To understand the functional interactions between the TrkA and p75 nerve growth factor (NGF) receptors, we employed several lines of investigation including biophysical, biochemical and cellular assays. A high-affinity nerve growth factor (NGF) receptor is thought to be a complex of two receptors, p75 and the receptor tyrosine kinase, TrkA. The existence of a gp75-TrkA complex was demonstrated by a copatching technique. p75 on the surface of intact cells is patched with an anti-p75 antibody and fluorescent secondary antibody, the cells are then fixed to prevent further antibody-induced redistributions, and the distribution of TrkA is probed with an anti-TrkA antibody and fluorescent secondary antibody. We utilize a baculovirus-insect cell expression system, which allows high level expression of wild-type and mutated NGF receptors. TrkA and p75 copatch in both the absence and presence of NGF. This association is specific, since p75 does not copatch with other tyrosine kinase receptors, including TrkB, platelet-derived growth factor receptor-β and Torso (Tor). To determine which domains of TrkA are required for copatching, we used a series of TrkA-Tor chimeric receptors and show that the extracellular domain of TrkA is sufficient for copatching with p75. A chimeric receptor with TrkA transmembrane and intracellular domains shows partial copatching with p75. Deletion of the intracellular domain of p75 decreases but does not eliminate copatching. A point mutation that inactivates the TrkA kinase has no effect on copatching, indicating that this enzymatic activity is not required for association with p75. Hence, although interactions between the p75 and TrkA extracellular domains are sufficient for complex formation, interactions involving other receptor domains also play a role. To study what signal transduction mechanisms were activated by the two receptors to bring about differentiation and survival, we stably transfected LAN5 neuroblastoma cells with an expression vector for ET-R, a chimeric receptor with the extracellular domain of the epidermal growth factor receptor (EGFR) and the TrkA transmembrane and intracellular domains. EGF activated the ET-R kinase and induced partial differentiation. NGF, which can bind to endogenous p75, did not induce differentiation, but enhanced the EGF-induced response, leading to differentiation of almost all of the cells. A mutated NGF, 3T-NGF, that binds to TrkA but not to p75 did not synergize with EGF. Enhancement of EGF-induced differentiation required at least nanomolar concentrations of NGF, consistent with the low-affinity p75 binding site. EGF may induce a limited number of neuronal cells because it also enhances apoptosis. Both NGF and a caspase inhibitor reduced apoptosis and, thereby, enhanced differentiation. NGF appears to enhance survival through the phosphatidylinositol-3 kinase (PI3K) pathway. Consistent with this hypothesis, Akt, a downstream effector of the PI3K pathway, was hyperphosphorylated in the presence of EGF+NGF. These results demonstrate that TrkA kinase initiates differentiation, and p75 enhances differentiation by rescuing differentiating cells from apoptosis via the PI3K pathway. Even though both EGF and NGF are required for differentiation of LAN5/ET-R cells, only NGF is required for survival of the differentiated cells. In the absence of NGF, the cells die by an apoptotic mechanism, involving caspase-3. An anti-p75 antibody blocked the survival effect of NGF. Brain-derived neurotrophic factor also enhanced cell survival, indicating that in differentiated cells, NGF acts through the p75 receptor to prevent apoptosis.

Dbp2p, a member of the large family of DEAD-box proteins and a yeast homolog of human p68, was shown to interact with Upf1p, an essential component of the nonsense-mediated mRNA decay pathway. Dbp2p:Upf1p interaction occurs within a large conserved region in the middle of Upf1p that is largely distinct from its Nmd2p and Sup35/45p interaction domains. Deletion of DBP2, or point mutations within its highly conserved DEAD-box motifs, increased the abundance of nonsense-containing transcripts, leading us to conclude that Dbp2p also functions in the nonsense-mediated mRNA decay pathway. Dbp2p, like Upf1p, acts before or at decapping, is predominantly cytoplasmic, and associates with polyribosomes. Interestingly, Dbp2p also plays an important role in rRNA processing. In dbp2Δ cells, polyribosome profiles are deficient in free 60S subunits and the mature 25S rRNA is greatly reduced. The ribosome biogenesis phenotype, but not the mRNA decay function, of dbp2Δ cells can be complemented by the human p68 gene. We propose a unifying model in which Dbp2p affects both nonsense-mediated mRNA decay and rRNA processing by altering rRNA structure, allowing specific processing events in one instance and facilitating dissociation of the translation termination complex in the other.

