statistical significance was determined by the student’s unpaired t-test.

47

6.4 TNFα enhances cell survival in a dose dependent manner Although TNFα has been reported to prevent the death of SAg-reactive cells in

vivo [193] it still remains unclear how this occurs. Moreover, since TNFα signaling has a range

of functions which include cytotoxicity, proliferation, and activation of nuclear factor kB, and act

through two different TNF receptors (TNFR) expressed on the majority of cell types [228], it is

unclear if TNFα acts on T cells directly, or indirectly via other cell types expressing TNFR.

To determine if TNFα plays a role in the survival of T cells in general, we

performed an experiment to see whether adding TNFα could enhance the viability of T cells in

the absence of SEB. Splenocytes were cultured in media in the presence or absence of

recombinant mouse TNFα. The survival of T cells was measured by staining with a viability dye

(sytox blue) and specific antibodies to CD3. The addition of TNFα increased the frequency of

viable cells (Sytox blue negative) from 11.8% to 27.2% as well as increased the frequency of live

T cells from 41.6% to 49.6% in the experiment shown (Figure 14A). Changes in the percentage

of total viable T cells were also plotted for n=3 independent experiments and was found to be

statistically significant (p = 0.0002) (Figure 14B). This effect was also found to be dose

dependent as increasing concentrations of TNFα led to increased T cell viability (Figure 14C).

Since SEB can also directly induce the production of TNFα by itself [107], we performed

another experiment to determine whether neutralization of endogenously produced TNFα would

affect Vβ8+ T cell death. The addition of neutralizing anti-TNFα antibody to cultures stimulated

with SEB, increased the frequency of apoptotic Vβ8+ T cells from 58.9% to 67.5% in the

experiment shown (Figure 15A). Changes in the percentage of apoptotic Vβ8+ T cells among

total Vβ8+ cells was plotted for n=4 independent experiments and was found to be statistically

significant (p = 0.0006) (Figure 15B).

48

Figure 14: TNFα enhances survival of T cells in a dose dependent manner

(A) Wildtype and Cbl-b -/- splenocytes were cultured with SEB (0.3μg/ml) for 5 days. Apoptosis of SEB-reactive T cells was determined by annexin V and Vβ8-TCR staining. Dead cells (7AAD+) were excluded from analysis and gating was determined by an isotype control stain. (B) The percentage of apoptotic Vβ8+ T cells within the total Vβ8+ population was plotted. Data from n=5 independent experiments are shown. Statistical significance was determined by the student’s paired t-test.

49

Figure 15: Neutralization of endogenous TNFα decreases survival of SEB-reactive T cells

(A) Wildtype splenocytes were cultured with SEB (0.3μg/ml) in the presence or absence of neutralizing anti-TNFα antibody for 5 days. Apoptosis of SEB-reactive T cells was determined by annexin V and Vβ8-TCR staining. Dead cells (7AAD+) were excluded from analysis and gating was determined by an isotype control stain. (B) The percentage of apoptotic Vβ8+ T cells within the total Vβ8+ population was plotted. Data from n=4 independent experiments are shown. Statistical significance was determined by the student’s paired t-test.

50

6.5 TNFα receptors on APC regulate superantigen-induced death of T cells

The previous results (see 6.4) support the idea that TNFα may have a pro-survival

role in SAg responses, but did not clearly demonstrate the mechanism of TNFα action. TNFRs

are expressed by both APC and T cells, thus it was unclear whether TNFα acts on T cells

directly, or indirectly via TNFRs on APC. To dissect the specific contributions of TNFRs, we

purified T cells and APC from wildtype and TNFR1-/-R2-/- mice, then mixed them together in

different combinations and stimulated them with SEB. In these mixing experiments, when

TNFR-deficient T cells were paired with wildtype APC, we observed no difference in the levels

of apoptosis of Vβ8+ T cells (mean Δ = -2.22 + 1.172%, p=0.1539) (Figure 16, right). However,

when TNFR-deficient APC were paired with wildtype T cells, we observed significant increases

in the apoptosis of SAg-reactive T cells compared to wildtype APC (mean Δ% = 9.77 + 1.234%,

p=0.0002) (Figure 16, left). These findings suggest that during SAg responses, TNFα acts

indirectly through the APC to affect cell survival.

51

Figure 16: TNFR on APC regulate survival of SEB-reactive T cells

T cells and APC from wildtype and TNFR1-/-R2-/- mice were purified then recombined and cultured with SEB (0.3μg/ml) at a ratio of 1:4. Apoptosis of SEB-reactive T cells was determined by annexin V and Vβ8-TCR staining. Dead cells (7AAD+) were excluded from analysis and gating was determined by an isotype control stain. The percentage of apoptotic Vβ8+ T cells within the total Vβ8+ population was plotted. Data from a minimum of n=4 independent experiments are shown. Statistical significance was determined by the student’s paired t-test.

52

6.6 TNFα promotes survival of T cells via TNFR on APC It remained unclear whether TNFα induced the survival of T cells directly via

TNFRs on T cells or indirectly via TNFRs present on APC. To dissect TNFα’s specific

contributions to T cell survival, splenic T cells and APC from wildtype and TNFR1-/-R2-/- mice

were purified, then recombined and stimulated with TNFα. After gating on CD3+ cells (Figure

17A), the percentage of dead T cells was quantified by staining with sytox blue (Figure 17B). In

these mixing experiments, when TNFR-deficient T cells were paired with wildtype APC, the

addition of TNFα decreased T cell death (mean Δ = 6.438 + 1.797%, p=0.009) (Figure 17B, top

row). In contrast, when TNFR-deficient APC were paired with wildtype T cells, the addition of

TNFα did not impact T cell death (mean Δ = -1.313 + 0.9255%, p=0.1991) (Figure 17B, bottom

row). Although an increase in T cell death displayed in the experiment shown (62.6% to 67%),

this difference was not statistically significant when the changes in the percentage of dead T cells

were plotted for n=8 independent experiments (Figure 17C). These findings suggest that TNFα

can act indirectly through the TNFR on APC to affect promote T cell survival.

53

Figure 17: TNFR on APC direct T cell survival

(A) T cells and APC from wildtype and TNFR1-/-R2-/- mice were purified, recombined, then cultured in the presence or absence of recombinant TNFα (12 ng/ml) for 2 days. T cell death was quantified by CD3 and sytox blue staining. T cells were gated on using light scattering properties and CD3 staining. (B) The percentage of sytox blue positive T cells (see gated) relative to all T cells was plotted. (C) Data from of n=8 independent experiments are shown here. Significance was determined using an unpaired student’s t-test.

54

6.7 TNFR does not affect SEB-induced proliferation TNFα’s promotion of cell survival of SEB-reactive Vβ8+ T cells could have

resulted from early differences in proliferation. For example, an increase in the initial

proliferation phase might contribute to differences in cell numbers observed after the expansion.

To investigate this, the proliferative potential of wildtype and TNFR1-/-R2-/- mice was assessed

by [3H] thymidine incorporation. Splenocytes were stimulated with different concentrations of

SEB and their proliferation measured at the peak of expansion (day 3). No significant difference

in proliferation was found between wildtype and TNFR1-/-R2-/- splenocytes at three different

doses of SEB (Figure 18).

Figure 18: Lack of TNFRs does not affect SEB-induced proliferation

Splenocytes from wildtype and TNFR1-/-R2-/- mice were stimulated with different doses of SEB for 3 days. [3H] thymidine was added to wells 24 hrs before scintillation counts. The stimulation index was calculated by dividing the mean disintegrations per minute (DPM) of stimulated cells by unstimulated controls. The average DPM was calculated from quadruplicate wells and data from n=3 independent experiments are shown. Statistical significance was determined by the student’s paired t-test.

55

6.8 Contribution of B7.2 on APC and T cells to survival of SEB-stimulated T cells

B7.2 expression occurs on B cells and T cells after SEB stimulation with SEB or

TNFα (Figure 8 and 9). Thus, it was possible that B7.2 expressing T cells could promote their

own survival by virtue of expression both B7.2 ligand and the CD28 receptor. To dissect the

contribution of B7.2 on APC vs T cells, additional mixing experiments were performed with

cells from B7.2 -/- mice. T cells and APC from wildtype and B7.2 -/- mice were purified, then

mixed them together in different combinations and stimulated with SEB. When B7.2-deficient

APC were paired with wildtype T cells, we observed no difference in the proportion of Vβ8+ T

cells undergoing apoptosis (Figure 19A, left). Similarly, no differences were observed when

B7.2-deficient T cells were paired with wildtype APCs (Figure 19A, right). Although the

proportion of Vβ8+ T cells undergoing apoptosis did not differ, a trend towards lower absolute

Vβ8+ T cell numbers was seen when B7.2-deficient APC were paired with wildtype T cells

(Figure 19B).

56

Figure 19: Contribution of B7.2 on APC and T cells to survival of SEB-reactive T cells

(A) T cells and APC from wildtype and B7.2-/-- mice were purified then recombined and cultured with SEB (0.3μg/ml) at a ratio of 1:4. Apoptosis of SEB-reactive T cells was determined by annexin V and Vβ8-TCR staining. Dead cells (7AAD+) were excluded from analysis and gating was determined by an isotype control stain. The percentage of apoptotic Vβ8+ T cells within the total Vβ8+ population was plotted. Data from n=8 independent experiments are shown. (B) The absolute viable Vβ8+ T cell counts were plotted. Statistical significance was determined by the student’s paired t-test.

57

6.9 Enhanced costimulation increases Bcl-XL and Bcl-2 expression

We have shown that costimulation can increase the survival of SEB-stimulated T

cells. Because studies had demonstrated that SAg-activated T cells downregulate pro-survival

proteins Bcl-XL and Bcl-2 [229], we also tested whether costimulation would alter the levels of

these anti-apoptotic proteins. Wildtype splenocyte were cultured with SEB in the presence or

absence of costimulatory anti-CD28. When costimulation was enhanced, Bcl-XL and Bcl-2

expression levels increased compared to SEB and Media controls (Figure 20A). Increased Bcl-XL

expression supports earlier data from Moolani et Al [192, 230] and is reproduced in Figure 20B-

left. Although increased Bcl-2 expression was also found in n=2 independent experiments,

repetitions will need to be performed to determine statistical significance (Figure 20B, right).

