The IL-12 family of heterodimeric cytokines, comprising IL-12, IL-23 and IL-27, is an integral component of the inflammatory response. The IL-12 p40 subunit, which is part of both IL-12 and IL-23, is expressed specifically in macrophages and dendritic cells in response to microbial stimuli which signal through toll-like receptors (TLRs) and nucleotide-oligomerization-binding domain (NOD) proteins. Dysregulated expression of IL-12 p40 could lead to prolonged, unresolved inflammation manifesting into chronic inflammatory disorders such as inflammatory bowel disease (IBD). Understandably, IL-12 p40 expression is tightly regulated. We have demonstrated the requirement of a complex comprising nuclear factor of activated T cells (NFAT) and interferon regulatory factor 8 (IRF8) in IL-12 p40 gene transcriptional regulation. Subsequently, IRF8 was shown to be a target for an important anti-inflammatory pathway activated by carbon monoxide (CO) and heme oxygenase-1 (HO-1) in murine IBD. This dissertation explored two molecular mechanisms of IL-12 p40 inhibition targeting the transcription factor NFAT using a novel cell-permeable inhibitory peptide and phosphoinositide 3-kinase (PI3K) mediated inhibition of IL-12. Firstly, VIVIT peptide which prevents the nuclear translocation of NFAT was employed to examine the role of NFAT in macrophage gene expression. We observed that NFAT inhibition attenuated the expression of inflammatory cytokines including IL-12, IL-23 and TNF. Secondly, we studied PI3K mediated inhibition of IL-12 and characterized immune responses and macrophage defects in PI3K p110δ mutant mice which spontaneously develop IBD. We observed that the PI3K mutant mice recapitulate certain aspects of human IBD, including a profound increase in several proinflammatory cytokines in the intestinal mucosa and in macrophages, such as IL-12, IL-23, IL-17, TNF and IFN-γ. Furthermore, macrophages from PI3K p110δ mutant mice were defective in IL-10 and C5a mediated inhibition of IL-12 p40. Thus, the expression of proinflammatory cytokines is coordinately regulated by transcription factors (such as NFAT and IRF8) and signaling molecules (such as PI3K), and mechanisms limiting inflammation are crucial for maintenance of immune homeostasis. Manipulation of such inhibitory pathways is a potential therapeutic approach for treating chronic inflammatory disorders.

Rheumatoid arthritis (RA) is an autoimmune disease affecting nearly 1% of the population. The joints of RA patients become inflamed and a pannus of synovial tissue invades cartilage and destroys bone. Current treatments, though moderately successful, are often administered systemically and are associated with significant side effects. Local treatment of arthritis by intra-articular injection of gene transfer vectors has the potential to overcome some of the obstacles of other treatment regimens. In particular, adeno-associated virus (AAV) is attractive as a gene transfer vector because it is non-pathogenic in humans, has low immunogenicity and broad tissue tropism, and provides for long-term transgene expression. The primary target of gene transfer in the joint is the fibroblast-like synoviocytes (FLS), which make up part of the pannus of tissue responsible for cartilage and bone destruction. However, FLS are only minimally transduced by AAV and provide only low levels of transgene expression. For local treatment with AAV vectors to be successful, improvement in the transduction of FLS is necessary. Several agents have been shown to increase AAV transduction in other cell types. The specific effect of inflammation and proteasome inhibition on AAV-mediated transgene expression in FLS has not been examined. Here we show that both inflammatory cytokines and the proteasome inhibitor zLLL can increase transgene expression in FLS dramatically. Remarkably, both agents are also able to regulate transgene expression. The cytokines appear to act in part by promoting uncoating of the virus, while proteasome inhibition works by both increasing the nuclear trafficking of the virus and by increasing transcription of the viral DNA. In addtion, we screened 3 serotypes of AAV (2, 2.5, and 5) for their ability to transduce FLS from 4 different species: human, mouse, horse, and rabbit. AAV2 provided the highest transgene expression in human and horse FLS, while AAV2.5 was better suited for mouse and rabbit FLS. These results have implications in the future use of AAV vectors in the treatment of RA.