Human immunodeficiency virus type-1 (HIV-1) is the etiologic agent of acquired immune deficiency syndrome (AIDS). In most cases HIV-1 infection in humans, leads to AIDS, which is characterized by opportunistic infections leading to death. The role various infectable cell types play in the course of infection is unclear. However, it is becoming increasingly more evident that cells of the monocyte/macrophage lineage are very important at several stages of disease. They are involved in the transmission, establishment and dissemination of infection as well as the AIDS related complication of dementia and pulmonary dysfunction. The regulation of virus expression in monocyte/macrophages while maintaining normal cell function would be of great benefit. Retinoic acid (RA) is a bioactive metabolite of vitamin A, an essential nutrient, and acts as a transcriptional regulator of many genes. RA is also a potent modulator of myeloid cell differentiation and function; it is currently used clinically. Clinical data indicate that serum vitamin A levels are inversely correlated with various aspects of HIV-1 induced disease. Furthermore, work done by several groups has demonstrated that RA directly modulates HIV-1 replication in cells of the myeloid lineage. RA is capable of either stimulating or repressing HIV-1 replication depending on the cell type used. This dichotomy appears to depend upon the differentiation state of the cells. Changes in differentiation states are associated with the altered expression of many cellular proteins including transcriptional regulators. Experiments indicate that the TATA box of HIV-1 is required for full levels of gene expression. I hypothesized that RA was modulating replication at the level of LTR-directed gene expression, and that the differentiation state of the cell influences the RA modulation. This thesis demonstrates that the RA effect is at the level of gene expression mapping to a promoter proximal element for both HIV-1 and simian immunodeficiency virus (SIVmac.) The ability of RA to stimulate or repress expression depends upon the differentiation state of the cells. Using U937 promonocyte cells, I demonstrate that RA increases SIVmac and HIV-1 transcription. When THP-1 monocytes or primary macrophages are used, I demonstrate that RA induces repression of HIV-1 and SIVmac. This RA modulation of expression is associated with altered complexes binding to the promoter proximal regions of HIV-1 and SIVmac. There has been a great deal of interest in CD8+ T cell derived factors which modulate HIV-1 replication. Work done by Levy and colleagues over a decade ago demonstrated that factors secreted by CD8+ T cells could block HIV-1 replication. Others have shown that the β-chemokines, released by activated CD8+ T cells, can block the entry of HIV-1 into macrophages. Center and colleagues identified a lymphocyte chemoattractant factor as IL-16. IL-16 is released by activated CD8+ T cells and it's receptor is CD4. IL-16 induces the migration of CD4+ T lymphocytes, and has been shown to activate many signaling pathways in CD4+ T lymphocytes. Kurth et al. demonstrated that IL-16 blocked the replication of HIV-1 in CD8+ depleted PBMC. In these experiments, it was not determined whether IL-16 was blocking viral entry (preventing viral binding to CD4) or whether IL-16 had inhibitory effects on subsequent steps in the virus life cycle. While IL-16 and HIV-1 share CD4 as their receptor, IL-16 binding was mapped to a separate epitope on CD4 from the HIV-1 binding site. Therefore I began experiments to determine how IL-16 regulates HIV-1 expression in T cells. I hypothesized that the IL-16 signaling pathway is involved in repressing HIV-1 gene expression. Experiments presented here demonstrate that IL-16 represses LTR-directed gene expression in T cell lines in a CD4 dependent manner. The IL-16 mediated repression is dependent on a DNA binding site contained within the viral core enhancer region. The data are also consistent with IL-16 inducing a repressor which binds within or adjacent to the HIV-1 core enhancer region.

The prevalence of type 2 diabetes mellitus is increasing worldwide and is considered one of the top health concerns globally. The occurrence of type 2 diabetes is linked to the rapidly increasing trend of obesity in both adults and children, which is proposed to be a contributing factor in the development of insulin resistance and type 2 diabetes. White adipose tissue, an insulin target tissue, is an important endocrine organ involved in the control of energy homeostasis through its direct influence on metabolism, insulin sensitivity and food intake. To better understand these functions, we studied adipocyte differentiation in 3T3-Ll cells, a white adipose tissue cell line. Many mitochondrial proteins exhibit an increase in expression levels during adipogenesis as identified by mass spectrometry. Moreover, increased mitochondrial mass and altered morphology was observed by light microscopy. Qualitative changes in mitochondrial gene expression were also observed during adipogenesis as revealed by Affymetrix GeneChip analysis. Additionally, striking changes in mitochondrial protein expression and morphology were identified following treatment with the insulin sensitizing agent, rosiglitazone. These results suggest that mitochondrial biogenesis and remodeling is inherent to white adipocyte differentiation. To investigate the physiological relevance of these findings, mRNA and protein expression profiles and mitochondrial morphology were studied during the development of insulin resistance and obesity and following treatment with rosiglitazone in ob/ob mice. These studies reveal a marked decrease in transcript levels for over 50% of mitochondrial genes with the onset of obesity in ob/ob mice. Rosiglitazone treatment stimulates enhanced expression in approximately half of these genes, as well as changes in mitochondrial mass and remodeling. Furthermore, these studies reveal that depressed oxygen consumption and fatty acid oxidation occur with obesity development and these alterations can be reversed with rosiglitazone treatment. This work identifies the previously underscored plasticity of mitochondria in white fat and suggests that mitochondrial biogenesis and remodeling in white adipose tissue may lead to systemic changes in insulin sensitivity and energy homeostasis. Lastly, these studies suggest that mitochondria may be an important therapeutic target for antidiabetic drugs.