58

Figure 20: Enhanced costimulation increases expression of anti-apoptotic proteins

(A) Splenocytes harvested from wildtype mice were stimulated with SEB (0.3μg/ml) in the presence or absence of costimulatory anti-CD28 antibody (0.5μg/ml) for 2 (Bcl-XL staining) or 5 (Bcl-2 staining) days. Staining with antibodies specific to Vβ8-TCR, Bcl-XL, and Bcl-2 was performed. Gating was set on Vβ8+ T cells. Fluorescence minus one (FMO) control staining was performed on SEB stimulated cells and is shown in grey. Histograms represent a minimum of n=2 independent experiments. (B) Changes in MFI were plotted for the experiments shown in (A) (Bcl-XL left, Bcl-2 right). The plot of Bcl-XL MFI values (left) was taken from earlier work by Moolani et Al [230]. Statistical significance was determined by the student’s paired t-test where applicable (in this figure * denotes a significance of p<0.05).

59

6.10 Loss of TNFR decreases Bcl-2 and Bcl-XL in SEB-reactive T cells

It was shown that costimulation can increase the survival of SEB-stimulated T

cells with specific increases in the expression of pro-survival molecules Bcl-XL and Bcl-2. Given

that TNFR promotes the expression of costimulatory molecules (see 6.2), additional experiments

were performed to see whether the loss of TNFR would affect the expression of Bcl-2 and Bcl-XL

in Vβ8+ T cells during responses to SEB. It was shown that TNFR1-/-R2-/- T cells stimulated

with SEB had decreased expression of Bcl-2 (Figure 21A) and Bcl-XL (Figure 21B) in

comparison to wildtype controls after the initial expansion phase. These findings demonstrate

that, after exposure to SEB, TNFR signaling contributes to the expression of pro-survival

molecules.

6.11 TNFR contributes to the activity of surviving SEB-reactive T cells

In these studies, TNFα was seen to drive the survival of Vβ8+ T cells after SEB

stimulation. Since TNFR signaling was found to contribute to the survival of T cells (see 6.4-

6.6) and expression of pro-survival molecules (see 6.10), additional experiments were performed

to determine whether TNFR would also affect the activity of surviving Vβ8+ T cells. To measure

the differences in T cell activity within the SEB-reactive population, flow cytometry was used to

measure proliferation by incorporation of BrdU. In comparison to wildtype, Vβ8+ T cells from

TNFR1-/-R2-/- displayed decreased amounts of proliferation after stimulation with SEB for 5

days (Figure 22). This finding suggests that TNFα not only preserves the survival of SEB-

reactive T cells, but also contributes to their proliferative capacity after expansion.

60

Figure 21: TNFR promotes the expression of survival molecules in SEB-reactive T cells

Wildtype or TNFR-deficient splenocytes were stimulated with SEB (0.3μg/ml) for 5 days. Staining with antibodies specific to Vβ8-TCR, Bcl-XL, and Bcl-2 was performed. Histograms depict the expression of (A) Bcl-2 and (B) Bcl-XL of Vβ8+ T cells. Light scatter properties and isotype controls were used to set the gates on histograms. The histograms represent a minimum of n=2 independent experiments.

61

Figure 22: TNFR contributes to the activity of surviving Vβ8+ T cells

(A) Wildtype or TNFR-deficient splenocytes were stimulated with SEB (0.3μg/ml) for 5 days. Brdu was added to cultures 24hrs before staining with antibodies specific to Vβ8-TCR and BrdU. Analysis was performed on SEB-reactive T cells after gating on Vβ8+ T cells with isotype controls. (B) The percentage of proliferating Vβ8+ T cells within the total Vβ8+ population was plotted. Data from n=3 independent experiments are shown. Statistical significance was determined by the student’s paired t-test.

62

7 DISCUSSION

7.1 TNFα drives survival of SAg-reactive T cells by costimulation

The presence of a persistent T cell infiltrate in the coronary artery of KD patients

is not in keeping with the established fates of SAg-activated T cells, which undergo death and

anergy. Previous studies demonstrated that TNFα has important roles in regulating apoptosis and

that TNFα may also be able to prevent the death of SAg-stimulated T cells (see below and 1.4.3).

In addition, work in the LCWE model has demonstrated that TNFα and its receptors are critically

required for pathogenesis. In these studies, we investigated the effects of TNFα on the fate of

SAg-activated T cells. Understanding how TNFα alters the fate of SAg-activated T cells may

help explain how T cells persist in KD, and further legitimizes the use of therapies targeting

TNFα for the treatment of KD.

The functions of TNFα and its receptors TNFR1(p55) and TNFR2(p75) have been

well studied. They include necrotic and apoptotic cell death, proliferation, differentiation, and

inflammation. The breadth of functions is thought to result from the diversity of proteins and cell

signaling pathways TNFα activates. Although the number of activated cell signaling pathways is

large, it has become clear that TNFRs have key roles in regulating apoptosis. Reports have

shown both pro-apoptotic and anti-apoptotic roles for TNFα. For example, TNFR1 is understood

to associate with TNF receptor associated death domain protein (TRADD), which can recruit

Fas-associated protein with death domain (FADD), to activate the initiator caspase 8. While, in

contrast, TNFR2 through its association with the TNF receptor associated factor 2 (TRAF2) can

activate the NF-κB pathway, which transcriptionally regulates cellular inhibitor of apoptosis

protein 2 (c-IAP2), and can block caspase 8 activation and apoptosis.

It has been reported that SEB-stimulated T cells undergo similar levels of

expansion and deletion in triple knockout mice lacking Fas, TNFR1, and TNFR2 [229], and thus

there may be alternative pathways to death in SAg-activated T cells. The authors of this paper

demonstrated the existence of an intrinsic cell death pathway triggered by SAg. It was found that

63

SEB-activation caused Bcl-2 downregulation in T cells, and that overexpression of Bcl-2 in

transgenic mice permitted the T cells to survive. Their results agree with other investigators who

have suggested alternative pathways to cell death in activated T cells that involve cytokine

withdrawal [231], as it has also been shown that administration of IL-2 [231] or TNFα [194] can

promote the survival of SEA-activated T cells. One interesting study by Vella et al. found that

coadministration of LPS could preserve survival of SAg reactive T cells in vivo [193]. Although

the precise mechanism was unclear, the increased survival mediated by LPS was found to

depend on the production of TNFα.

In this study, TNFα was found to alter the fate of SAg-stimulated T cells,

promoting their survival though regulating costimulation. It was shown that stimulation with

SEB or TNFα, which is produced during SAg-responses, can upregulate B7.2 expression on

splenocytes, and more specifically, on B cells (Figure 8 and 9). Although B cells have been

known to upregulate B7.2 when activated with other stimuli such as alloantigen, LPS, or

concanavalin A [232], it has not been demonstrated that TNFα and its receptors could directly

drive B7.2 expression. Interestingly, upregulation of B7.1 was not detected on splenocytes in

response to SEB or TNFα. This might have been due to the time frame of analysis, as it has been

suggested that B7.1 and B7.2 are expressed with different kinetics [233]. However, given that a

kinetic analysis of B7.1 was explored in this study, and that B7.1 and B7.2 have been found to be

expressed on B cells after 18-42 hrs of LPS stimulation [234], it is unlikely that B7.1 was

undetected because of the time frame chosen. Instead, it could be that different stimuli, such as

SEB or TNFα, drive differential expression of B7.1 and B7.2. This idea is supported by

published data showing a decreased role for B7.1 than B7.2 during SAg stimulation [226].

While it was demonstrated that stimulation with recombinant TNFα could

upregulate B7.2 on wildtype splenic B cells in a dose dependent manner (Figure 9B), this effect

was also partially seen in TNFR1-/-R2-/- B cells. This finding could be due to the preparation of

TNFα used, as commercial preparations may contain traces of endotoxin, which are known to

upregulate B7 expression [234]. It is also possible that the differences are a result of inherent

64

experimental variability. However, a direct relationship between higher doses of TNFα and

larger differences in MFI was found in TNFR1-/-R2-/- cultures (data not shown).

During SEB immune mediated responses, TNF receptors were shown to be

required for B7.2 expression as upregulation of B7.2 was only seen in wildtype and not TNFR1-/-

R2-/- cultures (Figure 10). This result indicated that TNF receptors are responsible for the

regulation of B7.2 expression during SAg immune responses, and suggested that TNFα might

alter the fate of SAg-stimulated T cells via the costimulatory pathway. It should be noted that

during these experiments, multiple doses of SEB (0.03, 0.3, and 3 μg/ml) were also compared.

Abrogation of B7.2 expression in TNFR1-/-R2-/- cultures was always seen at 0.03μg/ml; however,

at higher doses, B7.2 expression was consistently diminished although not completely disrupted

(data not shown).

Although increases in B7.2 expression were measured when SEB or TNFα was

added to cultures (Figures 8, 9 and 10), it should be acknowledged that these differences could

be, in part, attributed to potential increases in the auto-fluorescence of stimulated cells. In

retrospect, even though control stains (isotype and fluorescence minus one) were performed on

activated cells, additional control stains on unactivated cells could have been included to

determine if changes in background auto-fluorescence occurred in the experiments.

Our lab had previously demonstrated that SEB-stimulated T cells could be

rescued from apoptosis by enhanced costimulation [192]. Since TNFα signaling was found to be

critical for costimulatory ligand expression during SEB responses (Figure 10), further

experiments were performed to explore the specific links between costimulation and the survival

of SEB-reactive T cells. In addition to recapitulating initial results which demonstrated that

adding anti-CD28 to SEB stimulated cultures could rescue Vβ8+ T cells from apoptosis (Figure

11), it was further demonstrated that blockade of B7 ligands with CTLA4-Ig increased the

apoptosis of SEB-reactive Vβ8+ T cells (Figure 12) and deletion of Cbl-b, which negatively

65

regulates vav during CD28 signaling (see 1.5.2), decreased apoptosis of SEB-reactive Vβ8+ T

cells (Figure 13). The results further suggest that costimulation provided by B7 plays an

important role in preventing the death of SEB-activated T cells. Although statistical significance

(p = 0.08) was not achieved with the number of Cbl-b -/- experiments performed (n = 6), a trend

towards increased survival of T cells was observed (Figure 13B). More experiments with Cbl-b -

/- mice would be expected to clearly resolve this difference in apoptosis.