The human chronic inflammatory bowel diseases (IBD), Crohns disease (CD) and ulcerative colitis (UC) are ostensibly disorders of innate immunity with an exaggerated inflammatory response and loss of tolerance to the normal enteric microbial flora. In this project, we have extensively characterized innate immune responses driven by Pathogen Associated Molecular Pattern Molecules (PAMPs) and the more recently recognized Damage Associated Molecular Pattern molecules (DAMPs). The prototype DAMP, a chromatin-associated protein, high mobility group box 1 (HMGB1), is released during cellular necrosis and is secreted from activated macrophages. Extracellularly, it binds the receptor for advanced glycation end products (RAGE), as well as toll-like receptor (TLR) 2 and TLR4, important in the recognition of PAMPs. PAMPs and DAMPs trigger inflammatory signaling pathways in neighboring cells through activation of the transcription factor family, NF-kappaB. Much attention has been given to the central role played by PAMPs in the form of the enteric bacterial flora in IBD pathogenesis. We hypothesize that DAMPs also play a pivotal role in this process. Accordingly, we have determined the significance of DAMPs and PAMPs in the mucosal inflammatory response in macrophages and in vivo in mouse models of IBD. We first investigated expression of TLRs in the gut to determine cell types in the intestinal epithelium that may respond to danger signals. TLR expression was most prominent on intestinal epithelial enteroendocrine cells (EEC). Using a murine EEC line, multiple functional consequences of TLR activation were demonstrated. Second, in IL-10 deficient (-/-) mice with chronic Th1-mediated enterocolitis, we demonstrate a role for HMGB1 in macrophage activation and IBD. Lastly, we examined an in vivo therapy targeted at inhibiting the prominent downstream effector of DAMP and PAMP signaling, NF-kappaB, in murine IBD. Inhibition of activated NF-kappaB with a short cell permeable peptide inhibited chronic enterocolitis in IL-10-/- mice. In summary, this dissertation provides new insight into our understanding of intestinal innate mucosal inflammatory responses. We demonstrate the relevance of TLRs on EECs and the contribution of DAMP and PAMP signaling in disease. These results also provide proof of concept for new therapeutic approaches in IBD.

An elusive objective for tumor immunologists has been the development of an effective tumor vaccine capable of inducing potent immune responses to eliminate established tumors and induce long-term protective antitumor immunity. Dendritic cells (DCs) are potent inducers of immunity and represent a promising tool for the purpose of immune-based tumor therapies. However, DC-based vaccines have enjoyed only limited success in clinical trials, probably due to the use of immature/intermediate mature DCs that maintain tolerance during the steady state, or to the use of non-polarized DCs which lack the proper cytokine production that favors cellular immune responses needed to eliminate established tumors. The failure of past tumor vaccines demonstrates a need to examine and enhance immunization strategies on multiple levels. The underlying hypothesis for these studies was that combining a DC1 polarization signal with an effective antigen-loading strategy will result in enhanced tumor immunotherapy. Our first aim was to compare cytosolic and membrane-bound antigen presentation of tumor-derived proteins by DCs following three different antigen-loading strategies; coculture of DCs and tumor cells, feeding DCs with tumor lysate, and fusion of DCs and tumor cells. We demonstrated that both DC-tumor coculture and fusion result in a higher level of tumor-derived peptide presentation compared to feeding DCs with tumor lysate. Our second aim was to develop a murine DC1 polarization model to evaluate DC1-based tumor immunotherapy. Herein, we described the synergistic affect of TLR3 and TLR9 ligation on IL-12p70 production by murine DCs, characterizing the timing and exhaustion of IL-12p70 production. Furthermore, we examined the ability of polarized DCs to stimulate T cell proliferation and cytokine secretion in response to a model antigen in vitro. For our third aim we examined the capacity of DC1s to stimulate immune responses to a model antigen as well as native tumor antigens in vivo and tested the therapeutic effect of tumor-loaded DC1 vaccines. These studies demonstrate the ability of antigen-loaded polarized DCs to induce strong Th1-mediated anti-tumor immunity characterized by tumor infiltrating CD4+ T cells and macrophages, but not CD8+ T cells, resulting in tumor growth inhibition.

