Search

Memory CD8 T cells protect against intracellular pathogens by scanning host cell surfaces, thus infection detection rates depend on memory cell number and distribution. Many cell population analyses rely on isolation from whole organs and interpretation is predicated on presumptions of near-complete cell recovery. Paradigmatically, T cell memory is parsed into central, effector, and resident subsets, ostensibly defined by immunosurveillance patterns, but in practice identified by phenotypic markers. Because isolation methods and subsequent phenotypic marker-based analyses ultimately inform models of memory T cell differentiation, protection, and vaccine translation, we tested their validity via quantitative immunofluorescence microscopy of a murine memory CD8 T cell population. We found that lymphocyte isolation fails to recover most cells and recovery is biased against certain subsets. Applying this approach to parabiotic mice we found that the overwhelming majority of memory CD8 T cells in non-lymphoid tissues are resident, rather than recirculating. Residence was not absolutely predicted by common phenotypic markers (CD103 & CD69), a finding that demonstrates heterogeneity in the resident memory population and insists that migration rather than solely phenotype be used for identification. Despite tissue-specific immune regulation, establishment of resident memory CD8 T cells was extended to male genital tract tissues, where they maintain local cytokine production in the presence of rechallenge. Our studies of male genital tract organs revealed non-canonical migration of effector CD8 T cells directly into visceral non-lymphoid tissues of recently infected mice. Together, these results provide a systematic quantification of the distribution and compartmentalization of virus-specific memory CD8 T cell subsets and highlight the relative numerical abundance of resident memory CD8 T cells, indicating that host immunosurveillance by memory CD8 T cells is conducted in a highly localized manner.

Technological advancements in glycobiology have added a novel dimension to immunotherapy when carbohydrates are considered as key targets in developing safe and effective vaccines to combat cancer, viral infections, bacterial infections etc. Carbohydrates have long been known to be T cell independent and only elicit the innate immune responses and weak adaptive immune responses via cross-linking with B cell receptors. In order to achieve T cell dependent immunity, carbohydrates are generally conjugated with immunogenic carrier proteins. However, carrier proteins, being self-immunogenic, can lead to increased peptide specific antibody production resulting in the suppression of immunity towards the targeted carbohydrate antigen. Therefore, as an alternative to carrier proteins, we have utilized a zwitterionic polysaccharide (PS A1), isolated from the commensal anaerobe Bacteroides fragilis ATCC® 25285™/NCTC® 9343™, which is known to bind with the MHC class II molecule and invoke T cell dependent immunity.We hypothesize that covalent attachment of carbohydrate antigens with PS A1 will lead to the generation of carbohydrate antigen specific robust humoral and cellular responses. To validate this idea, we have synthesized a number of carbohydrates or glycopeptide antigens and conjugated these antigens with PS A1. Initially, a stereoselective glycosylation strategy has been developed to synthesize a complete set of tumor associated Thomsen-nouveau (Tn) antigen, D-GalNAc-O-Ser/Thr (Tn).5 Phthaloyl protected amino acid acceptors played a key role in generating a- or ß- anomers in excellent yields (72-96%) and selectivity (~100%) when appropriate C-2 substitution was installed. Protected Tn antigens were divergently transformed into Tn-based carboxylates or hydrazides by hydrogenolysis followed by hydrazine hydrate treatment or hydrazine hydrate treatment alone.a-L-Thr-Tn hydrazide was coupled to PS A1 utilizing sodium metaperiodate (NaIO4) mediated oxidation of PS A1 in acetate buffer (pH 5.0), followed by overnight incubation with the antigen in 1X PBS buffer (pH 7.2). Structural modification of PS A1 can modulate immunogenicity of PS A1. Therefore, it is required to ascertain that the oxidation and conjugation condition will not alter the intrinsic biological nature of PS A1. As it is challenging to precisely identify the reaction outcome from a polysaccharide, we have synthesized 1-propyl-D-galactofuranose and 1-thiophenyl-4,6-O-pyruvate acetal-D-galctopyranose, which mimics two monosaccharides present in the core structure of PS A1. These two saccharide moieties can be affected by the reaction conditions employed for conjugation. In order to evaluate the specific oxidation of the vicinal-diol containing a primary alcohol over the trans-diol, 1-propyl-D-galactofuranose was subjected to NaIO4 mediated oxidation with a relatively harsher conditions (more concentrated NaIO4 solution, longer reaction time) and experimental outcomes revealed that only the vicinal-diol containing a primary alcohol was oxidized to form an aldehyde even under harsh reaction conditions. This observation suggests that a regioselective conjugation of antigen to the PS A1 is possible under the reaction conditions. Again, an NMR based stability study was performed with 1-thiophenyl-4,6-O-pyruvate acetal-D-galctopyranose and it was determined that the pyruvate acetal is very stable at pH 5.0. Therefore, the zwitterionic charge character of PS A1 remains intact in the reaction conditions that is generally employed for antigen-PS A1 conjugation.L-Thr-a-Tn antigen was coupled with poliovirus derived immunogenic peptide (PV), which was further conjugated to oxidized PS A1 to access a glycopeptide-PS A1 based vaccine (Tn-PV-PS A1). We were under the assumption that the Tn-PV-PS A1 construct would elicit Tn antigen specific robust cellular and humoral responses via both MHC I and MHC II dependent pathways. To validate our hypothesis, we injected our vaccine and Tn-PV in C57BL/6 mice and a series of assays, including ELISA, FACS, anti-proliferation, T cell proliferation, IFN¿ production and JAM assay, were performed to decipher the immunogenicity of the vaccine giving mechanistic insight as well. It was revealed that the vaccine candidate (Tn-PV-PS A1) was able to generate more anti-Tn IgG antibodies than Tn-PV alone and the induced antibodies were able to recognize Tn expressing tumor cell lines. Mice immunized with our vaccine construct showed higher antigen specific CD4+ T cell proliferation than the Tn-PV immunized group. Again, better Tn specific CD8+ T cell responses were also induced with the Tn-PV-PS A1 construct than Tn-PV. Considering the aforementioned observations, it could be concluded that PS A1 is an unique immunogen which is suitable to use for generating carbohydrate specific CD4+ and CD8+ T cell dependent immunity.An aminooxy derivative of a tetrasaccharide repeating unit of S. dysgalactiae 2023 rhamnose rich polysaccharide was synthesized by utilizing a strategy inclusive of stereoselective glycosylations, which ultimately promoted facile purification of the desired compound in overall good isolated yields. This repeating unit was conjugated with PS A1 and this conjugation delivered the first semi-synthetic entirely carbohydrate immunogen targeting bacterial S. dysgalactiae 2023. The immunogenicity of this vaccine will be evaluated in the near future. Generating immune responses against this bacteria can potentially combat S. dysgalactiae related bovine mastitis, which is recognized as the most economically devastating disease in the dairy industry. We anticipate that immunological studies of this tetrasaccharide-PS A1 construct will produce a promising result because it contains two highly antigenic rhamnose moieties.