Other studies which have looked at costimulatory molecules CD28 and B7 during

superantigen responses have mainly focused on the initial activation and proliferation of T cells

[222, 223]. Conflicting data show that the precise role of B7 ligands in regulating cell survival is

unclear. For example, a study using transgenic mice expressing soluble CTLA4-Ig found no

differences in the numbers of viable Vβ8+ cells after treatment with SEB [225], while another

other group showed that injection of CTLA4-Ig decreased the numbers of SEA-reactive cells

[194]. Complicating this discrepancy further is the acknowledgement that blocking costimulation

would also inhibit the initial activation of T cells, and thus might prevent a clear measurement of

cell survival later on. To avoid this problem, experiments were performed where costimulation

was not blocked, but enhanced with anti-CD28 antibody or with Cbl-b -/- mice (Figure 11 and

13). In this way, cells from both conditions would not have defects in the initial proliferation and

the differences in the survival of cells would not be attributed to defects in expansion. However,

this approach has limitations as well. In optimization experiments, increased levels of

proliferation were noted in SEB stimulated cultures when anti-CD28 antibody was added (data

not shown), thus it remains possible that the differences in apoptosis measured may also reflect

different expansion kinetics of Vβ8+ T cells. This consideration would also affect experiments

performed with Cbl-b -/- mice, as they are also noted to have increased capacity to proliferate

[205]. Determining whether this difference in cell survival is due to increased costimulatory

signaling or due to increased expansion remains an interesting question that could be addressed

in the future. To resolve the expansion and survival components in the presence or absence of

costimulation, CFSE dye could be used to track successive generations of Vβ8+ T cells. With

this approach, the levels of apoptosis between conditions could be compared while normalizing

between generations.

66

Additional experiments revealed that TNFα itself also had a pro-survival effect on

resting T cells and on SEB-activated T cells. In the absence of SAg, the addition of recombinant

TNFα enhanced T cell viability in a dose dependent manner (Figure 14), while in the presence of

SEB-stimulation, neutralization of TNFα increased apoptosis (Figure 15). These findings suggest

that TNFα contributes to T cell survival during SAg-responses and are in agreement with

published data which demonstrated that the addition of recombinant TNFα promotes the survival

of SEA-stimulated T cells [194]. The authors of this paper did not clarify the specific

mechanism(s) by which TNFα interferes with the apoptosis of SEA-activated cells.

In this thesis, experiments also aimed to dissect the specific contributions of

TNFR from various different cell types to the survival of SAg-reactive T cells. Since TNFRs are

expressed on many different cells types, including B cells and T cells (see 1.4.1), TNFα-driven

survival might have resulted from stimulation of T cells, by upregulation of costimulatory

molecules APC, or by both. Mixing experiments revealed that TNFR present on APC drive the

survival of SEB-reactive T cells while the absence of TNFR signaling on T cells did not affect

survival (Figure 16A). Additional mixing experiments also demonstrated that stimulation with

recombinant TNFα promotes the survival of T cells via TNFRs on the APC, and not by TNFRs

present on the T cells themselves (Figure 16). Thus, these results, viewed together with previous

observations, suggest that TNFα acts to promote the survival of SAg-stimulated cells indirectly,

through modulating costimulatory molecule expression on APC.

Similar to previous costimulation experiments, it was also possible that

differences in cell survival could be attributed to differences in proliferation between wildtype

and TNFR-deficient cells as TNFR have been reported to promote proliferation [158]. However,

thymidine incorporation experiments revealed no significant statistical difference in proliferation

between wildtype and TNFR -/- cultures in response to three doses of SEB (Figure 18). Taken

together, the results suggest that the differences in apoptosis were not due to a significant failure

67

of TNFR -/- cells to expand, but could result from defects in survival pathways which are driven

by TNFR such as B7.2.

Additional mixing experiments using B7.2 -/- mice revealed that B7.2 present on

the APC may be important for cell survival after exposure to SEB (Figure 19). Although this

result did not reach statistical significance, a trend towards increased survival was seen when

absolute cell counts were graphed. The inability to reach statistical significance could be due to

the inherent variability between experiments and the small magnitudes of change being

measured. In particular, compensatory mechanisms may diminish the effect of a single

deficiency in costimulation as preliminary experiments from our lab have shown that 4-1BBL

(another costimulatory ligand and receptor of the TNF superfamily [235]) is also upregulated

after exposure to SAg (unpublished observation) and may thus be able to compensate for a loss

in B7.2 function.

Apoptosis is triggered by the extrinsic or intrinsic pathways. In the extrinsic

pathway, receptors containing death domains, such as Fas and TNFR1, induce apoptosis by

recruiting the adaptor FADD which leads to activation of caspase 8 and initiation of apoptosis. In

the intrinsic pathway, external stresses, growth factor withdrawal, or cytotoxic drugs are shown

to initiate apotosis in a Fas/TNFR1-independent pathway which strongly regulated by Bcl

proteins Bim and Bcl-2 [236]. Bcl proteins regulate the permeability of the mitochondria and

release of cytochrome c, which interacts with APAF1 and caspase 9 to initiate apoptosis. The

cell death induced by exposure to SAg may involve both extrinsic and intrinsic pathways. For

example, some report that SEB-activated CD4+ T cells express Fas and commit fratricide [237],

while others report have shown that SEB-stimulated results in an intrinsic death pathway,

independent of death receptor signaling. For example, triple knockout mice lacking Fas, TNFR1,

and TNFR2 are seen to undergo similar amounts of apoptosis in response to SAg [229]. An

intrinsic death pathway is thought to be activated and regulated by Bcl proteins such as Bid and

Bcl-2. For example, Vβ8+ T cells from Bid -/- mice or from Bcl-2 overexpression transgenic

mice are shown to survive after stimulation with SEB. Apoptosis by the extrinsic or intrinsic

pathways is thought to be due to differences in the initial stimulation has been suggested to be

68

due to differences in experimental design. In experiments using recurrent SAg stimulation, roles

for Fas have been observed, while experiments using a single dose of SAg are found to be Fas-

independent [229]. It is argued that restimulation studies are not realistic because clearance of

SAg occurs quickly in vivo, which precludes restimulation [238]. In this thesis, a single exposure

to SEB was always used.

In this study, TNFα was seen to drive the expression of costimulatory ligands and

interfere with the apoptosis of SEB-activated T cells. This TNFα-dependent survival may be

driven by costimulatory signals which could alter the intrinsic pathway of cell death triggered by

SEB exposure. Given that Bcl-2 is strongly downregulated after SEB stimulation and that

decreased levels correlate with the death of T cells [229], experiments were performed to

determine whether enhanced costimulation could alter the expression of Bcl-2. It was shown here

that after SEB-activation, enhanced costimulation with anti-CD28 increased the expression of

Bcl-2 and Bcl-XL in Vβ8+ T cells (Figure 20). These finding suggests that costimulation can

rescue Vβ8+ T cells by increasing anti-apoptotic Bcl family members, Bcl-2 and Bcl-XL. This

result is also in agreement with another report which demonstrated that CD28 signaling promotes

T cell survival by enhancing Bcl-XL expression [213]. Interestingly, the study also reported that

CD28 costimulation does not affect Bcl-2 expression, while an increase in Bcl-2 was found in

this work. This discrepancy could be due different initial stimulatory agents, as their study used

anti-CD3 antibodies to activate T cells, and in this report, SEB was used. Thus, it could be that

the increased expression of Bcl-2 seen with enhanced costimulation does not reflect a direct

upregulation of Bcl-2 expression per se, but may interfere with the downregulation of Bcl-2

during SAg activation. In this study, it was also demonstrated that mice deficient in TNFRs have

decreased expression of Bcl-2 and Bcl-XL in response to SEB (Figure 21). This finding agrees

with the idea that costimulation driven by TNFR promotes an interference of the intrinsic Bcl-2

death pathway triggered by SEB stimulation. The defect in the survival of these T cells observed

after expansion also correlated with less proliferation of Vβ8+ TNFR-deficient T cells (Figure

22). These findings also help explain why growth factors, such as TNFα, can promote the

survival of SEB-activated T cells, yet may not exert their effects directly through TNFRs

expressed on T cells.

69

7.2 TNFα and Autoimmunity Proinflammatory cytokines coordinate and cause most of the generalized events

observed during inflammation. Since inflammation seems to accompany most forms of

autoimmune disease, it seemed intuitive that key cytokines, such as TNFα, may also be involved

in autoimmunity [239]. Over the years, many studies have reported key roles for TNFα in

different disease models such as type 1 diabetes mellitus (T1DM) and rheumatoid arthritis (RA).

Interestingly, TNFα is seen to have complex functions as it can be both harmful and helpful in

the development of autoimmunity [240]. Of particular relevance to our laboratory, TNFα does

seem to have a pathogenic role (see 1.3.3 TNFα in the LCWE model), although the precise

mechanism(s) by which TNFα mediates disease is not well understood. By understanding how

TNFα acts in different autoimmune modalities, we can begin to infer how TNFα may be acting

in LCWE-initiated KD.

In the NOD mouse model of type I diabetes mellitus, TNFα is thought to have an

overall pathogenic role [240]. TNFα is expressed in the in tra-islet infiltrate, predominantly by

CD4+ T cells, and contributes to the development of spontaneous diabetes [241]. It has also been

shown that when NOD-scid mice are injected with lymphocytes from NOD donors, they develop

insulitis and diabetes within 120 days; however, with early exposure to a chimeric TNFα

inhibitor [242], diabetes is prevented entirely [243]. Yet, in contrast to these results, it was

shown that systemic administration of TNFα has a protective effect in NOD mice [244]. In view

of these results it becomes clear that TNFα’s role is complex; likely having multiple non-

mutually exclusive effects. For example, it has been suggested that the harmful effects of TNFα

are driven locally in the islet, while systemic TNFα has immunoregulatory functions that dampen

autoimmunity [240].

TNFα has also been demonstrated to have multiple roles in RA. Clinical studies

have shown that TNFα is highly expressed in the synovial fluid of RA patients [245, 246].

Numerous studies using experimental models of RA have demonstrated pathogenic roles for

TNFα, in the initiation, maintenance, and progression of the local inflammatory response and

70

tissue injury in arthritis (reviewed in [247]). Similar to experimental models of diabetes, use of

neutralizing TNFα antibodies was seen to ameliorate arthritic lesions [248, 249], yet in contrast

to diabetes, systemic administration of TNFα exacerbated the disease [250]. Thus TNFα likely

has disease-specific roles, despite sharing many similarities in driving and perpetuating

autoimmunity.

In view of the functions of TNFα in RA and T1DM, KD shares many similarities.

The LCWE model (see 1.3.1) displays characteristics of dysregulated TNFα production as well

as localized TNFα expression within the target organ of autoimmunity. In fact, it has been shown

by our laboratory that local TNFα expression drives the production and enzymatic activity of

matrix metalloproteinase 9 (MMP9), which degrades the elastin layer of vessels resulting in

aneurysm formation [86, 87]. This localized production of TNFα and subsequent destruction of

target tissue in KD is quite similar to RA and T1DM, where TNFα is seen to stimulate MMP9

production from synoviocytes [251], as well as, potentiate the toxic effects of IL-1 on pancreatic

islets cells [252]. Additionally, interfering with TNFα with neutralizing antibodies or by genetic

deletion of the receptor is also seen to ameliorate disease [147]. Thus, with knowledge that TNFα

acts in multiple capacities in different autoimmune disorders, it may be the case that TNFα plays

a larger systemic role in dysregulating the immune system in KD. Aside from inducing MMP9

production locally, we suggest in this study that TNFα contributes to the survival of a

subpopulation of pathogenic cells T cells which remain after SAg-induced expansion.

