<p>Toll-like receptor (TLR) signaling pathways have been demonstrated to be important in many aspects of innate and adaptive immunity. Binding of TLRs by their respective ligands initiates signaling cascades which promote a number of cellular responses including, but not limited to, pathogen recognition, co-stimulation, cell maturation and activation and initiation of adaptive immune responses. While many roles for TLRs have been rigorously tested, the exact contributions of these receptors to the quality of B-cell immune responses remain unclear. Specifically, a role for intracellular TLR signaling in B-cell development is one area that remains largely uncharacterized. Endogenous TLRs that recognize viral and microbial DNAs and RNAs are also capable of recognizing self-nucleic acids. B cells, which express TLRs, undergo tolerizing mechanisms in the bone marrow (BM) and periphery to eliminate self-reactive clones. I hypothesize that endogenous TLR recognition of self-nucleic acids during B cell development promotes the elimination of these autoreactive cells from the B cell repertoire. </p><p>TLRs are also known to be integral in the subsequent triggering of the adaptive immune response. A series of controversial studies has attempted to determine whether TLR signaling is required for antibody responses to thymus-dependent protein antigens. Initial reports indicated that MyD88 was required for antibody responses to a native protein antigen. However, later studies using haptenated protein antigens did not confirm a requirement for TLR signaling in adjuvant-enhanced antibody responses. In an effort to resolve these discordant results, it was suggested that haptenated protein antigens, unlike native proteins, are fundamentally distinct in that they do not require MyD88 signaling. I hypothesize that antibody and germinal center (GC) responses to native and haptenated proteins are independent of MyD88 signaling. I further propose that the unique immunogenicity ascribed to haptenated proteins is simply immunodominance of hapten, a phenomena observed and described almost 100 years ago. </p><p>Deficiency in the global adaptor protein for TLR signaling, myeloid differentiation primary response protein 88 (MyD88), through which all TLRs signal (with the exception of TLR 3), effectively silences the TLR pathway. Utilizing mice deficient in MyD88, I show that TLR signaling through MyD88 mediates central tolerance of B cells. Specifically, crossing the MyD88 deficiency onto a 3H9 autoreactive heavy-chain knock-in rescues the immature (imm) and transitional 1(T1) B cell compartments in the BM. This finding demonstrates the requirement for MyD88 signaling in the removal of autoreactive B cell clones at the first tolerance checkpoint. </p><p> I also find that MyD88 signaling is not required for antibody, GC or memory responses to native or haptenated proteins. Additionally, affinity maturation, determined by BCR mutation frequency in GC was comparable between MyD88 deficient and sufficient mice. Both MyD88 deficient and sufficient mice were able to elicit secondary immune responses to native and haptenated proteins. Furthermore, I demonstrate that the MyD88 independent immunogenicity attributed to haptenated protein is a misinterpretation of the established concept of immunodominance- haptenated proteins elicited hapten-specific responses that were approx. 20- to 100-fold greater than to the carrier. Regardless of MyD88 signaling, native proteins elicited significantly less serum Ab than their haptenated forms.</p><p>I conclude that TLR signaling through MyD88 mediates tolerance during B cell development in the BM, but is not required for B cell immune responses.</p> / Dissertation

Rod-shaped bacteria were previously shown to be associated with the small intestinal epithelium of children with celiac disease (CD). Using culture-dependent and independent methods, we characterized the microbiota of small intestine in children with CD and controls. The normal microbiota constitutes an unique organ-specific biofilm. Dominant bacteria are Streptococcus, Neisseria, Veillonella, Gemella, Actinomyces, Rothia and Haemophilus. Altogether 162 Genus Level Operational Taxonomic Units (GELOTU) of six different phyla were identified in a total of 63 children. In biopsies collected during 2004- 2007 we did not find major differences in the microbiota between CD patients and controls. However, in biopsies collected earlier from children born during the “Swedish CD epidemic” and demonstrated to have rod-shaped bacteria by electron microscopy, we found that unclassified-Clostridales and Prevotella species were associated with CD. These anaerobic, rod-shaped bacteria showed marked affinity for the intestinal epithelium. Changes in breast-feeding practice and/or regiments for introduction of gluten containing food probably affect the composition of the bacterial flora in small intestine. We