One of the consequences of human aging is an overall decline in immune function. A hallmark of this aging process, termed immunosenescence, is the downregulation of co-receptor CD28 from the surface of CD8+ T cells. This population of terminally differentiated CD8+CD28– T cells accumulates during aging and chronic infections, and is linked to many age-related diseases observed in elderly individuals. In this thesis, we explore the molecular factors regulating CD8+CD28– T cell fate and function. In Chapter III, we find that resting CD8+CD28– T cells possess a unique glycolytic profile that is associated with their enhanced cytotoxicity and decreased expression of NAD+-dependent protein deacetylase SIRT1. Global gene–expression profiling identified the transcription factor FoxO1 as a SIRT1-target involved in the transcriptional reprogramming of CD8+CD28– T cells. FoxO1 is proteasomally degraded in CD8+CD28– T cells, and inhibiting its activity in CD8+CD28+ T cells recapitulates the metabolic and cytotoxic phenotype of resting CD8+CD28– T cells. In Chapter IV, we explore the signaling and functional consequences of IL-15, a pro-memory cytokine that serves as a TCR-independent activation signal for CD8+CD28– T cells. We find that IL-15 signals through mTOR to upregulate a metabolic program conducive for effector T cell function. Targeting CD8+CD28– T cells with a SIRT1 activator or glucose-starvation suppresses IL-15-induced cytotoxicity and proliferation. Altogether, this body of work identifies new molecular pathways that regulate the function of human CD8+CD28– T cells—providing deeper insight into the mechanisms of immune aging, as well as novel therapeutic opportunities to target it.