7.3 Future Directions Since this work has demonstrated that TNFα alters the fate of SAg-stimulated T

cells via costimulation on APC, it would be interesting to investigate the role of TNFα-driven

costimulation more closely in vivo. One future experiment could be performed using mice which

lack TNFR1 and TNFR2 on their APC only. The receptors could be knocked out using a

Cre/Lox system with Cre recombinase expressed under an MHC class II promoter and loxP sites

added to both TNFR1 and TNFR2. Injection of LCWE into these mice would determine the

contribution of TNFR on APC in our disease model. This question is particularly interesting

71

because APC have been found to localize with T cells and accumulate in coronary artery lesions

[46]. In addition, because costimulation was found to be important for the survival of SAg-

reactive T cells, future experiments could also investigate this effect in our disease model by

injecting LCWE into CD28-/- mice or transgenic mice which overexpress Bcl-2, as it has been

shown that overexpression of Bcl-2 prevents the death of SAg-reactive T cells [229]. The

severity of disease would be expected to be decreased in CD28-/- mice and increased when Bcl-2

is overexpressed. In support of these experiments, our lab has recently demonstrated that

coinjection of LCWE and a 4-1BB agonist can aggravate coronary disease [230].

7.4 Conclusions

It is known that TNFα is important in human KD and the LCWE mouse model.

Expression of TNFα is upregulated during disease induction and the lack of TNFα signaling in

the LCWE model was seen to completely prevent disease [147]. The findings in this work

suggest that TNFα promotes the survival of SEB-stimulated T cells by regulating costimulatory

molecule expression on APC. These observations are important because they shed light on the

mechanisms which govern T cell death after SAg exposure and can be represented in a disease

model (Figure 23). Perhaps most interestingly, these findings provide a way through a paradox:

that SAg-mediated autoimmunity can occur, despite the majority of activated cells being deleted.

Studies in mouse models have already shown that TNFα may contribute to different autoimmune

conditions such as diabetes [243, 253], KD [147], and arthritis [248, 249]. In addition, this study

also lends support to the autoimmune therapies which target TNFα and B7.

72

Figure 23: Model for TNFα-driven survival of SAg activated T cells

A single exposure to SAg results in initiation of the intrinsic cell death pathway which is regulated by Bcl proteins [229] and not the extrinsic pathway [229, 236]. Production of TNFα from APC and T cells drives the expression of B7.2 on APC, which provides costimulation to T cells and rescues them from cell death and modulates Bcl-2 and Bcl-XL expression. Given the importance of TNFα in autoimmune diseases (see 1.4.2), autoreactive T cells stimulated by SAg might escape cell death and persist with TNFα stimulation.

73

8 REFERENCES

1. Kawasaki, T., [Acute febrile mucocutaneous syndrome with lymphoid involvement with

specific desquamation of the fingers and toes in children]. Arerugi, 1967. 16(3): p. 178-222.

2. Newburger, J.W., et al., Diagnosis, treatment, and long-term management of Kawasaki

disease: a statement for health professionals from the Committee on Rheumatic Fever,

Endocarditis and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American

Heart Association. Circulation, 2004. 110(17): p. 2747-71.

3. Committee, T.J.K.D.R., Diagnostic Guideline of Kawasaki Disease [in Japanese]. 1984:

Tokyo, Japan.

4. Kim, D.S., Kawasaki disease. Yonsei Med J, 2006. 47(6): p. 759-72.

5. Dajani, A.S., et al., Diagnosis and therapy of Kawasaki disease in children. Circulation,

1993. 87(5): p. 1776-80.

6. Wolff, A.E., K.E. Hansen, and L. Zakowski, Acute Kawasaki disease: not just for kids. J

Gen Intern Med, 2007. 22(5): p. 681-4.

7. Taubert, K.A. and S.T. Shulman, Kawasaki disease. Am Fam Physician, 1999. 59(11): p.

3093-102, 3107-8.

8. Durongpisitkul, K., et al., The prevention of coronary artery aneurysm in Kawasaki

disease: a meta-analysis on the efficacy of aspirin and immunoglobulin treatment. Pediatrics,

1995. 96(6): p. 1057-61.

9. Furusho, K., et al., High-dose intravenous gammaglobulin for Kawasaki disease. Lancet,

1984. 2(8411): p. 1055-8.

10. Newburger, J.W., et al., The treatment of Kawasaki syndrome with intravenous gamma

globulin. N Engl J Med, 1986. 315(6): p. 341-7.

74

11. Muta, H., et al., Early intravenous gamma-globulin treatment for Kawasaki disease: the

nationwide surveys in Japan. J Pediatr, 2004. 144(4): p. 496-9.

12. Newburger, J.W., et al., A single intravenous infusion of gamma globulin as compared

with four infusions in the treatment of acute Kawasaki syndrome. N Engl J Med, 1991. 324(23):

p. 1633-9.

13. Hashino, K., et al., Re-treatment for immune globulin-resistant Kawasaki disease: a

comparative study of additional immune globulin and steroid pulse therapy. Pediatr Int, 2001.

43(3): p. 211-7.

14. Wright, D.A., et al., Treatment of immune globulin-resistant Kawasaki disease with

pulsed doses of corticosteroids. J Pediatr, 1996. 128(1): p. 146-9.

15. Furukawa, S., et al., Pentoxifylline and intravenous gamma globulin combination therapy

for acute Kawasaki disease. Eur J Pediatr, 1994. 153(9): p. 663-7.

16. Burns, J.C., et al., Infliximab treatment of intravenous immunoglobulin-resistant

Kawasaki disease. J Pediatr, 2008. 153(6): p. 833-8.

17. Burns, J.C., et al., Infliximab treatment for refractory Kawasaki syndrome. J Pediatr,

2005. 146(5): p. 662-7.

18. Rowley, A.H., F. Gonzalez-Crussi, and S.T. Shulman, Kawasaki syndrome. Adv Pediatr,

1991. 38: p. 51-74.

19. Lin, Y.T., et al., Repeated systematic surveillance of Kawasaki disease in Ontario from

1995 to 2006. Pediatr Int, 2010. 52(5): p. 699-706.

20. Holman, R.C., et al., Kawasaki syndrome hospitalizations in the United States, 1997 and

2000. Pediatrics, 2003. 112(3 Pt 1): p. 495-501.

21. Esper, F., et al., Association between a novel human coronavirus and Kawasaki disease. J

Infect Dis, 2005. 191(4): p. 499-502.

75

22. Boehme, K.W., M. Guerrero, and T. Compton, Human cytomegalovirus envelope

glycoproteins B and H are necessary for TLR2 activation in permissive cells. J Immunol, 2006.

177(10): p. 7094-102.

23. Catalano-Pons, C., et al., Primary cytomegalovirus infection, atypical Kawasaki disease,

and coronary aneurysms in 2 infants. Clin Infect Dis, 2005. 41(5): p. e53-6.

24. Huber, B.T., P.N. Hsu, and N. Sutkowski, Virus-encoded superantigens. Microbiol Rev,

1996. 60(3): p. 473-82.

25. Usta Guc, B., et al., Cytomegalovirus infection in a patient with atypical Kawasaki

disease. Rheumatol Int, 2008. 28(4): p. 387-9.

26. Gaudreault, E., et al., Epstein-Barr virus induces MCP-1 secretion by human monocytes

via TLR2. J Virol, 2007. 81(15): p. 8016-24.

27. Kikuta, H., et al., Kawasaki disease and an unusual primary infection with Epstein-Barr

virus. Pediatrics, 1984. 73(3): p. 413-4.

28. Kikuta, H., et al., Epstein-Barr virus genome-positive T lymphocytes in a boy with

chronic active EBV infection associated with Kawasaki-like disease. Nature, 1988. 333(6172): p.

455-7.

29. Holm, J.M., L.K. Hansen, and H. Oxhoj, Kawasaki disease associated with parvovirus

B19 infection. Eur J Pediatr, 1995. 154(8): p. 633-4.

30. Matsubara, K. and T. Fukaya, The role of superantigens of group A Streptococcus and

Staphylococcus aureus in Kawasaki disease. Curr Opin Infect Dis, 2007. 20(3): p. 298-303.

31. Shinomiya, N., et al., Variant Streptococcus sanguis as an etiological agent of Kawasaki

disease. Prog Clin Biol Res, 1987. 250: p. 571-2.

32. Shimizu, T., Y. Kida, and K. Kuwano, Triacylated lipoproteins derived from

Mycoplasma pneumoniae activate nuclear factor-kappaB through toll-like receptors 1 and 2.

Immunology, 2007. 121(4): p. 473-83.

76

33. Wang, J.N., et al., Mycoplasma pneumoniae infection associated with Kawasaki disease.

Acta Paediatr, 2001. 90(5): p. 594-5.

34. Nakamura, Y., et al., Epidemiologic features of Kawasaki disease in Japan: results from

the nationwide survey in 2005-2006. J Epidemiol, 2008. 18(4): p. 167-72.

35. Fujita, Y., et al., Kawasaki disease in families. Pediatrics, 1989. 84(4): p. 666-9.

36. Harada, F., et al., Genetic analysis of Kawasaki syndrome. Am J Hum Genet, 1986.

39(4): p. 537-9.

37. Hirata, S., Y. Nakamura, and H. Yanagawa, Incidence rate of recurrent Kawasaki disease

and related risk factors: from the results of nationwide surveys of Kawasaki disease in Japan.

Acta Paediatr, 2001. 90(1): p. 40-4.

38. Yanagawa, H., et al., Results of the nationwide epidemiologic survey of Kawasaki

disease in 1995 and 1996 in Japan. Pediatrics, 1998. 102(6): p. E65.

39. Amano, S., et al., General pathology of Kawasaki disease. On the morphological

alterations corresponding to the clinical manifestations. Acta Pathol Jpn, 1980. 30(5): p. 681-94.

40. Brown, T.J., et al., CD8 T lymphocytes and macrophages infiltrate coronary artery

aneurysms in acute Kawasaki disease. J Infect Dis, 2001. 184(7): p. 940-3.

41. Rowley, A.H., et al., IgA plasma cells in vascular tissue of patients with Kawasaki

syndrome. J Immunol, 1997. 159(12): p. 5946-55.

42. Rowley, A.H., et al., IgA plasma cell infiltration of proximal respiratory tract, pancreas,

kidney, and coronary artery in acute Kawasaki disease. J Infect Dis, 2000. 182(4): p. 1183-91.