hypotesize that these bacteria contribute to contraction of CD. An in vitro model for studies of immune mechanisms of the intestinal epithelium was established. Polarized tight monolayers of the human colon carcinoma cell lines, T84 and Caco2, were developed by culture in a two-chamber system. The two cell lines showed the features of mature- and immature columnar epithelial cells respectively. Polarized monolayers were challenged with bacteria and proinflammatory cytokines. Immune responses were estimated as quantitative changes in mRNA expression levels of a secreted mucin (MUC2), glycocalyx components (CEACAMs, MUC3), antimicrobial factors and cytokines (IFN-g, TNF-a, IL-6 and IL-8). Tight monolayer cells were more resistant to bacterial attack than ordinary tissue culture cells and only B. megaterium induced the defensin, hBD2. Tight monolayer cells responded to cytokine challenge suggesting awareness of basolateral attack. TNF-a induced markedly increased levels of IL-8 and TNF- a itself in both cell lines suggesting recruitment and activation of immune cells. Cytokine challenge also increased levels of CEACAM1, which includes two functionally different forms, CEACAM1-L and CEACAM1-S. In T84 cells, IFN-g was selective for CEACAM1- L while TNF-a upregulated both forms. Increased CEACAM1 expression may influence epithelial function and communication between epithelial cells and intraepithelial lymphocytes. As a pathogenic enteric bacterium, Vibrio cholerae secretes cholera toxin that is the major factor of cholera diarrhea. However, some strains of O1 serogroup lacking the cholera toxin still cause enterocolitis and most V. cholerae vaccines candidates exhibit reactogenicity in clinical trails. An extracellular metalloprotease PrtV was characterized. It was associated with killing of bacteria predators such as the nematode Caenorhabditis elegans. Its role in human intestine was addressed by using the T84 tight monolayer in vitro model. We found that Vibrio Cholera Cytolysin (VCC), a pore-forming toxin, induces an inflammatory response in intestinal epithelial cells that includes increased epithelial permeability and induction of IL-8 and TNF-a and hence could be responsible for enterocolitis. The inflammatory response was abolished by PrtV thus VCC is indeed an autologous substrate for PrtV. In protein rich environment PrtV degradation of VCC was inhibited, suggesting that the magnitude of the inflammatory response is modulated by the milieu in the small intestine. Thus, VCC is likely to be part of the pathogenesis of cholera diarrhea and the causative agent of enteropathy in V. cholerae strains lacking the cholera toxin.

<p>The unique functional attributes possessed by T cells are initiated after stimulation via their T cell receptor (TCR). Due to the T cell's integral role in immune function, the regulation of signaling events downstream of the TCR have been, and continue to be, an area of intense research. An effective T cell immune response is dependent upon the proper relaying of the signals initiated by the TCR-antigen-MHC interaction, whereas a disruption in the signaling cascades downstream of the TCR can result in the inability to control pathogen replication or the initiation of T cell mediated autoimmunity. The discovery of specific pharmacological inhibitors and the generation of genetically modified mouse models have allowed investigators to take a stepwise approach in understanding TCR induced signaling hierarchies. </p><p>In this study, we have utilized genetically modified mouse models, in which the targets of gene deletion, TSC1 and DGK, are molecules that regulate signals emanating from the TCR. Additionally, we have demonstrated that both of these molecules negatively regulate the mammalian target of rapamycin (mTOR) in both naïve and activated T cells. mTOR is a critical regulator of cell growth and metabolism. By virtue of its kinase ability, mTOR has been shown to initiate signaling in response to a variety of extracellular signals including cytokines, growth factors, amino acids, and toll-like receptor (TLR) ligands. The versatility with which this protein kinase interprets multiple signaling inputs simultaneously has led to mTOR being referred to as the "rheostat" of the cell. Recent investigation has begun to elucidate the role of the mTOR pathway in basic T cell biology, however, the mechanisms by which mTOR controls basic T cell homeostasis and function is unclear, and the importance of tight control of the TSC-mTOR pathway in T cells is not known.