The incidence of immunopathology is increasing; and treatments usually involve systemic immunosuppression, with significant side-effects. Regulatory T cells (Tregs) inhibit immune responses in a targeted manner, and are being studied as potential therapeutic agents. Interleukin 2 (IL-2) is among the few well-characterized modulators of Tregs. IL-2 stimulation increases Treg function, but inhibits suppression in co-cultures with T helper (Th) cells, due to dominant effects on Th cells. Characterizing cytokine modulation of Tregs is important because Treg-targeted therapy would be used primarily to treat diseases that induce cytokine production. Recent work has implicated IL-6 in the regulation of Tregs in the lung and in the pathogenesis of several diseases. We investigated the influence of IL-2 and IL-6 on Tregs, and found that they increased Treg suppressive function, proliferation, and expression of FoxP3 and CTLA4. Interestingly, both cytokines are regulated by suppressor of cytokine signaling 3 (SOCS3), and we found that unstimulated Tregs lacked SOCS3, whereas naïve Th cells expressed it abundantly. SOCS3 over-expression in Tregs inhibited proliferation, FoxP3 and CTLA4 expression, and suppressive function. Whereas IL-2 and SOCS3 seemed to act during homeostatic conditions, IL-4 is active in disease states. IL-4 is required for experimental asthma induction in mice due to its critical role in the development of Th2 cells, which protect against helminth infections. Consequently, we analyzed regulation of Tregs by IL-4 in vivo during allergic airway inflammation. Pulmonary Treg numbers were similar in tolerized mice compared to those with inflammation, and adoptive transfer of Tregs inhibited inflammation, implying that cytokines might have inhibited Treg function. However, IL-4Rα-/- Tregs, which lack the IL-4 receptor alpha chain, were less effective in reducing inflammation. In vitro, IL-4 increased Treg proliferation and maintenance of FoxP3 expression, did not alter Treg suppressive function, and increased Th cell resistance to suppression. Also, IL-4-mediated effects on Tregs and Th cells required signal transducer and activator of transcription 6 (STAT6). Therefore, IL-4 increased Treg function in vivo by enhancing Treg proliferation through a STAT6-dependent mechanism. In summary, SOCS3 may serve as a future therapeutic agent, whereas the net effect of IL-4 blockade therapy remains unclear.

CD8+ T cell responses are crucial for immunity against intracellular infections and can mediate tumor regression. While CD8+ T cells are widely recognized as cytolytic effector cells (cytolytic T cells; CTLs), little is known about their immunoregulatory functions and their impact on dendritic cells (DCs). A similar area of controversy is the role of DC in regulating the induction of CD8+ T cell effector functions and CD8+ T cell memory. This dissertation addresses the impact of bidirectional communication between DCs and CD8+ T cells, during different phases of the immune response, upon the functions of both these cell types. In order to reconcile the apparently contrasting notions that CD8+ T cells perform both suppressor and helper functions, I compared the DC-modulating activity of CD8+ T cells at different stages of activation. I observed that DC-killing and DC-activating (and protective) functions are exerted sequentially by activated CD8+ T cells. In contrast to the effector cells that kill DCs in a granzyme B/perforin-dependent manner, memory CD8+ T cells promote IL-12 production in DCs and support CD4+ and CD8+ T cell responses. Moreover, memory CD8+ T cells instruct DC to over-express granzyme B inhibitor PI-9, protecting them from elimination by CTLs. I observed that the inclusion of heterologous CD8+ T cell epitopes in cancer vaccines, promoting the interaction of vaccine-bearing DCs with large numbers of tumor-unrelated CD8+ T cells, strongly enhances the immunologic and therapeutic activity of vaccination against established tumors that are resistant to standard vaccines. Since the character of the vaccination-induced CD8+ T cells is important for the efficacy of cancer immunotherapy, I have analyzed the role of DCs in influencing the cytolytic function and peripheral tissue-homing ability of CD8+ T cells. I observed that short-term-activated inflammatory-type DCs, capable of producing high levels of IL-12 and other pro-inflammatory cytokines, support induction of cytotoxic function and a switch from lymphoid to peripheral chemokine receptors in CD8+ T cells. In contrast, exhausted DCs matured for extended periods of time or matured under the influence of the mediators of chronic inflammation, favor CD8+ T cell expansion alone without acquisition of effector functions. Collectively, the findings presented in this dissertation broaden our understanding of the feedback circuitry between CD8+ T cells and DCs and will help us to design improved vaccines against cancer and chronic infections.