43. Furukawa, F., G. Ohshio, and Y. Hamashima, Possible polyclonal B cell activation in

mucocutaneous lymph node syndrome. Eur J Pediatr, 1986. 145(1-2): p. 104-8.

44. Rowley, A.H., et al., Searching for the cause of Kawasaki disease--cytoplasmic inclusion

bodies provide new insight. Nat Rev Microbiol, 2008. 6(5): p. 394-401.

77

45. Rowley, A.H., et al., RNA-containing cytoplasmic inclusion bodies in ciliated bronchial

epithelium months to years after acute Kawasaki disease. PLoS One, 2008. 3(2): p. e1582.

46. Yilmaz, A., et al., Activated myeloid dendritic cells accumulate and co-localize with

CD3+ T cells in coronary artery lesions in patients with Kawasaki disease. Exp Mol Pathol,

2007. 83(1): p. 93-103.

47. Nagata, S., et al., Immunohistochemical studies on small intestinal mucosa in Kawasaki

disease. Pediatr Res, 1993. 33(6): p. 557-63.

48. Kimura, J., et al., Th1 and Th2 cytokine production is suppressed at the level of

transcriptional regulation in Kawasaki disease. Clin Exp Immunol, 2004. 137(2): p. 444-9.

49. Furuno, K., et al., CD25+CD4+ regulatory T cells in patients with Kawasaki disease. J

Pediatr, 2004. 145(3): p. 385-90.

50. Harrington, L.E., et al., Interleukin 17-producing CD4+ effector T cells develop via a

lineage distinct from the T helper type 1 and 2 lineages. Nat Immunol, 2005. 6(11): p. 1123-32.

51. Bettelli, E., et al., Reciprocal developmental pathways for the generation of pathogenic

effector TH17 and regulatory T cells. Nature, 2006. 441(7090): p. 235-8.

52. Jia, S., et al., The T helper type 17/regulatory T cell imbalance in patients with acute

Kawasaki disease. Clin Exp Immunol, 2010. 162(1): p. 131-7.

53. Onouchi, Y., et al., A genomewide linkage analysis of Kawasaki disease: evidence for

linkage to chromosome 12. J Hum Genet, 2007. 52(2): p. 179-90.

54. Onouchi, Y., Molecular genetics of Kawasaki disease. Pediatr Res, 2009. 65(5 Pt 2): p.

46R-54R.

55. Onouchi, Y., et al., ITPKC functional polymorphism associated with Kawasaki disease

susceptibility and formation of coronary artery aneurysms. Nat Genet, 2008. 40(1): p. 35-42.

56. Lewis, R.S., Calcium signaling mechanisms in T lymphocytes. Annu Rev Immunol,

2001. 19: p. 497-521.

78

57. Pattni, K. and G. Banting, Ins(1,4,5)P3 metabolism and the family of IP3-3Kinases. Cell

Signal, 2004. 16(6): p. 643-54.

58. Chun, J.K., et al., Programmed death-1 (PD-1) gene polymorphisms lodged in the genetic

predispositions of Kawasaki Disease. Eur J Pediatr, 2010. 169(2): p. 181-5.

59. Huang, F.Y., et al., Genetic polymorphisms in the CD40 ligand gene and Kawasaki

disease. J Clin Immunol, 2008. 28(5): p. 405-10.

60. Onouchi, Y., et al., CD40 ligand gene and Kawasaki disease. Eur J Hum Genet, 2004.

12(12): p. 1062-8.

61. Armitage, R.J., et al., Molecular and biological characterization of a murine ligand for

CD40. Nature, 1992. 357(6373): p. 80-2.

62. Yamazaki, T., et al., Expression of programmed death 1 ligands by murine T cells and

APC. J Immunol, 2002. 169(10): p. 5538-45.

63. Freeman, G.J., et al., Engagement of the PD-1 immunoinhibitory receptor by a novel B7

family member leads to negative regulation of lymphocyte activation. J Exp Med, 2000. 192(7):

p. 1027-34.

64. Armitage, R.J., et al., CD40 ligand is a T cell growth factor. Eur J Immunol, 1993. 23(9):

p. 2326-31.

65. Wang, C.L., et al., Expression of CD40 ligand on CD4+ T-cells and platelets correlated

to the coronary artery lesion and disease progress in Kawasaki disease. Pediatrics, 2003. 111(2):

p. E140-7.

66. Henn, V., et al., CD40 ligand on activated platelets triggers an inflammatory reaction of

endothelial cells. Nature, 1998. 391(6667): p. 591-4.

67. Yellin, M.J., et al., Functional interactions of T cells with endothelial cells: the role of

CD40L-CD40-mediated signals. J Exp Med, 1995. 182(6): p. 1857-64.

79

68. Eberhard, B.A., et al., Evaluation of the cytokine response in Kawasaki disease. Pediatr

Infect Dis J, 1995. 14(3): p. 199-203.

69. Maury, C.P., E. Salo, and P. Pelkonen, Elevated circulating tumor necrosis factor-alpha

in patients with Kawasaki disease. J Lab Clin Med, 1989. 113(5): p. 651-4.

70. Hirao, J., et al., High levels of circulating interleukin-4 and interleukin-10 in Kawasaki

disease. Int Arch Allergy Immunol, 1997. 112(2): p. 152-6.

71. Rowley, A.H., et al., Serum interferon concentrations and retroviral serology in Kawasaki

syndrome. Pediatr Infect Dis J, 1988. 7(9): p. 663-6.

72. Lin, C.Y., et al., Serial changes of serum interleukin-6, interleukin-8, and tumor necrosis

factor alpha among patients with Kawasaki disease. J Pediatr, 1992. 121(6): p. 924-6.

73. Leung, D.Y., et al., Endothelial cell activation and high interleukin-1 secretion in the

pathogenesis of acute Kawasaki disease. Lancet, 1989. 2(8675): p. 1298-302.

74. Sohn, M.H., et al., Circulating interleukin 17 is increased in the acute stage of Kawasaki

disease. Scand J Rheumatol, 2003. 32(6): p. 364-6.

75. Mosmann, T.R. and R.L. Coffman, TH1 and TH2 cells: different patterns of lymphokine

secretion lead to different functional properties. Annu Rev Immunol, 1989. 7: p. 145-73.

76. Acosta-Rodriguez, E.V., et al., Interleukins 1beta and 6 but not transforming growth

factor-beta are essential for the differentiation of interleukin 17-producing human T helper cells.

Nat Immunol, 2007. 8(9): p. 942-9.

77. Furukawa, S., et al., Serum levels of p60 soluble tumor necrosis factor receptor during

acute Kawasaki disease. J Pediatr, 1994. 124(5 Pt 1): p. 721-5.

78. Weng, K.P., et al., IL-10 polymorphisms are associated with coronary artery lesions in

acute stage of Kawasaki disease. Circ J, 2010. 74(5): p. 983-9.

79. Burns, J.C., et al., Family-based association analysis implicates IL-4 in susceptibility to

Kawasaki disease. Genes Immun, 2005. 6(5): p. 438-44.

80

80. Wu, S.F., et al., Association of IL-1Ra gene polymorphism, but no association of IL-

1beta and IL-4 gene polymorphisms, with Kawasaki disease. J Clin Lab Anal, 2005. 19(3): p. 99-

102.

81. Lau, A.C., et al., Elastolytic matrix metalloproteinases and coronary outcome in children

with Kawasaki disease. Pediatr Res, 2007. 61(6): p. 710-5.

82. Lepidi, S., et al., MMP9 production by human monocyte-derived macrophages is

decreased on polymerized type I collagen. J Vasc Surg, 2001. 34(6): p. 1111-8.

83. Oviedo-Orta, E., et al., Comparison of MMP-2 and MMP-9 secretion from T helper 0, 1

and 2 lymphocytes alone and in coculture with macrophages. Immunology, 2008. 124(1): p. 42-

50.

84. Gavin, P.J., et al., Systemic arterial expression of matrix metalloproteinases 2 and 9 in

acute Kawasaki disease. Arterioscler Thromb Vasc Biol, 2003. 23(4): p. 576-81.

85. Takeshita, S., et al., Elevated serum levels of matrix metalloproteinase-9 (MMP-9) in

Kawasaki disease. Clin Exp Immunol, 2001. 125(2): p. 340-4.

86. Lau, A.C., et al., Inhibition of matrix metalloproteinase-9 activity improves coronary

outcome in an animal model of Kawasaki disease. Clin Exp Immunol, 2009. 157(2): p. 300-9.

87. Lau, A.C., et al., Matrix metalloproteinase 9 activity leads to elastin breakdown in an

animal model of Kawasaki disease. Arthritis Rheum, 2008. 58(3): p. 854-63.

88. Llewelyn, M. and J. Cohen, Superantigens: microbial agents that corrupt immunity.

Lancet Infect Dis, 2002. 2(3): p. 156-62.

89. Choi, Y., J.W. Kappler, and P. Marrack, A superantigen encoded in the open reading

frame of the 3' long terminal repeat of mouse mammary tumour virus. Nature, 1991. 350(6315):

p. 203-7.

90. Choi, Y., P. Marrack, and J.W. Kappler, Structural analysis of a mouse mammary tumor

virus superantigen. J Exp Med, 1992. 175(3): p. 847-52.

81

91. Acharya, K.R., et al., Structural basis of superantigen action inferred from crystal

structure of toxic-shock syndrome toxin-1. Nature, 1994. 367(6458): p. 94-7.

92. Fields, B.A., et al., Crystal structure of a T-cell receptor beta-chain complexed with a

superantigen. Nature, 1996. 384(6605): p. 188-92.

93. Swaminathan, S., et al., Crystal structure of staphylococcal enterotoxin B, a superantigen.

Nature, 1992. 359(6398): p. 801-6.

94. White, J., et al., The V beta-specific superantigen staphylococcal enterotoxin B:

stimulation of mature T cells and clonal deletion in neonatal mice. Cell, 1989. 56(1): p. 27-35.

95. Hudson, K.R., H. Robinson, and J.D. Fraser, Two adjacent residues in staphylococcal

enterotoxins A and E determine T cell receptor V beta specificity. J Exp Med, 1993. 177(1): p.

175-84.

96. Alber, G., D.K. Hammer, and B. Fleischer, Relationship between enterotoxic- and T

lymphocyte-stimulating activity of staphylococcal enterotoxin B. J Immunol, 1990. 144(12): p.

4501-6.

97. O'Hehir, R.E. and J.R. Lamb, Induction of specific clonal anergy in human T

lymphocytes by Staphylococcus aureus enterotoxins. Proc Natl Acad Sci U S A, 1990. 87(22): p.

8884-8.