</p><p> In the first model, mice were strategically bred so that TSC1 was deleted exclusively in the murine T cell lineage. TSC1, in complex with TSC2, acts as an inhibitor of mTOR by inhibiting Rheb, an activator of mTORC1. Using this model we found that deletion of TSC1 at the double-positive (DP) stage of thymocyte development has several profound effects on T cell signaling, homeostasis and survival. Specifically, the loss TSC1 in T cells results in constitutive activation of mTORC1 and decreased activation of mTORC2. Reduced T cell numbers were also observed in the peripheral lymphoid organs, which correlated with the finding of increased cell death ex vivo as well as after TCR stimulation in vitro. Furthermore, we found that TSC1 deficiency resulted in altered mitochondrial homeostasis and function, which could be rescued in vitro with co-stimulation and/or antioxidants. These observations give us clear evidence that the TSC-mTOR pathway regulates T cell survival and normal mitochondrial homeostasis. </p><p> In the second model, I utilized diacylglycerol kinase (DGK) deficient mice in our examination of the mechanisms by which the TCR affects T cell biology. To date, ten DGK isoforms have been identified in mammals, with DGK&#945; and DGK&#950; being expressed in T cells. Immediately following TCR stimulation, PIP2 is hydrolyzed into the secondary messengers inostitol triphosphate (IP3) and diacylglycerol (DAG). By converting DAG into phosphatidic acid, DGKs effectively terminate signaling mediated by DAG and serve to dampen T cell activation. Additionally, previous work from our lab has shown that mTOR kinase activity is negatively regulated in T cells by DGKs, and reveals a novel signaling relationship between the TCR, DGKs, and mTOR kinase activation. Given the ability of DGK&#945; and &#950; to regulate mTOR and other signaling cascades, we hypothesized that DGK activity may play an important role during an anti-viral immune response. Using DGK&#945;- and DGK&#950;-deficient (germline) mice in conjunction with MHC class I restricted tetramers and synthetically generated viral antigens, we were able to enumerate the primary and memory CD8+ anti-viral immune response in the absence of diacylglycerol (DAG) metabolism and enhanced mTOR activity. In response to LCMV infection, DGK-deficient CD8+ T cells expand more aggressively and produce elevated amounts of anti-viral cytokines, which results in reduced viral titers in DGK-deficient mice 7 days after infection. Additionally, we found that while DGK activity serves to suppress the CD8+ T cell response during the primary infection phase, it promotes the expansion of antigen specific CD8+ T cells during the memory phase of an immune response. The diminished response by DGK-deficient memory CD8+ T cells highlights opposing roles for DAG metabolism during the primary and memory immune phases. </p><p> The studies reported in this dissertation provide novel insights into the intrinsic mechanisms that regulate T cell homeostasis and function.</p> / Dissertation

MUC1 has conventionally been studied as an epithelial cell surface molecule. Its glycosylation and expression change when those cells are transformed into adenocarcinomas. These changes have led to focus on MUC1 as a tumor antigen and also its role in adhesion to blood vessels and signaling within the tumor cell. The recent discovery that T cells also express MUC1 on their surface extends the physiological role of MUC1, with the possibility that functions observed in tumors may be reproduced on T cells. Expression of MUC1 on T cells was first characterized in terms of timing, location and structure. T cells activated both in vivo and in vitro express MUC1. Expression in vitro is maintained over long time periods as the T cell population acquires the memory phenotype. Activated T cells induced to polarize by inflammatory conditions focus MUC1 expression to their leading edge, the sensory compartment of polarized T cells. Reactivity with glycosylation-sensitive antibodies and induction of glycosyltransferases indicates that the glycosylation of MUC1 on T cells is similar to that on normal epithelial cells. A MUC1-negative T cell line was transfected with MUC1 cDNA and used as a model to investigate consequences of MUC1 expression on the T cell surface. Interaction of MUC1+ T cells with resting or activated endothelial cells revealed that MUC1 aids in adhesion under both normal and inflammatory conditions. Analysis of interactions with individual adhesion molecules demonstrated MUC1 specific enhancement of binding to ICAM-1 but inhibition of binding to E-selectin. Phosphorylation of the MUC1 intracytoplasmic tail is constitutive but decreases upon interaction with activated endothelium. MUC1 expression on T cells is also associated with differential phosphorylation of proteins in the molecular weight ranges of 39 kDa, ~80 kDa and 190 kDa, with the ~80 kDa band identified as beta-catenin. While human T cells express MUC1 on their surface upon activation, this does not appear to be a characteristic of mouse T cells from the human MUC1 transgenic mouse model. However, as recent work indicates that mouse T cells express mouse Muc-1 after activation, human and mouse T cells may similarly depend on MUC1 for normal functioning.