Epstein-Barr virus (EBV) in a human herpesvirus. It infects about 90% of the human population, and is the main causative agent of infectious mononucleosis. The incubation period, the time between viral acquisition and onset of symptoms, is unusually long in patients presenting with infectious mononucleosis, lasting about six weeks. In addition to causing acute illness, there can also be long-term consequences as the result of acquisition of the virus, including nasopharyngeal carcinoma and lymphoma. Nevertheless, there remains a surprising dearth of knowledge regarding the establishment of and immune response to persistent EBV infection in its natural hosts, especially during the incubation period. We sought to address many of these gaps by studying the incubation period, acute phase, and convalescence of undergraduates experiencing infectious mononucleosis during primary natural EBV infection in a cohort of prospectively studied volunteers. Particular attention was paid to the previously uncharacterized incubation period. Our findings have focused on understanding the immune response that occurs in young adults presenting with infectious mononucleosis, via gene expression changes as observed in peripheral blood mononuclear cells and innate and adaptive immune cells. Using a systems biology approach we discovered that important gene expression changes occur during the immune response to primary EBV infection. A typical antiviral type I interferon response was not observed at onset of infectious mononucleosis symptoms, but rather up to two weeks prior. The gene expression signature at symptom onset was dominated by cell cycle related genes, probably due to the CD8 T cell lymphocytosis, and type II interferon regulated genes. Interestingly, comparison of the EBV signature with other acute viral infections revealed very little similarity. The EBV signature showed the greatest similarity with hemophagocytic syndromes. This result is consistent with the view that infectious mononucleosis is an immunopathologic disease, and is supported by evidence that EBV can cause hemophagocytic lymphohistiocytosis. As an extension of this work, we carefully examined changes in cellular phenotypes and population frequencies to determine if there were significant alterations to certain compartments during the response to primary EBV infection. We observed a type I interferon signature in a larger subset of study participants during the incubation period. This response was concurrent with the transition of virally infected B cells from the oral cavity to the blood, a decline in plasmacytoid dendritic cells from the circulation, and a polyclonal CD8 T cell activation. No EBV specific CD8 T cells activation was observed until the onset of infectious mononucleosis symptoms. A major obstacle to understanding EBV related sequelae has been the lack of an efficient animal model for EBV infection, although progress in primate and mouse models has recently been made. Taken together, the data compiled in this thesis provide important first descriptions of the immune responses that occur during the establishment of a natural persistent infection in humans. Key future challenges are to develop protective vaccines and effective treatment regimens.

Abstract: CD8+ T cells are vital cytotoxic lymphocytes against intracellular pathogens and tumor cells. CD8+ T cells utilize a fast-acting degranulation of granzyme/perforin and slow-acting receptor:ligand, FasL:Fas, pathway of inducing target cell lysis. The importance of degranulation based target cell killing by CD8+ T cells has been well established while the FasL:Fas based mechanism of killing has been far less characterized. Additionally, it remains unclear as to the purpose of possessing the slow-acting FasL:Fas mechanism of killing in CD8+ T cells. In this thesis, I investigated the differential expression patterns of intracellular FasL and GrB in stimulated CD8+ T cells to further delineate the roles between degranulation and FasL. Interestingly, while I found GrB expression in CD8+ T cells to be proportional to the level of stimulation, weak stimulation is more favorable for FasL expression. In ex vivo stimulated CD8+ T cells, while strong stimulation induced expression of intracellular FasL, only weakly stimulated cells were able to sustain FasL expression over an extended period of time. Using allogeneic tumors, I further demonstrated that high antigen availability stimulates CD8+ T cells to express GrB and FasL while low antigen availability favors intracellular expression of FasL expression. Collectively, these results suggest two strategies of effector molecule expression: a low antigen level strategy favoring FasL expression and high antigen level strategy of GrB and FasL expression. In addition to CD8+ T cells, NK cells are also potent cytotoxic lymphocytes. As part of the innate immune system, NK cells were not believed to possess adaptive immune cell features. However, recent studies have demonstrated that following stimulation, NK cells may possess a number of memory cell-like features such as longevity, antigen specificity and enhanced secondary responses. While no definitive memory marker(s) have been found for memory NK cells, Ly-6C, a memory marker for CD8+ T cells, is a promising candidate. In my study, I first demonstrated that Ly-6C positive NK cells are mature NK cells based on effector function and phenotype. Additionally, based on Ly-6C staining with monoclonal AL-21 and iMAP antibodies, two subsets of Ly-6C positive NK cells are visible, an AL-21 only reactive subset, Ly-6CAL-21, and an AL-21/iMAP reactive subset, Ly-6CAL-21/iMAP. I further demonstrated that Ly-6C positive NK cells develop from Ly-6C negative NK cells and that Ly-6CAL-21/iMAP NK cells develop from Ly-6CAL-21 NK cells. In long term studies, this progression of Ly-6C expression suggests that Ly-6CAL-21/iMAP NK cells may represent a previously undescribed differentiation state of long-lived NK cells. Collectively, AL-21 and iMAP co-reactivity towards Ly-6C may further define the population of memory NK cells.