98. Fleischer, B. and H. Schrezenmeier, T cell stimulation by staphylococcal enterotoxins.

Clonally variable response and requirement for major histocompatibility complex class II

molecules on accessory or target cells. J Exp Med, 1988. 167(5): p. 1697-707.

99. Herrmann, T., et al., Activation of MHC class I-restricted CD8+ CTL by microbial T cell

mitogens. Dependence upon MHC class II expression of the target cells and V beta usage of the

responder T cells. J Immunol, 1990. 144(4): p. 1181-6.

100. Kappler, J., et al., V beta-specific stimulation of human T cells by staphylococcal toxins.

Science, 1989. 244(4906): p. 811-3.

82

101. MacDonald, H.R., et al., Peripheral T-cell reactivity to bacterial superantigens in vivo:

the response/anergy paradox. Immunol Rev, 1993. 133: p. 105-17.

102. Renno, T., M. Hahne, and H.R. MacDonald, Proliferation is a prerequisite for bacterial

superantigen-induced T cell apoptosis in vivo. J Exp Med, 1995. 181(6): p. 2283-7.

103. Kawabe, Y. and A. Ochi, Programmed cell death and extrathymic reduction of Vbeta8+

CD4+ T cells in mice tolerant to Staphylococcus aureus enterotoxin B. Nature, 1991. 349(6306):

p. 245-8.

104. Gonzalo, J.A., et al., Expansion and clonal deletion of peripheral T cells induced by

bacterial superantigen is independent of the interleukin-2 pathway. Eur J Immunol, 1992. 22(4):

p. 1007-11.

105. Lenardo, M.J., Interleukin-2 programs mouse alpha beta T lymphocytes for apoptosis.

Nature, 1991. 353(6347): p. 858-61.

106. Mountz, J.D., et al., Increased susceptibility of fas mutant MRL-lpr/lpr mice to

staphylococcal enterotoxin B-induced septic shock. J Immunol, 1995. 155(10): p. 4829-37.

107. Miethke, T., et al., T cell-mediated lethal shock triggered in mice by the superantigen

staphylococcal enterotoxin B: critical role of tumor necrosis factor. J Exp Med, 1992. 175(1): p.

91-8.

108. Attinger, A., H.R. MacDonald, and H. Acha-Orbea, Lymphoid environment limits

superantigen and antigen-induced T cell proliferation at high precursor frequency. Eur J

Immunol, 2001. 31(3): p. 884-93.

109. MacDonald, H.R., S. Baschieri, and R.K. Lees, Clonal expansion precedes anergy and

death of V beta 8+ peripheral T cells responding to staphylococcal enterotoxin B in vivo. Eur J

Immunol, 1991. 21(8): p. 1963-6.

110. Rellahan, B.L., et al., In vivo induction of anergy in peripheral V beta 8+ T cells by

staphylococcal enterotoxin B. J Exp Med, 1990. 172(4): p. 1091-100.

83

111. Baschieri, S., et al., Clonal anergy to staphylococcal enterotoxin B in vivo: selective

effects on T cell subsets and lymphokines. Eur J Immunol, 1993. 23(10): p. 2661-6.

112. Gutierrez-Ramos, J.C., I. Moreno de Alboran, and C. Martinez, In vivo administration of

interleukin-2 turns on anergic self-reactive T cells and leads to autoimmune disease. Eur J

Immunol, 1992. 22(11): p. 2867-72.

113. Lavoie, P.M., et al., Understanding the mechanism of action of bacterial superantigens

from a decade of research. Immunol Rev, 1999. 168: p. 257-69.

114. Attinger, A., H. Acha-Orbea, and H.R. MacDonald, Cutting edge: cell autonomous rather

than environmental factors control bacterial superantigen-induced T cell anergy in vivo. J

Immunol, 2000. 165(3): p. 1171-4.

115. Migita, K., et al., Defective TCR-mediated signaling in anergic T cells. J Immunol, 1995.

155(11): p. 5083-7.

116. Sundstedt, A., et al., In vivo anergized CD4+ T cells express perturbed AP-1 and NF-

kappa B transcription factors. Proc Natl Acad Sci U S A, 1996. 93(3): p. 979-84.

117. Webb, S., C. Morris, and J. Sprent, Extrathymic tolerance of mature T cells: clonal

elimination as a consequence of immunity. Cell, 1990. 63(6): p. 1249-56.

118. McCormack, J.E., et al., Profound deletion of mature T cells in vivo by chronic exposure

to exogenous superantigen. J Immunol, 1993. 150(9): p. 3785-92.

119. Rocha, B., C. Tanchot, and H. Von Boehmer, Clonal anergy blocks in vivo growth of

mature T cells and can be reversed in the absence of antigen. J Exp Med, 1993. 177(5): p. 1517-

21.

120. Pape, K.A., et al., Direct evidence that functionally impaired CD4+ T cells persist in vivo

following induction of peripheral tolerance. J Immunol, 1998. 160(10): p. 4719-29.

121. Tanchot, C., et al., Adaptive tolerance of CD4+ T cells in vivo: multiple thresholds in

response to a constant level of antigen presentation. J Immunol, 2001. 167(4): p. 2030-9.

84

122. Gonzalo, J.A., et al., Glucocorticoid-mediated control of the activation and clonal

deletion of peripheral T cells in vivo. J Exp Med, 1993. 177(5): p. 1239-46.

123. Leung, D.Y., et al., Prevalence of superantigen-secreting bacteria in patients with

Kawasaki disease. J Pediatr, 2002. 140(6): p. 742-6.

124. Nomura, Y., et al., Possible relationship between streptococcal pyrogenic exotoxin A and

Kawasaki syndrome in patients older than six months of age. Pediatr Infect Dis J, 2003. 22(9): p.

794-8.

125. Yoshioka, T., et al., Quantitative analysis of the usage of human T cell receptor alpha and

beta chain variable regions by reverse dot blot hybridization. J Immunol Methods, 1997. 201(2):

p. 145-55.

126. Abe, J., et al., Selective expansion of T cells expressing T-cell receptor variable regions

V beta 2 and V beta 8 in Kawasaki disease. Proc Natl Acad Sci U S A, 1992. 89(9): p. 4066-70.

127. Abe, J., et al., Characterization of T cell repertoire changes in acute Kawasaki disease. J

Exp Med, 1993. 177(3): p. 791-6.

128. Curtis, N., et al., Evidence for a superantigen mediated process in Kawasaki disease.

Arch Dis Child, 1995. 72(4): p. 308-11.

129. Nomura, Y., et al., Twenty-five types of T-cell receptor Vbeta family repertoire in

patients with Kawasaki syndrome. Eur J Pediatr, 1998. 157(12): p. 981-6.

130. Reichardt, P., et al., Analysis of T-cell receptor V-beta 2 in peripheral blood lymphocytes

as a diagnostic marker for Kawasaki disease. Infection, 2002. 30(6): p. 360-4.

131. Yoshioka, T., et al., Polyclonal expansion of TCRBV2- and TCRBV6-bearing T cells in

patients with Kawasaki disease. Immunology, 1999. 96(3): p. 465-72.

132. Yoshioka, T., et al., Relation of streptococcal pyrogenic exotoxin C as a causative

superantigen for Kawasaki disease. Pediatr Res, 2003. 53(3): p. 403-10.

85

133. Leung, D.Y., et al., Evidence for superantigen involvement in cardiovascular injury due

to Kawasaki syndrome. J Immunol, 1995. 155(10): p. 5018-21.

134. Matsubara, K., et al., Development of serum IgM antibodies against superantigens of

Staphylococcus aureus and Streptococcus pyogenes in Kawasaki disease. Clin Exp Immunol,

2006. 143(3): p. 427-34.

135. Yeung, R.S.M., The etiology of Kawasaki disease: a superantigen-mediated process.

Prog. Pediatr. Cardiol., 2004. 19: p. 109-113.

136. Onouchi, Z., et al., Coronary artery aneurysms develop in weanling rabbits with serum

sickness but not in mature rabbits. An experimental model for Kawasaki disease in humans.

Angiology, 1995. 46(8): p. 679-87.

137. Onouchi, Z., et al., Comparative study of the coronary arterial involvement in infant and

adult rabbits with experimental allergic angiitis. Prog Clin Biol Res, 1987. 250: p. 529-30.

138. Felsburg, P.J., et al., Immunologic abnormalities in canine juvenile polyarteritis

syndrome: a naturally occurring animal model of Kawasaki disease. Clin Immunol

Immunopathol, 1992. 65(2): p. 110-8.

139. Philip, S., et al., A swine model of horse serum-induced coronary vasculitis: an

implication for Kawasaki disease. Pediatr Res, 2004. 55(2): p. 211-9.

140. Murata, H., Experimental candida-induced arteritis in mice. Relation to arteritis in the

mucocutaneous lymph node syndrome. Microbiol Immunol, 1979. 23(9): p. 825-31.

141. Murata, H., et al., The pathogenesis of experimental arteritis induced by Candida alkali-

extract in mice. Jpn J Exp Med, 1987. 57(6): p. 305-13.

142. Ohkuni, H., et al., Coronary arteritis in mice after systemic injection of bacterial cell wall

peptidoglycan. Jpn Circ J, 1987. 51(12): p. 1357-61.

143. Lehman, T.J., et al., Coronary arteritis in mice following the systemic injection of group

B Lactobacillus casei cell walls in aqueous suspension. Arthritis Rheum, 1985. 28(6): p. 652-9.

86

144. Lehman, T.J., et al., Variable expression of Lactobacillus casei cell wall-induced

coronary arteritis: an animal model of Kawasaki's disease in selected inbred mouse strains. Clin

Immunol Immunopathol, 1988. 48(1): p. 108-18.

145. Takahashi, K., et al., Histopathological features of murine systemic vasculitis caused by

Candida albicans extract--an animal model of Kawasaki disease. Inflamm Res, 2004. 53(2): p.

72-7.

146. Myones, B.L., J.M Bathoria, T.J. A. Lehman, and S.T. Shulman., Human IVIG inhibits

Lactobacillus casei-inducible coronary arteritis in a murine model., in The 5th International

Kawasaki Disease Symposium. 1995, Ensevier Science: Fukuoka, Japan.

147. Hui-Yuen, J.S., T.T. Duong, and R.S. Yeung, TNF-alpha is necessary for induction of

coronary artery inflammation and aneurysm formation in an animal model of Kawasaki disease.

J Immunol, 2006. 176(10): p. 6294-301.

148. Chan, W.C., T.T. Duong, and R.S. Yeung, Presence of IFN-gamma does not indicate its

necessity for induction of coronary arteritis in an animal model of Kawasaki disease. J Immunol,

2004. 173(5): p. 3492-503.