Herpes simplex virus type-1 (HSV-1) infection results in the establishment of a latent infection in sensory ganglia of the peripheral nervous system (PNS). In humans the virus can sporadically reactivate and gain access to the primary sites of infection and cause considerable tissue damage. The mechanisms involved in establishment and maintenance of latency and the sporadic reactivation of HSV-1 infected neurons in sensory ganglia is poorly defined. Elucidating the mechanisms involved in maintenance of latency of HSV-1 would be extremely valuable in controlling the high prevalence of disease in communities around the world. We have shown that CD8+ T lymphocytes can block HSV-1 reactivation from latency in sensory neurons of the trigeminal ganglion ex vivo. The mechanism that the CD8+ T cells employ to prevent reactivation includes release of the antiviral cytokine IFNã, but other mechanisms, such as the lytic effect of perforin and granzymes appear to be important as well. In order to further understand the cellular immune response against the latent HSV infection we have attempted to dissect the CD8+ T cell immune response in the TG during acute and latent infection. Specifically, we determined the antigen specificity and TCR expression of the CD8+ T cells that accumulate in the TG at various times post infection. We have investigated the state of CD8+ T cell memory after HSV-1 infection and clarified the mechanisms used by CD8+ T cells to prevent reactivation from latency.

Abstract To generate an effective cellular immune response, it is necessary to elicit both antigen-specific CTL (CD8+) and Th (CD4+) T cell recognition. The design of an effective therapeutic vaccine must incorporate means by which to generate novel T cell responses or enhance existing responses of a clinically-preferred functional type. In the cancer setting, one vaccine strategy is to target the immune system to specifically recognize tumor-associated antigens (TAAs). TAAs can be sub-categorized in many ways (i.e. onco-viral, mutated self-proteins, overexpressed and fetal-like proteins). Thus far, in part based on the historical prioritization applied to epitope searches, the number of defined CTL epitopes greatly outnumbers that of Th epitopes. Our goal in the current studies was to define and then characterize functional CD4+ T cell responses directed against the HPV-16 E7 oncoprotein in patients with cervical intraepithelial neoplasia (CIN) or cancer using dendritic cell (DC)-based vaccine strategies. The importance of this work stems from the unequivocal linkage between oncogenic HPV-infection and the development of cervical carcinoma. Over 50% of all cervical carcinomas are HPV-16 positive, making it the most salient HPV type for integration into therapeutic vaccine designs. For cellular transformation to occur and be maintained, expression of the HPV early region gene products E6 and E7 is mandatory. Due to this unique requirement for sustained expression of the E6 and E7 proteins in transformed cells, these proteins make excellent candidates for protective or therapeutic vaccinations. While numerous HPV-16 E7-derived CTL epitopes have been identified over the past several years, surprisingly, only a single Th epitope has been reported thus far. In this thesis, I have defined three novel, naturally-processed and -presented epitopes derived from the HPV-16 E7 oncoprotein that are recognized by CD4+ T helper cells in patients with cervical intraepithelial neoplasia (CIN) or cervical carcinoma. Since the functional polarization state of the E7-specific CD4+ T cells remains of the Th1-type until the development of cancer in situ in these patients, DC-based vaccines that include E7-derived peptides or the whole E7 protein, and which are capable of selectively maintaining or enhancing Type-1 immunity may prove clinically beneficial in preventing or treating HPV-16+ malignancies, including cervical cancer.