Herpes Simplex Virus-1 (HSV-1) is the primary cause of sporadic viral encephalitis in the United States. While prompt treatment results in high survival rates, greater than fifty percent of patients develop severe, life-altering neurological deficits subsequent to Herpes Simplex encephalitis (HSE). Inflammatory cues generated during non-HSE viral encephalitis have been shown to alter reparative neurogenesis in humans and mice. However, interactions between induced immune responses and the normal processes of adult neurogenesis remain unexplored during HSE. The present thesis hypothesized that inflammation induced by experimental HSV-1 infection of laboratory mice alters neural stem/progenitor cell (NSC) proliferation. The response of NSCs was dynamic throughout HSV-1 infection of adult BALB/c mice, with a significant increase in the NSC population during acute HSE at 6 d p.i. and a significant decrease by 15 d p.i.. The 6 d p.i. and 15 d p.i. time points coincided with macrophage infiltration and CD8(+) T cell infiltration respectively. At 6 d p.i., infiltrating macrophages were observed to be alternatively activated (M2) whereas macrophages at 15 d p.i. were classically activated (M1 macrophages). Treatment with M2 macrophages resulted in significant increases in the NSC population size both in vivo and in vitro. In vitro, the increase in NSC proliferation was found to be mediated via Wnt5a. In vivo, transplanted M2 macrophages increased the number of NSCs and Wnt5a(+)CD11b(+) macrophages were observed in the brains of 5 d p.i. mice. On the other hand, M1 macrophages activated with interferon-γ did not support NSC proliferation, which suggested that macrophage effects may be modulated by other immune factors present in the infected brain. CD8(+) T cells are the largest population of infiltrating immune cells during chronic HSE and produce robust amounts of IFN-�. The interactions of CD8(+) T cells and NSCs were addressed in vitro and in vivo experiments. Co-culture of activated CD8(+) T cells and NSCs abrogated NSC proliferation in vitro while in vivo depletion of CD8(+) T cells restored the NSC population in infected mice. Further evaluation of the effect of CD8(+) T cells on NSC proliferation was dependent on IFN-� both in vivo and in vitro, as IFN-� receptor 1 (IFN-�R1) knockout mice were unresponsive to CD8(+) T cells in culture and infected IFN-�R1 knockout mice had no reduction in the size of the NSC population. To continue evaluation of interactions of NSC proliferation and IFN-�, alternative models of HSE were assessed in C57BL/6 mice. The kinetics of immune cell and NSC proliferation were similar between BALB/c and C57BL/6 mice, but macrophage and T cell phenotypes were different between the strains. Macrophages infiltrating the brain, at any time point examined, were persistently classically activated with an M1 phenotype, and T cell infiltration was dominated by CD4(+) T cells in C57BL/6 mice. On the other hand, NSC proliferation was significantly decreased when IFN-� was expressed suggesting that the interactions of IFN-� and NSCs were similar in both strains of mice tested. The studies presented in this dissertation are the first to suggest a connection between IFN-� produced by infiltrating T cells and impaired NSC proliferation. The mechanism by which T cells mediate this effect may be direct or through interaction with other cell types in the brain. However, there are implications that NSC proliferation and associated reparative neurogenesis may be impacted in any model of brain damage that exhibits prominent IFN-� expression.

Immunotherapy has emerged as a promising tool to treat diseases via enhancing or suppressing immune responses of T cells. For infectious diseases and cancer, inducing expansion of antigen–specific CD8 T cells has proved effective in preclinical and clinical studies for eliminating infected or malignant cells. In this thesis a subcellular vaccine prepared by sonication of disrupted mature dendritic cells (DCs) followed by loading with specific peptides was developed and used to stimulate naïve CD8 T cells. To achieve this, a new culture method was developed to prepare large numbers of mature DCs from mouse bone marrow (BM). Cell disruption and sonication of these BMDCs yielded membrane vesicle nano-particles that were immunogenic for CD8 T cells both in vitro and in vivo. Notably, the in vivo immunogenicity of the vesicles was considerably increased when co-delivered with anti-CD9 antibodies and CpG oligodeoxynucleotides, especially when given in multiple injections. Preliminary experiments demonstrated the potential of membrane vesicles for cancer immunotherapy in murine models.

Characterizing the cellular participants in tissue immune responses is critical for understanding infection, cancer, autoimmunity, allergy, graft rejection, and other immunological processes. Leukocytes recirculate through blood vessels before localizing to tissue sites of immune responses. Thus, in experimental animal models, tissues are often perfused to putatively remove blood-borne leukocytes that may conflate analysis. Here, we develop and validate an intravascular staining methodology that distinguishes between vascular and tissue-localized cells of the immune system. We demonstrate that perfusion both fails to remove many blood-borne leukocytes and also may remove tissue-localized populations of interest, and we provide examples of how this issue distorts interpretation of leukocyte differentiation state, migration, and phenotype in healthy mice, as well as those responding to viral or Mycobacterium tuberculosis infection or tumor challenge. Additionally, we utilize intravascular staining to examine resident memory T cells in non-lymphoid tissues, such as the lung, liver and female reproductive tract. This study highlights the breadth and gravity of the issues regarding tissue leukocyte composition, outlines simple methods for identification of various intravascular leukocyte populations, reviews the limitations of the technology, and demonstrates that these methods should be routinely adopted in lieu of perfusion for interpretable and accurate analyses of immune responses in many tissues.