The TIM proteins belong to a novel family of molecules contained within a single gene locus positioned on mouse chromosome 11B1.1 and human chromosome 5q33.2- a locus that has been repeatedly linked with susceptibility to atopy and autoimmune diseases. Early evidences from genetic, epidemiological and immune functional studies have also implicated these proteins in the regulation of immune responses associated with asthma and autoimmunity. However, at the time of commencing these studies there was a paucity of information regarding the underlying molecular basis of TIM function that directed these different effector responses. Based on existing information on putative signaling motifs contained within the TIM protein sequences and early evidence of TIM-2 tyrosine phosphorylation we proposed that TIM-1 and TIM-2 could transduce intra-cellular biochemical signals in response to ligation of the receptor, by coupling to phosphotyrosine dependent signaling mechanisms in order to regulate T cell activation. Employing a combination of biochemical, pharmacological and genetic approaches, our studies establish TIM-1 and TIM-2 proteins as -signal transducing- cell surface receptors. We show that, TIM-2 functions as a negative regulator of T cell activation by inhibiting NFAT/AP-1 dependent transcription. In contrast, TIM-1 can provide a co-stimulatory signal for T cell activation, and augment cytokine and NFAT/AP-1 dependent transcription.TIM-1- mediated signal transduction requires the cytoplasmic tail and the conserved tyrosine 276 contained within the tail. Furthermore, TIM-1 can amplify classical TCR signaling cascades for NFAT/AP-1 dependent transcription and this occurs in a manner that requires the TCR and contribution of most of the key signaling components elicited by the TCR. TIM-1 can also trigger distinct pathways that involve Lck-dependent tyrosine phosphorylation followed by recruitment and activation of p85-PI3K for up-regulation of surface markers associated with T cell activation. In this context, TIM-1 requires Akt to enhance NFAT/AP-1 dependent transcription. Eventually, elucidation of the biochemical signals underlying the mechanistic function of a family of molecules significantly involved in the regulation of T helper cell responses would present targets for therapeutic modulation of Th1 and Th2- type immunity in health and in immune-mediated disease.

Approximately one third of the worlds population is infected with Mycobacterium tuberculosis, which was responsible for about 1.6 million deaths in 2005. In spite of continuing advances in understanding host response to this infection, generation and maintenance of the host immune response remains unclear. In this thesis, we investigate molecules involved in the generation and maintenance of the host response, specifically the granuloma, to M. tuberculosis. We investigated the role of TNF antagonists in reactivation of tuberculosis, and showed that while anti-TNF antibody is superior to TNFR2-Fc fusion molecule in penetrating the granuloma, any blockade of TNF compromises control of acute tuberculosis. We hypothesized that TNF is required for priming T cell responses and that TNF-inducible chemokine receptors function redundantly, allowing one chemokine to compensate in the absence of another. Here, we show that TNF is not required to prime the adaptive immune response, and that TNF-inducible chemokines CXCR3 and CCR5 are simultaneously expendable, refuting the compensation hypothesis in these two chemokines. Reports have implicated unexplored inflammatory molecules in host response to M. tuberculosis infection. We hypothesized that the small chemotactic molecule LTB4 and its receptor BLT1 increase pathology during M. tuberculosis infection. We also hypothesized that osteopontin is required for mediating an effective immune response to tuberculosis by mediating Th1 priming and lymphocyte migration. We show here that neither BLT1 nor osteopontin play a significant role in the inflammatory response to M. tuberculosis. Finally, we investigated the role of ICAM-1 in priming effector and regulatory T cells in response to tuberculosis. We report that ICAM-1 is dispensable for priming and migration of effector T cells, but that ICAM-1 is required for production of inducible Foxp3+ T regulatory cells via TGFâ1 stimulation. We hypothesize that the reduction in T regulatory cells exacerbates the immune response, allowing greater inflammation in the lungs, potentially causing overwhelming inflammation. This body of work contributes to the understanding of the host response to tuberculosis by investigating activity of cytokines, chemotactic molecules and adhesion molecules in balancing the host response to M. tuberculosis infection.