Murine CD8alpha+ dendritic cells (DC) are antigen-presenting cells with tolerogenic properties, including ability to prolong allograft survival. Little is known, however, regarding their migratory ability, either in vitro or in vivo. Limited work to date has yielded inconsistent findings, and potential impediments to use of this subset to promote tolerance induction are reported discrepancies regarding their ability to home to secondary lymphoid tissue. Despite functional differences from classic CD8alpha- DC, the two subsets exhibit similar phenotypes and commonly coexist in the same tissues, suggesting expression of similar homing molecules/receptors. The central hypothesis underlying these studies was that CD8alpha+ and CD8alpha- DC have equal ability to respond to migration-inducing factors, both in vitro and in vivo. Our aim was to determine the factors and potential therapeutic targets that regulate CD8alpha+ DC migration. Herein we have employed an in vitro chemotaxis assay for the determination of which, if any, CC chemokines specifically regulate the migration of CD8alpha+ (and CD8alpha-) murine spleen DC. We also used this assay with the addition of endothelial cell layers, to assess the adhesion molecules that facilitate DC transendothelial migration. Two chemokines (CCL19 and CCL21) elicited CD54-dependent migration of both mature DC subsets, in the presence or absence of endothelial cells, but CD8alpha+ DC migrated in consistently fewer numbers than CD8alpha- DC in vitro. Our findings led us to investigate the importance of the in vitro migration-inducing chemokines in vivo. In these studies we compared the ability of DC subsets to migrate to T cell areas of wild type and specific chemokine-deficient mice. Unlike our in vitro results, the DC subsets migrated with equal efficiency in normal recipients. After sc injection into CCL19/CC21-deficient mice, CD8alpha- DC trafficked more efficiently than CD8alpha+ DC to draining lymphoid tissue. The necessity of specific chemokine-directed DC migration in vivo for alloimmune responses was borne out in transplant studies in which impaired migration of DC resulted in significant prolongation of murine cardiac allograft survival. These studies demonstrate, for the first time, potential chemokine and adhesion molecule targets for manipulation of murine DC subset migration in vivo.

In order to generate and monitor effective and specific immune responses against tumors, a clear understanding of relevant tumor antigens and their derivative epitopes recognized by T lymphocytes is warranted. The characterization of tumor antigen epitopes recognized by T lymphocytes has been a major focus of study over the past decade. Both CD8+ and CD4+ T lymphocytes contribute to the immune response against tumors, and the determination of the epitopes they recognize is necessary for their incorporation into immunotherapy protocols for cancer. The tumor antigens recognized by T lymphocytes fall into 3 major categories: Tumor-specific (TSA), Cancer Testis (CT), and Tumor Associated Antigens (TAA). Our goal in the following studies was to characterize a novel TAA, EphA2, since this protein has been linked to metastasis in numerous cancer settings. The definition of epitopes seen by T lymphocytes will assist in vaccine strategies for immunotherapy protocols against EphA2+ tumors. In the following studies, I have defined 8 novel EphA2 T cell epitopes (5 HLA-A2 restricted and 3 HLA-DR4 restricted) recognized by CD8+ and CD4+ T lymphocytes, respectively. The anti-EphA2 CD4+ functional response was skewed based on the presence of disease or increased staging of RCC disease, with patients with active disease exhibiting a Th2-biased CD4+ response. I have also linked the expression of EphA2 in primary RCC tumors to the time to recurrence in patients affected with RCC. Furthermore, I have demonstrated that the cell surface expression of EphA2 on tumors can be modulated using EphA2 agonists. This agonist treatment results in the enhanced recognition of EphA2+ tumors by specific CTLs. With reports of the overexpression of protein phosphatases (PPs) in several cancer settings, we discovered the EphA2 was constitutively underphosphorylated in certain cancer cell lines, likely as the consequence of overexpressed PP activity. Finally, I have shown that by neutralizing the activity of cellular phosphatases utilizing phosphatase inhibitors, that we can induce the phosphorylation of EphA2 and its subsequent degradation via a largely proteasome-dependent pathway. As a result, this thesis has defined a novel tumor antigen, EphA2, and demonstrated the possibility that modulation of its expression in tumor cells may result in increased recognition by specific T effector cells that may be germane to the design of improved and efficacious therapies for the treatment of patients with EphA2+ tumors.