TNF-receptor associated factor (TRAF) 2 integrates multiple TNF-family signaling pathways in T cells making it a unifying target for immunomodulation based on co-stimulation. This thesis examines the role of TRAF2 in T cell effector responses towards allografts in T cell specific TRAF2 knockout (TRAF2TKO) mice. TRAF2TKO mice permanently accepted the majority of full MHC-mismatched islet allografts without exogenous immunosuppression. Further, rejection of vascular heterotopic heart and skin allografts was delayed. Without TRAF2, CD4+ T cells exhibited reduced proliferation and an altered T cell activation profile correlating with reduced NF-κB and JNK activation. Thus CD4+ T cells require a TRAF2-NF-κB and TRAF2-JNK signalling cascade for activation. Under specific polarizing conditions TRAF2TKO CD4+ T cells more readily converted to a Th2 and Treg phenotype but exhibited impaired Th17 effector differentiation highlighting a role for TRAF2 in T cell effector differentiation. The impaired alloresponse of TRAF2TKO mice was confined to the effector (CD25-) compartment as depletion of CD25+ Tregs did not restore alloresponses to control levels both in vitro and in vivo. Whilst T cell survival and peripheral turnover were unchanged by loss of TRAF2, TRAF2TKO CD4+ T cells failed to persist in a lymphopenic environment, likely contributing to long-term allograft survival in RAG-/- mice receiving TRAF2TKO CD4+ T cells. Phenotypic analysis of T cell subsets in TRAF2TKO mice revealed striking reductions in IL-15 dependent T cell populations: CD8+ effector-memory, CD8+ central-memory, and NKT cells. The loss of IL-15 dependent T cell populations was mechanistically linked to an inability to respond to physiological levels of IL-15. Further, TRAF2TKO mice exhibited reduced CD4+ effector-memory and central-memory T cells, a possible consequence from impaired IL-15 responses. TRAF2TKO CD8+ central-memory and NKT cells were rescued by a super-physiological in vivo boost of IL-15 and exhibited control levels of IL-15 signalling components indicating that loss of TRAF2 does not abolish IL-15 responses but modulates IL-15 sensitivity in vivo. Therefore, targeting the TRAF2 system may be useful to blunt CD4+ T cell dependent allograft rejection and reduce CD8+ memory-induced transplant tolerance by modulating IL-15 responses.

Persistent viral infection places a significant burden on global heath. Pathogens such as human immunodeficiency virus (HIV), and hepatitis B and C are able to replicate at high levels within their host, outpacing the immune response and ultimately leading to its attenuation. This attenuation is largely the result of host-derived suppressive factors produced in response to the chronic inflammation and immune-mediated damage to critical lymphoid organs accrued throughout the course of the immune response itself. The landscape of the persistently infected immune environment is thus profoundly altered from its naïve state, raising the possibility that new CD4 and CD8 T cell responses that develop and differentiate in the midst of persistent infection could function differently than T cell responses in a healthy environment. Using lymphocytic choriomeningitis virus (LCMV) as a model system of persistent infection, we determined that de novo virus-specific T cell responses primed in the midst of persistent infection are impaired. Virus-specific CD8 T cells primed in the midst of persistent infection assume a terminally differentiated phenotype and form defective peripheral effector memory, while virus-specific CD4 T cell responses fail to generate Th1 immunity, instead becoming almost exclusively T follicular helper cells (Tfh). This alteration in the T cell differentiation pattern was mediated by chronic type I interferon (IFN-I) signaling, and blockade of IFN-I signaling restored normal differentiation. Moreover, we observed that IFN-I signaling in the context of persistent LCMV infection led to profound thymic atrophy and disrupted the development of T cell precursors. Together, our data demonstrate several novel mechanisms by which IFN-I acts to modulate the immune response during persistent viral infection. Moreover, the data suggest that modulation of the IFN-I signaling pathway or its downstream targets could lead to treatments to purge persistent viral infection.