Despite successes in animals, cytokine gene expression selectively in human tumors is difficult to achieve due to lack of efficient delivery methods. We previously demonstrated that adoptively transferred, IL-2 activated natural killer (A-NK) cells are very effective in trafficking to, infiltrating and, subsequently, reducing B16 lung tumors in tumor-bearing animals. We therefore speculated that the tumor-seeking A-NK cells could be used for the delivery of cytokines selectively to the tumor microenvironment. However, tumor infiltration by A-NK cells depends on high and often toxic doses of IL-2 to support the transferred A-NK cells survival and anti-tumor functions. To address this problem, we hypothesized that A-NK cells transduced to express pro-NK cell cytokines would become less dependent on high and potentially toxic amounts of IL-2. Assessments of transduction efficiency in vitro demonstrated that adenoviral transduction consistently resulted in high (>60%) transduction rates and substantial expression of transgenes such as GFP, luciferase, and mIL-12 for at least 4 days. In vivo experiments illustrated that mIL-12 transduced A-NK cells localized and survived significantly better than mock transduced cells within lung metastases than in the surrounding normal lung tissue when supported with low, non-toxic doses of IL-2. The intratumoral survival of non-transduced bystander A-NK cells also increased when they were co-injected with IL-12 gene-transduced A-NK cells. The enhanced survival of exogenously delivered, IL-12 gene-transduced A-NK cells resulted in greater anti-tumor responsiveness. This led to a 7-10 day increase in median survival time compared to tumor-bearing mice receiving mock-transduced A-NK cells. These data show that the presence of IL-12 around tumor-infiltrating A-NK cells enhances their anti-tumor activity while reducing their requirement for systemically administered IL-2. This A-NK cell delivered IL-12 has lead to an enhanced host anti-tumor reactivity, which appeared to be mediated through cytokine involvement, namely IFNγ, rather than T, B, and NK cellular activity. Thus, adoptive transfer of A-NK cells represents an efficient method for targeting products of genes to tumor sites and eliciting anti-tumor responses.

HSV-1 establishes latency in sensory neurons of the trigeminal ganglia (TG) following corneal infection. The concept that latently infected neurons are ignored by the host immune response has given way to the notion that CD8 T cells maintained in the TG monitor infected neurons thereby subverting reactivation. The tendency of HSV-1 to periodically reactivate in humans and mediate recurrent disease is associated with significant morbidity. A desire to understand the complex interactions of this pathogen, the neurons that harbor it, and the immune system that monitors latency define this study. Two populations of CD8 T cells rapidly infiltrate the TG coincident with resolution of replicating virus and juxtapose with neurons for the life of the mouse. One population recognizes the immunodominant glycoprotein B (gB) epitope while the other does not (gB-nonspecific). We establish that the homeostatic cytokine IL-15 does not contribute to the maintenance of gB-specific or gB-nonspecific CD8 T cells within the TG during latency. However, IL-15 is crucial for the regulation of gB-specific memory CD8 T cells in noninfected tissues. These findings led us to question whether gB-nonspecific CD8 T cells are important in the HSV-1 response. We demonstrate that gB-nonspecific CD8 T cells upregulate the effector molecule granzyme B, and produce IFNã and proliferate in response to HSV-1-infected but not gB-transfected targets. This data conclusively shows that gB-nonspecific CD8 T cells in the infected TG are HSV-1 specific. This population is also capable of preventing reactivation following explant of latent TG. Contrary to their gB-specific CD8 counterparts, gB-nonspecific CD8 T cells have a reduced capacity to produce IFNã during latency and this reduction in function is associated with increased expression of PD-1. Surprisingly, blockade of PD-L1 did not rescue effector function yet increased the viral burden during latency. We show that a population of neurons expressing PD-L1 contains an enriched reservoir of HSV-1 latency that is highly prone to reactivate.