One of the main challenges facing tumor immunologists is to develop strategies that would effectively stimulate Type-1 anti-tumor T cell responses, which have been correlated with better clinical outcome and prolonged survival of cancer patients. As CD4+ T cells were shown to play a critical role in mediating these responses, it was of interest to examine novel ways of effectively stimulating and enhancing Type-1 CD4+ T cell responses. For these studies I used MAGE-A6, a tumor associated antigen (TAA) expressed by a broad range of human cancer types. Two novel MAGE-A6 T-helper epitopes were identified and were shown to be recognized by CD4+ T cells isolated from the majority of normal donors or patients with melanoma, regardless of their HLA genotype (i.e. poly-DR presented epitopes). Furthermore, peptide-specific T cells also recognized autologous monocytes pulsed with recombinant MAGE-A6 protein, supporting the natural processing and MHC presentation of these epitopes. Interestingly, one of the novel MAGE-A6 epitopes possesses a high-degree of homology with a microbial peptide. CD4+ T cells stimulated in vitro with this microbial peptide cross-reacted against the MAGE-A6 homologue peptide, and could recognize naturally-processed MAGE-A6 epitopes more effectively than T cells stimulated with MAGE-A6 peptides. This study showed that it is possible to stimulate, and even enhance tumor-specific T cell responses using microbial epitopes that are homologous to TAA-derived peptides. In the final study, human dendritic cells (DC) were engineered to secrete high levels of IFN-×-inducing cytokines IL-12p70 and IL-18 via recombinant adenoviral infection to generate an in vitro stimulus capable of promoting previously deficient patient Th1-type responses. DC engineered to secrete both of these cytokines simultaneously (DC.IL-12/18) were highly effective at stimulating MAGE-A6-specific Th1-type CD4+ T cell responses from patients with melanoma, particularly when loaded with MAGE-A6 protein. Poly-DR presented epitopes and MAGE-A6 protein defined in this thesis, if loaded onto DC.IL-12/18, could prove clinically useful as a vaccine modality capable of promoting the recovery and/or enhancement of tumor antigen-specific, Th1-type CD4+ T cell responses in the majority of patients harboring MAGE-A6+ cancers.

Patients with inflammatory bowel disease (IBD), a chronic inflammatory disease of the colon, have an increased incidence of colon cancer. This has led to the hypothesis that chronic inflammation causes malignant transformation and promotes tumor progression. However, an alternative hypothesis can be made that everything starts with early malignant lesions, which activate innate but not adaptive immunity thus driving chronic inflammation. This imbalance between the innate and the adaptive immunity at the intestinal site may speed up colon cancer progression. To test this hypothesis we are examining development of colonic inflammation and associated colon cancer from the perspective of de novo expression of the tumor antigen MUC1 in both settings and innate and adaptive immune responses against it. We have created an animal model that recapitulates de novo MUC1 expression in human IBD by crossing IL10-/- mice that develop IBD and colon cancer, with human MUC1 transgenic mice that express MUC1 under its own promoter, thereby maintaining human tissue specific expression of this molecule. Mice were sacrificed at various time points and colonic tissue sections assessed for inflammatory and malignant changes and MUC1 expression. We found that, like in humans, expression of normal MUC1 as well as hypoglycosylated (tumor) MUC1 increases with the severity of inflammation in IBD. In other experiments, MUC1+/IL10-/- mice were vaccinated with TnMUC100mer, representing the hypoglycosylated (tumor) form of MUC1. MUC1-specific vaccination slows the progression to IBD as measured by rectal prolapse. Vaccinated animals, that develop rectal prolapse, have fewer tumors than unvaccinated animals. We have developed an animal model of MUC1+ IBD and colon cancer that mimics human disease. We show that MUC1-specific vaccination slows the progression to IBD and has a protective anti-tumor effect. We postulate that induction of MUC1-specific immunity, including effector and regulatory T-cells, restores the balance between adaptive and innate immunity, which resolves chronic inflammation and stops progression of premalignant lesions to cancer.