A hallmark of age-related immunosenescence is the accumulation of late-differentiated memory CD8 T cells in peripheral blood with features of replicative senescence, such as inability to proliferate, absence of CD28 expression, shortened telomeres, loss of telomerase activity, enhanced activation, and increased secretion of inflammatory cytokines. Oligoclonal expansions of these late-differentiated memory CD8 T cells in peripheral blood are correlated with increased risk of morbidity and mortality in elderly humans. The studies described in this thesis focus on furthering our understanding of CD8 T cell replicative senescence in humans and its relationship to aging and HIV disease, with special emphasis on immune features of gut-associated lymphoid tissue (GALT), which contains 40-65% of lymphocytes (as opposed to peripheral blood, which houses only 2%) and is an area of high antigenic exposure. We first review the data for CD8 T cell replicative senescence in peripheral blood. This includes a comprehensive discussion of experiments defining the phenotype and function of senescent cells in vitro. We then present evidence that CD8 T cell replicative senescence occurs in vivo, and discuss how senescent cells contribute to, and are affected by, dysregulated immunity in aging and HIV disease, ultimately contributing towards "inflammaging" and an immune risk phenotype (IRP) predictive of morbidity and mortality. We next move the focus of our studies to the GALT and compare blood and gut (specifically rectosigmoid colon) CD8 T cells, focusing on age-related phenotypic and functional alterations previously linked to senescence in peripheral blood. Overall, our results support the hypothesis that gut CD8 T cells have profiles suggestive of greater "aging" than those in peripheral blood. These results include a significant increase in RA-CD8+ (memory), CD28-CD8+ (differentiated), RA-CD28-CD8+ (differentiated memory), DR+CD38+CD8+ (activated), CD8 &alpha/&alpha, RO+PD-1+CD8+, and a (non-significant) decrease in baseline telomerase activity. However, other criteria from this study (lower levels of CD57 expression, lack of inverted CD4:CD8 ratio, increase in KI-67 (proliferation) expression, and proliferative potential that is similar to blood) suggest that gut CD8 T cells may not be more senescent than peripheral blood. Our data suggest that additional work (such as looking at CD4 cells, other regions of the gut, and CD8 T cell gut senescence in aging and HIV disease) is needed to further elucidate this relationship. We then describe the development and optimization of a novel protocol to test our hypothesis of compartment-specific differences of in vitro proliferative dynamics of CD8 T cells. The protocol involves a 5 day culture of mononuclear leukocyte populations from blood and gut, respectively, labeled with CFSE and BrdU and stimulated with anti-CD2/3/28-linked microbeads. The specific experimental variables we have optimized include: mode of T cell stimulation, CFSE concentration, inclusion of second proliferation marker BrdU, culture duration, initial cell concentration and inclusion of autologous irradiated feeder cells. Finally, we transition back to peripheral blood in order to examine a link between peripheral blood T cell senescence and a mild presentation of Kaposi's sarcoma in treated HIV-infected subjects with long-term suppression of HIV replication. This study indicates these subjects exhibit some features of immunosenescence relative to treated controls without sarcoma (increased frequency of CD57+ and CD28- CD4 and CD8 T cells, decreased proportion of naïve CD27+CD28+CD45RA+ CD4 and CD8 cells, increased expression of CCR5 in CD4 and CD8 T cells), indicating immunologic perturbations associated with immune senescence might be casually associated with development of this disease. Together, results from this research further our understanding of CD8 T cell senescence, especially in the rectosigmoid colon region of the GALT, and pave the way for further research elucidating the relationship of CD8 T cell immunosenescence and pathologies such as aging, HIV disease and Kaposi's sarcoma.

Regulatory T cells (Tregs) are critical to the maintenance of immune homeostasis and the prevention of autoimmunity. In the NOD mouse model of autoimmune diabetes, treatment with therapeutic Tregs can prevent and even reverse disease onset. The means by which Tregs are able to exert their protective effects in vivo have remained incompletely understood. This work focused on elucidating the critical cellular targets of Tregs at the site of inflammation and the effects of Tregs on these target cells. This was accomplished by examining the target tissue, the pancreatic islets of Langerhans, before and after Treg treatment with a variety of approaches. Islets and their immune infiltrates were analyzed at the transcriptional and post-transcriptional levels. Islet immune infiltrates were characterized at the phenotypic and functional levels by flow cytometry and by multiple modes of microscopic imaging. Therapeutic Tregs were found to quickly home to inflamed islets, where they engaged in dynamic interactions with resident dendritic cells (DCs). Treg arrival was followed by a rapid decrease in islet CD8 T cells that corresponded with a reduction in the expression of cytotoxic effectors and cytokines. T cells that persisted in the islets following Treg treatment showed no decrease in proliferative capacity but were suppressed at the level of effector function, where they lacked production of IFNg protein despite continued expression of IFNg mRNA. In CD8 T cells, this inhibition of effector function was related to a decrease in mTOR signaling that was required for sustained IFNg production. In addition to these acute effects on CD8 T cells following Treg treatment, we also observed a more gradual impact on islet DCs. The recruitment of DCs was reduced in the weeks following Treg treatment. Islet DCs were found to arise from a blood monocyte precursor. Overall, our findings that therapeutic Tregs target CD8 T cells during acute disease control and DCs during later stages implicate the importance of these targets in disease pathogenesis and elucidates the mechanisms by which therapeutic Tregs exert their protective effects.