149. Okitsu-Negishi, S., et al., The induction of cardioangitis by Lactobacillus casei cell wall

in mice. I. The cytokine production from murine macrophages by Lactobacillus casei cell wall

extract. Clin Immunol Immunopathol, 1996. 78(1): p. 30-40.

150. Schulte, D.J., et al., Involvement of innate and adaptive immunity in a murine model of

coronary arteritis mimicking Kawasaki disease. J Immunol, 2009. 183(8): p. 5311-8.

151. Duong, T.T., et al., Superantigenic activity is responsible for induction of coronary

arteritis in mice: an animal model of Kawasaki disease. Int Immunol, 2003. 15(1): p. 79-89.

152. Bette, M., et al., Distribution and kinetics of superantigen-induced cytokine gene

expression in mouse spleen. J Exp Med, 1993. 178(5): p. 1531-9.

153. Lin, C.Y., et al., Cytokines predict coronary aneurysm formation in Kawasaki disease

patients. Eur J Pediatr, 1993. 152(4): p. 309-12.

87

154. Bruns, P., Die Heilwirkung des Erysipels auf Geschwulste. Beitr. Klin. Chir., 1868. 3: p.

443-446.

155. Shear, M.J.T., F.C. , Chemical treatment of tumors. V. Isolation of the hemorrhage-

producing fraction from Serratia marcescens (Bacillus prodigiosus) culture filtrate. J. Natl.

Cancer Inst., 1943. 4: p. 81-97.

156. O'Malley, W.E.A., B. Shear, M.J., Action of bacterial polysaccharide on tumors. II.

Damage of Sarcoma 37 by serum of mice treated with Serratia marcescens polysaccharide, and

induced tolerance. J. Natl Cancer Inst., 1962. 29: p. 1169-1175.

157. Pennica, D., et al., Human tumour necrosis factor: precursor structure, expression and

homology to lymphotoxin. Nature, 1984. 312(5996): p. 724-9.

158. Aggarwal, B.B., Signalling pathways of the TNF superfamily: a double-edged sword. Nat

Rev Immunol, 2003. 3(9): p. 745-56.

159. Bebo, B.F., Jr. and D.S. Linthicum, Expression of mRNA for 55-kDa and 75-kDa tumor

necrosis factor (TNF) receptors in mouse cerebrovascular endothelium: effects of interleukin-1

beta, interferon-gamma and TNF-alpha on cultured cells. J Neuroimmunol, 1995. 62(2): p. 161-

7.

160. Botchkina, G.I., et al., Expression of TNF and TNF receptors (p55 and p75) in the rat

brain after focal cerebral ischemia. Mol Med, 1997. 3(11): p. 765-81.

161. Al-Lamki, R.S., et al., TNF receptors differentially signal and are differentially expressed

and regulated in the human heart. Am J Transplant, 2009. 9(12): p. 2679-96.

162. Vancheri, C., et al., Different expression of TNF-alpha receptors and prostaglandin E(2

)Production in normal and fibrotic lung fibroblasts: potential implications for the evolution of the

inflammatory process. Am J Respir Cell Mol Biol, 2000. 22(5): p. 628-34.

163. Al-Lamki, R.S., et al., Expression of tumor necrosis factor receptors in normal kidney

and rejecting renal transplants. Lab Invest, 2001. 81(11): p. 1503-15.

88

164. Walter, U., et al., Monitoring gene expression of TNFR family members by beta-cells

during development of autoimmune diabetes. Eur J Immunol, 2000. 30(4): p. 1224-32.

165. Ryffel, B., et al., Tumour necrosis factor receptor distribution in human lymphoid tissue.

Immunology, 1991. 74(3): p. 446-52.

166. Austgulen, R., et al., Expression of receptors for tumor necrosis factor in human placenta

at term. Acta Obstet Gynecol Scand, 1992. 71(6): p. 417-24.

167. Forest, V., et al., Apc+/Min colonic epithelial cells express TNF receptors and ICAM-1

when they are co-cultured with large intestine intra-epithelial lymphocytes. Cell Immunol, 2003.

223(1): p. 70-6.

168. MacEwan, D.J., TNF receptor subtype signalling: differences and cellular consequences.

Cell Signal, 2002. 14(6): p. 477-92.

169. Weiss, T., et al., TNFR80-dependent enhancement of TNFR60-induced cell death is

mediated by TNFR-associated factor 2 and is specific for TNFR60. J Immunol, 1998. 161(6): p.

3136-42.

170. Chan, F.K., et al., A domain in TNF receptors that mediates ligand-independent receptor

assembly and signaling. Science, 2000. 288(5475): p. 2351-4.

171. Hsu, H., et al., TNF-dependent recruitment of the protein kinase RIP to the TNF receptor-

1 signaling complex. Immunity, 1996. 4(4): p. 387-96.

172. Chinnaiyan, A.M., et al., FADD, a novel death domain-containing protein, interacts with

the death domain of Fas and initiates apoptosis. Cell, 1995. 81(4): p. 505-12.

173. Eggermont, A.M., et al., The role of isolated limb perfusion for melanoma confined to the

extremities. Surg Clin North Am, 2003. 83(2): p. 371-84, ix.

174. Feinberg, B., et al., A phase I trial of intravenously-administered recombinant tumor

necrosis factor-alpha in cancer patients. J Clin Oncol, 1988. 6(8): p. 1328-34.

89

175. Kuprash, D.V., et al., Novel tumor necrosis factor-knockout mice that lack Peyer's

patches. Eur J Immunol, 2005. 35(5): p. 1592-600.

176. Kuprash, D.V., et al., TNF and lymphotoxin beta cooperate in the maintenance of

secondary lymphoid tissue microarchitecture but not in the development of lymph nodes. J

Immunol, 1999. 163(12): p. 6575-80.

177. Ratner, A. and W.R. Clark, Role of TNF-alpha in CD8+ cytotoxic T lymphocyte-

mediated lysis. J Immunol, 1993. 150(10): p. 4303-14.

178. Kuwano, K., T. Kawashima, and S. Arai, Antiviral effect of TNF-alpha and IFN-gamma

secreted from a CD8+ influenza virus-specific CTL clone. Viral Immunol, 1993. 6(1): p. 1-11.

179. Adams, A.B., et al., The role of TNF receptor and TNF superfamily molecules in organ

transplantation. Am J Transplant, 2002. 2(1): p. 12-8.

180. Mackay, F., et al., BAFF AND APRIL: a tutorial on B cell survival. Annu Rev Immunol,

2003. 21: p. 231-64.

181. Rothe, J., et al., Mice lacking the tumour necrosis factor receptor 1 are resistant to TNF-

mediated toxicity but highly susceptible to infection by Listeria monocytogenes. Nature, 1993.

364(6440): p. 798-802.

182. Bolger, A.P. and S.D. Anker, Tumour necrosis factor in chronic heart failure: a peripheral

view on pathogenesis, clinical manifestations and therapeutic implications. Drugs, 2000. 60(6):

p. 1245-57.

183. Diwan, A., et al., Inflammatory mediators and the failing heart: a translational approach.

Curr Mol Med, 2003. 3(2): p. 161-82.

184. Romas, E., M.T. Gillespie, and T.J. Martin, Involvement of receptor activator of

NFkappaB ligand and tumor necrosis factor-alpha in bone destruction in rheumatoid arthritis.

Bone, 2002. 30(2): p. 340-6.

185. Micheau, O. and J. Tschopp, Induction of TNF receptor I-mediated apoptosis via two

sequential signaling complexes. Cell, 2003. 114(2): p. 181-90.

90

186. Lu, X., S. Nemoto, and A. Lin, Identification of c-Jun NH2-terminal protein kinase

(JNK)-activating kinase 2 as an activator of JNK but not p38. J Biol Chem, 1997. 272(40): p.

24751-4.

187. Tobiume, K., et al., ASK1 is required for sustained activations of JNK/p38 MAP kinases

and apoptosis. EMBO Rep, 2001. 2(3): p. 222-8.

188. Song, H.Y., M. Rothe, and D.V. Goeddel, The tumor necrosis factor-inducible zinc finger

protein A20 interacts with TRAF1/TRAF2 and inhibits NF-kappaB activation. Proc Natl Acad

Sci U S A, 1996. 93(13): p. 6721-5.

189. De Smaele, E., et al., Induction of gadd45beta by NF-kappaB downregulates pro-

apoptotic JNK signalling. Nature, 2001. 414(6861): p. 308-13.

190. Papa, S., et al., Gadd45 beta mediates the NF-kappa B suppression of JNK signalling by

targeting MKK7/JNKK2. Nat Cell Biol, 2004. 6(2): p. 146-53.

191. McCormack, J.E., J. Kappler, and P. Marrack, Stimulation with specific antigen can

block superantigen-mediated deletion of T cells in vivo. Proc Natl Acad Sci U S A, 1994. 91(6):

p. 2086-90.

192. Moolani Y, D.T., Yeung RSM., The Role of Co-stimulation in Sustaining the Immune

Reponse in Kawasaki Disease. Arthritis and Rheumatism, 2008. 57:S502.

193. Vella, A.T., et al., Lipopolysaccharide interferes with the induction of peripheral T cell

death. Immunity, 1995. 2(3): p. 261-70.

194. Vella, A.T., et al., CD28 engagement and proinflammatory cytokines contribute to T cell

expansion and long-term survival in vivo. J Immunol, 1997. 158(10): p. 4714-20.

195. Gross, J.A., E. Callas, and J.P. Allison, Identification and distribution of the

costimulatory receptor CD28 in the mouse. J Immunol, 1992. 149(2): p. 380-8.

196. Smith-Garvin, J.E., G.A. Koretzky, and M.S. Jordan, T cell activation. Annu Rev

Immunol, 2009. 27: p. 591-619.

91

197. Coyle, A.J., et al., The CD28-related molecule ICOS is required for effective T cell-

dependent immune responses. Immunity, 2000. 13(1): p. 95-105.

198. Hutloff, A., et al., ICOS is an inducible T-cell co-stimulator structurally and functionally

related to CD28. Nature, 1999. 397(6716): p. 263-6.

199. Suh, W.K., et al., The inducible costimulator plays the major costimulatory role in

humoral immune responses in the absence of CD28. J Immunol, 2004. 172(10): p. 5917-23.

200. Brand, O., S. Gough, and J. Heward, HLA , CTLA-4 and PTPN22 : the shared genetic

master-key to autoimmunity? Expert Rev Mol Med, 2005. 7(23): p. 1-15.

201. Waterhouse, P., et al., Lymphoproliferative disorders with early lethality in mice

deficient in Ctla-4. Science, 1995. 270(5238): p. 985-8.

202. Felix, N.J., et al., Targeting lymphocyte co-stimulation: from bench to bedside.

Autoimmunity, 2010. 43(7): p. 514-25.

203. Linsley, P.S., et al., Immunosuppression in vivo by a soluble form of the CTLA-4 T cell

activation molecule. Science, 1992. 257(5071): p. 792-5.

204. Brunet, J.F., et al., A new member of the immunoglobulin superfamily--CTLA-4. Nature,

1987. 328(6127): p. 267-70.

205. Li, D., et al., Cutting edge: Cbl-b: one of the key molecules tuning CD28- and CTLA-4-

mediated T cell costimulation. J Immunol, 2004. 173(12): p. 7135-9.

206. Bachmaier, K., et al., Negative regulation of lymphocyte activation and autoimmunity by

the molecular adaptor Cbl-b. Nature, 2000. 403(6766): p. 211-6.

207. Samelson, L.E., Signal transduction mediated by the T cell antigen receptor: the role of

adapter proteins. Annu Rev Immunol, 2002. 20: p. 371-94.

208. Yamasaki, S., et al., Lck-independent triggering of T-cell antigen receptor signal

transduction by staphylococcal enterotoxins. J Biol Chem, 1997. 272(23): p. 14787-91.

92

209. Morgan, M.M., et al., Superantigen-induced T cell:B cell conjugation is mediated by

LFA-1 and requires signaling through Lck, but not ZAP-70. J Immunol, 2001. 167(10): p. 5708-

18.

210. Bavari, S. and R.G. Ulrich, Staphylococcal enterotoxin A and toxic shock syndrome

toxin compete with CD4 for human major histocompatibility complex class II binding. Infect

Immun, 1995. 63(2): p. 423-9.

211. Oyaizu, N., et al., Superantigen staphylococcal enterotoxin B-induced T-helper cell

activation is independent of CD4 molecules and phosphatidylinositol hydrolysis. Proc Natl Acad

Sci U S A, 1992. 89(17): p. 8035-9.

212. Bueno, C., et al., Bacterial superantigens bypass Lck-dependent T cell receptor signaling

by activating a Galpha11-dependent, PLC-beta-mediated pathway. Immunity, 2006. 25(1): p. 67-

78.

213. Boise, L.H., et al., CD28 costimulation can promote T cell survival by enhancing the

expression of Bcl-XL. Immunity, 1995. 3(1): p. 87-98.

214. Adams, J.M. and S. Cory, The Bcl-2 protein family: arbiters of cell survival. Science,

1998. 281(5381): p. 1322-6.

215. Hockenbery, D., et al., Bcl-2 is an inner mitochondrial membrane protein that blocks

programmed cell death. Nature, 1990. 348(6299): p. 334-6.

216. Fraser, J.D., et al., Regulation of interleukin-2 gene enhancer activity by the T cell

accessory molecule CD28. Science, 1991. 251(4991): p. 313-6.

217. Lindstein, T., et al., Regulation of lymphokine messenger RNA stability by a surface-

mediated T cell activation pathway. Science, 1989. 244(4902): p. 339-43.

218. Kabelitz, D. and S. Wesselborg, Life and death of a superantigen-reactive human CD4+

T cell clone: staphylococcal enterotoxins induce death by apoptosis but simultaneously trigger a

proliferative response in the presence of HLA-DR+ antigen-presenting cells. Int Immunol, 1992.

4(12): p. 1381-8.

93

219. Groux, H., et al., CD3-mediated apoptosis of human medullary thymocytes and activated

peripheral T cells: respective roles of interleukin-1, interleukin-2, interferon-gamma and

accessory cells. Eur J Immunol, 1993. 23(7): p. 1623-9.

220. Marrack, P. and J. Kappler, Control of T cell viability. Annu Rev Immunol, 2004. 22: p.

765-87.

221. Watts, T.H., Staying alive: T cell costimulation, CD28, and Bcl-xL. J Immunol, 2010.

185(7): p. 3785-7.

222. Blankson, J.N. and S.S. Morse, The CD28/B7 pathway costimulates the response of

primary murine T cells to superantigens as well as to conventional antigens. Cell Immunol, 1994.

157(1): p. 306-12.

223. Sansom, D.M., et al., B7/CD28 but not LFA-3/CD2 interactions can provide 'third-party'

co-stimulation for human T-cell activation. Immunology, 1993. 80(2): p. 242-7.

224. Krummel, M.F., T.J. Sullivan, and J.P. Allison, Superantigen responses and co-

stimulation: CD28 and CTLA-4 have opposing effects on T cell expansion in vitro and in vivo.

Int Immunol, 1996. 8(4): p. 519-23.

225. Lane, P., C. Haller, and F. McConnell, Evidence that induction of tolerance in vivo

involves active signaling via a B7 ligand-dependent mechanism: CTLA4-Ig protects V beta 8+ T

cells from tolerance induction by the superantigen staphylococcal enterotoxin B. Eur J Immunol,

1996. 26(4): p. 858-62.

226. Muraille, E., et al., B7.2 provides co-stimulatory functions in vivo in response to

staphylococcal enterotoxin B. Eur J Immunol, 1995. 25(7): p. 2111-4.

227. Chiang, Y.J., et al., Cbl-b regulates the CD28 dependence of T-cell activation. Nature,

2000. 403(6766): p. 216-20.

228. Tartaglia, L.A. and D.V. Goeddel, Two TNF receptors. Immunol Today, 1992. 13(5): p.

151-3.

94

229. Hildeman, D.A., et al., Activated T cell death in vivo mediated by proapoptotic bcl-2

family member bim. Immunity, 2002. 16(6): p. 759-67.

230. Moolani Y, Y.R., The Role of Costimulation in the Persistence of the Immune Response

in Kawasaki Disease, in Institute of Medical Science. 2009, University of Toronto: Toronto. p.

100.

231. Kuroda, K., et al., Implantation of IL-2-containing osmotic pump prolongs the survival of

superantigen-reactive T cells expanded in mice injected with bacterial superantigen. J Immunol,

1996. 157(4): p. 1422-31.

232. Lenschow, D.J., et al., Expression and functional significance of an additional ligand for

CTLA-4. Proc Natl Acad Sci U S A, 1993. 90(23): p. 11054-8.

233. Freeman, G.J., et al., Murine B7-2, an alternative CTLA4 counter-receptor that

costimulates T cell proliferation and interleukin 2 production. J Exp Med, 1993. 178(6): p. 2185-

92.

234. Hathcock, K.S., et al., Comparative analysis of B7-1 and B7-2 costimulatory ligands:

expression and function. J Exp Med, 1994. 180(2): p. 631-40.

235. Watts, T.H., TNF/TNFR family members in costimulation of T cell responses. Annu Rev

Immunol, 2005. 23: p. 23-68.

236. Bouillet, P. and L.A. O'Reilly, CD95, BIM and T cell homeostasis. Nat Rev Immunol,

2009. 9(7): p. 514-9.

237. Gorak-Stolinska, P., D.M. Kemeny, and A. Noble, Activation-induced cell death in

human T cells is a suicidal process regulated by cell density but superantigen induces T cell

fratricide. Cell Immunol, 2002. 219(2): p. 98-107.

238. Strasser, A. and M. Pellegrini, T-lymphocyte death during shutdown of an immune

response. Trends Immunol, 2004. 25(11): p. 610-5.

239. Beutler, B. and F. Bazzoni, TNF, apoptosis and autoimmunity: a common thread? Blood

Cells Mol Dis, 1998. 24(2): p. 216-30.

95

240. Jacob, C.O., Tumor necrosis factor alpha in autoimmunity: pretty girl or old witch?

Immunol Today, 1992. 13(4): p. 122-5.

241. Held, W., et al., Genes encoding tumor necrosis factor alpha and granzyme A are

expressed during development of autoimmune diabetes. Proc Natl Acad Sci U S A, 1990. 87(6):

p. 2239-43.

242. Peppel, K., D. Crawford, and B. Beutler, A tumor necrosis factor (TNF) receptor-IgG

heavy chain chimeric protein as a bivalent antagonist of TNF activity. J Exp Med, 1991. 174(6):

p. 1483-9.

243. Brown, G.R., et al., Lymphoid hyperplasia, CD45RBhigh to CD45RBlow T-cell

imbalance, and suppression of Type I diabetes mellitus result from TNF blockade in NOD--

>NOD-scid adoptive T cell transfer. Diabetologia, 1998. 41(12): p. 1502-10.

244. Jacob, C.O., et al., Prevention of diabetes in nonobese diabetic mice by tumor necrosis

factor (TNF): similarities between TNF-alpha and interleukin 1. Proc Natl Acad Sci U S A,

1990. 87(3): p. 968-72.

245. Heilig, B., et al., Elevated TNF receptor plasma concentrations in patients with

rheumatoid arthritis. Clin Investig, 1992. 70(1): p. 22-7.

246. Brennan, F.M., et al., Enhanced expression of tumor necrosis factor receptor mRNA and

protein in mononuclear cells isolated from rheumatoid arthritis synovial joints. Eur J Immunol,

1992. 22(7): p. 1907-12.

247. Maini, R.N., et al., Monoclonal anti-TNF alpha antibody as a probe of pathogenesis and

therapy of rheumatoid disease. Immunol Rev, 1995. 144: p. 195-223.

248. Kuiper, S., et al., Different roles of tumour necrosis factor alpha and interleukin 1 in

murine streptococcal cell wall arthritis. Cytokine, 1998. 10(9): p. 690-702.

249. Mori, L., et al., Attenuation of collagen-induced arthritis in 55-kDa TNF receptor type 1

(TNFR1)-IgG1-treated and TNFR1-deficient mice. J Immunol, 1996. 157(7): p. 3178-82.

96

250. Brahn, E., et al., Effects of tumor necrosis factor alpha (TNF-alpha) on collagen arthritis.

Lymphokine Cytokine Res, 1992. 11(5): p. 253-6.

251. Ray, A., B.S. Bal, and B.K. Ray, Transcriptional induction of matrix metalloproteinase-9

in the chondrocyte and synoviocyte cells is regulated via a novel mechanism: evidence for

functional cooperation between serum amyloid A-activating factor-1 and AP-1. J Immunol,

2005. 175(6): p. 4039-48.

252. Yoon, J.W., H.S. Jun, and P. Santamaria, Cellular and molecular mechanisms for the

initiation and progression of beta cell destruction resulting from the collaboration between

macrophages and T cells. Autoimmunity, 1998. 27(2): p. 109-22.

253. Yang, X.D., et al., Effect of tumor necrosis factor alpha on insulin-dependent diabetes

mellitus in NOD mice. I. The early development of autoimmunity and the diabetogenic process.

J Exp Med, 1994. 180(3): p. 995-1004.