\(CD8^+\) cytotoxic T lymphocytes (CTLs) play a critical role in controlling human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) replication. The CTL responses that are thought to be the most protective against HIV and SIV are those that are of high frequency, recognize multiple epitopes, and perform multiple antiviral functions. Therefore, current vaccines aim to elicit CTLs possessing these characteristics. However, the phenomenon of immunodominance likely limits the potential of vaccines from generating such CTL responses by restricting the breadth of epitopes recognized by CTLs and the frequency and functionality of these CTL responses. In this dissertation, we explored the relationship between SIV epitope dominance and the functionality of the epitope-specific CTL populations. We also examined factors that contribute to the development of SIV epitope immunodominance hierarchies. We initially investigated the relationship between SIV epitope dominance and the antiviral functionality of the epitope-specific CTL populations in rhesus monkeys. We performed a gene expression analysis in dominant and subdominant epitope-specific CTLs during the acute phase of SIV infection and observed differential expression of a number of genes during this time. Subsequent in vitro functional studies of these epitope-specific CTL populations during the chronic phase of infection confirmed the presence of differences in maturation phenotype and functional capacity of dominant and subdominant epitope-specific CTLs. These studies demonstrate a relationship between epitope dominance and antiviral functionality of epitopespecific CTLs and suggest that dominant and subdominant epitope-specific CTLs may differ in their protective role against HIV acquisition and replication. This has important implications for vaccine design. In subsequent studies, we investigated the contribution of the binding of the peptide:MHC (pMHC) complex to the T cell receptor (TCR) in the development of immunodominance hierarchies. Using surface plasmon resonance, we measured the kinetics and the affinity of the interactions between dominant and subdominant epitope pMHC complexes with their respective TCRs. We found that epitope dominance was associated with higher affinities of pMHC:TCR binding. These findings indicate a molecular interaction that may be manipulated in vaccine-induced CTL responses to enhance their frequency and functional capacity.

Memory phenotype CD8+ T cells are typically thought to have undergone an immune response to foreign antigen and to have differentiated from antigen-specific precursors in the naïve pool. However, using a peptide-MHC I tetramer enrichment technique, we identified foreign antigen-specific memory-phenotype CD8+ T cells in unimmunized mice. These cells (termed "virtual memory" T or VM cells) were observed in mice maintained in both specific-pathogen- and germ-free (SPF and GF respectively) housing. This thesis focuses on the relationship between VM cells and "conventional" memory cells: memory cells arising from homeostatic proliferation (HP), and innate-like memory CD8+ T cells such as IL-4 bystander memory CD8+ T cells. Our data indicate physiological HP and IL-4-driven bystander processes are the main mechanisms that drive the generation of VM cells and not the exposure to foreign antigens. VM cells arise in the periphery during the neonatal period and are maintained long term. We also show that VM cells respond in vitro to innate cytokines (similar to conventional memory CD8+ T cells) and they outcompete antigen-specific naive CD8+ T cells in in vivo responses. Overall our observations suggest that VM cells arise out of normal homeostatic and IL-4-driven bystander processes in unimmunized SPF and GF mice, and express at least some memory-like capabilities.

The limiting factor to the broad use of bone marrow transplantation (BMT) to cure hematopoietic malignancy is graft verses host disease (GVHD). GVHD is caused by mature T cells present in the bone marrow graft that recognize and destroy host epithelial tissue. Mature T cells facilitate several beneficial effects including elimination of residual malignant cells, the so-called graft-vs-leukemia (GVL) effect. The goal of this work was to develop a strategy for attenuating GVHD pathology mediated by donor T cells without compromising beneficial GVL effects. During GVHD, integrins present on the surface of T cells dictate T cell localization and accumulation patterns. CD103 is an integrin that is upregulated on CD8 T cells during GVHD. Given that the only known ligand of CD103 is the epithelial-specific molecule E-cadherin, we postulated that CD103 expression is required for destruction of the gut epithelium during GVHD, but is not required for effective clearance of a blood malignancy. To test this overall hypothesis, we first characterized CD103 expression and its role in promoting GVHD pathology in the host gut in both an MHC-disparate and a clinically relevant MHC-matched GVHD model. We then assessed the capacity of donor T cells on the CD103-deficient background to mediate GVL effects. Our studies showed that CD103 expression was required for optimal accumulation of CD8 T cells in the gut epithelium. Co-depletion ofiiiCD4 and CD103 expressing cells completely prevented mortality following MHC-disparate BMT. Interestingly, CD103 depletion exacerbated GVHD pathology following MHC-matched GVHD induction. Adoptive transfer studies showed that T cells from CD103-deficient donors cleared malignant B cells as effectively as those from wild-type donors. Our data show a conflicting role for CD103 following BMT, however, our data suggest that CD103 expression on CD8 T cells promotes intestinal pathology, but is not required for host defense against hematopoietic malignancy. Together, these data suggest that CD103 blockade may have the potential to prevent gut GVHD pathology without compromising GVL effects.

Compelling research has linked CD8 T cells to neuronal injury, a pathological hallmark of multiple sclerosis. The epitopes recognized by CD8 T cells in MS patients remains unknown. Defining these epitopes is of potentially critical importance to developing new therapeutics for MS, an aspect of which has been explored by our lab. In previous studies using the established Theiler's murine encephalomyelitis virus (TMEV) model of MS we demonstrated that expansion of an immunodominant population of CD8 T cells could be inhibited by the administration of the epitope peptide that they recognized. Inhibiting the expansion of this CD8 T cell population resulted in reduced levels of neuronal injury and sparing of motor ability. In work presented here we have developed an unbiased, molecular approach for determining CD8 T cell epitopes using the TMEV model. Such an approach could be potentially translated from this animal model to MS, possibly contributing to future therapeutics. Further, defining the putative mechanism utilized by CD8 T cells to perpetuate neuronal pathology in diseases such as Rasmussen's encephalomyelitis, neurological neoplastic disorders and MS remains a critical topic of research. Defining this mechanism could have important ramifications for the development of new treatments for these conditions. In studies presented here using the TMEV model we demonstrate evidence that strongly suggests CD8 T cells form traditional immune synapses with neurons. Further, we demonstrate that CD8 T cells also polarize granzyme B, an effector molecule that can potentially induce apoptosis, towards target neurons. Taken together this evidence provides a potential mechanism utilized by CD8 T cells to perpetuate neuronal pathology. Lastly, we demonstrate in these studies that morphological changes observed in CD8 T cells are iii associated with movement through CNS tissue. The motility patterns of these CNS infiltrating CD8 T cells was similar to the movement observed in other studies examining T cells within lymph nodes. Such observations may suggest that a common cellular mechanism controls morphological changes and motility among CD8 T cells regardless of the organ within which they are derived.

Pediatric HIV infection is a major worldwide public-health challenge, with an estimated rate of 1,200 children being infected daily in sub-Saharan Africa alone. In spite of this, the HIV-specific immune response in perinatally infected children has been relatively understudied. These studies were designed to probe specific aspects of the immune response to HIV in a group of vertically infected children/adolescents, treated at the Jacobi Medical Center in the Bronx, NY. The first study investigated the association between ethnicity and HIV-specific CD8+ T cell responses. Using a cytokine-enhanced ELISpot assay to measure IFN-gamma secretion in response to stimulation by HIV proteins Gag, Nef, and Tat, we show that African-American subjects displayed an immune response of higher magnitude than age and disease-matched Hispanic subjects. With the second study, we focused on defects in monocytes, specifically the signaling within, resulting from HIV infection. We utilized a recently developed intracellular staining technique, phospho-flow, to investigate the Jak/Stat signaling pathway in the monocytes of HIV-infected children. We found significantly lower levels of Jak/Stat signaling in response to stimulation in HIV-infected individuals compared to non-infected individuals. The last study focused on immunodominance patterns and differentiation profiles of Gag-specific CD8+ T cell responses in two progression groups within the same cohort of perinatally infected children. Using CDC guidelines, we classified patients into those with no immunosuppression or with severe immunosuppression. Using the IFN-gamma ELISpot assay, we then identified all epitope-specific CD8+ T cell responses to the Gag protein and compiled the immunodominance hierarchy of each patient. Utilizing multi-parameter cytokine flow cytometry, we then identified the differentiation profile and cytokine secretion profile of each epitope-specific response. We found significant differences between the progression groups with respect to immunodominance hierarchies of Gag-specific CD8+ T cell responses as well as striking differences in the differentiation profiles of CD4+ T cells, specifically an accumulation of the CCR7-CD45RA+ subset. Taken together, these studies add to the field of pediatric HIV immune responses and identify several factors, namely race, monocyte defects, and immunodominance patterns, which can play an important role in the role of the immune response in progression of HIV disease in perinatally infected children.