The hemagglutinin-neuraminidase (HN) molecule of Newcastle disease virus (NDV) is an integral membrane glycoprotein that is oriented with its N-terminus in the cytoplasm and its C-terminus external to the infected cell. Single spanning membrane proteins with this type of topology (N-terminus in, C-terminus out) have been classified as Type II glycoproteins, in contrast to the more common Type I glycoproteins, which are oriented in the opposite direction. (C-terminus in, N-terminus out). The membrane integration of HN protein was investigated using a wheat germ translation system to synthesize and integrate HN protein into microsomal membranes in vitro. The insertion and translocation of HN protein into microsomal vesicles was found to occur cotranslationally without signal sequence cleavage. The membrane targeting required both signal recognition particle (SRP) and SRP receptor. Membrane binding assays utilizing HN nascent chain/ribosome/SRP complexes demonstrated that the membrane insertion of HN polypeptide required the presence of GTP, in a way similar to that described for secretory, multispanning and Type I proteins. To investigate further the membrane translocation process of HN protein, the amino terminal region of HN was mutated to determine the role of this region in the membrane integration of HN. The cDNA sequence encoding the bulk of the cytoplasmic tail of the HN glycoprotein was deleted. When transcripts produced from the mutated cDNA were translated in wheat germ extract in the presence of membranes, several abnormalities were identified in the interaction of the mutant protein with membranes. Although translocation and glycosylation of the mutant protein was detected, the efficiency of membrane translocation and the stability of the mutant protein's membrane interaction were reduced. Even though a large proportion of the mutant products remained nontranslocated and unglycosylated, many of these products were inserted into membrane vesicles in a reverse orientation from the wild type HN protein. The aberrant insertion of the mutant protein required both SRP and SRP receptor. Ribosome-bound mutant nascent chains were able to insert into membranes without the addition of GTP or SRP, but this GTP-independent insertion was in reverse. Therefore, the cytoplasmic tail of the HN glycoprotein appears to playa critical role in the maintanence of faithful directionality of the protein's membrane insertion.

Histone Nuclear Factor P (HiNF-P) is a known transcriptional regulator that is critical for the activation of replication dependent histone H4 genes during S phase. HiNF-P is a 65 kDa zinc finger protein that binds to its consensus binding sequence in the Cell Cycle Control Element (Site II) of the proximal promoter region of 11 of the 14 histone H4 genes. HiNF-P is a known co-factor of the global histone gene regulator and cyclinE/CDK2 substrate p220NPAT, however it was not known if this regulatory function reflected a physical interaction. In addition, other HiNF-P interacting proteins have yet to be identified. The work presented in this thesis identifies and characterizes HiNF-P interactions with various proteins within the cell, including p220NPAT. A yeast two-hybrid interaction screen identified candidate interacting proteins of HiNF-P and provided insight into novel cellular functions and transcriptional targets. A candidate yeast two-hybrid approach identified an interaction between HiNF-P and p220NPAT. This direct physical interaction links the cyclin E/CDK2 signaling pathway governing the G1/S phase transition with replication dependent histone gene transcription in S phase. An unbiased yeast two-hybrid screen for HiNF-P interacting proteins revealed an interactome library which suggests roles of HiNF-P in multiple cellular processes. This screen identified 67 candidate HiNF-P interacting proteins that are RNA processing factors, known and putative gene regulators, uncharacterized proteins, proliferation related proteins, as well as metabolic and signaling proteins. Identification of multiple RNA binding and processing factors, including the splicing cofactor, SRm300, links HiNF-P to mRNA processing. HiNF-P is potentially functioning in mRNA processing by interacting with these proteins directly and functioning in complex with them, or more likely, by recruiting these and other splicing factors to sites of transcription. We identified a number of known and putative gene regulators which are candidate HiNF-P interacting proteins. We isolated the atypical C2CH zinc finger protein, THAP7, a known transcriptional repressor. THAP7 interacts with HiNF-P by co-immunoprecipitation and co-immunofluorescence experiments. We show forced expression of THAP7 abrogates HiNF-P/p220 mediated activation of histone H4 gene transcription. THAP7 may represent a novel co-factor of HiNF-P and p220 mediated regulation of histone H4 genes. Identification of interacting proteins of HiNF-P that are involved in transcriptional regulation provides insight into other transcriptional targets of HiNF-P. HiNF-P is localized throughout the nucleus, presumably at multiple gene foci. These interacting proteins may represent novel co-factors of HiNF-P regulation of these other multiple target genes. HiNF-P has been identified as a regulator of cell cycle dependent histone genes, therefore we were interested in identifying other proliferation related proteins with which HiNF-P is interacting. We identified a number of proteins thought to be involved in cellular proliferation, including Ki-67 and an unknown protein XTP2. The functions of these proteins have not been identified. An interaction with HiNF-P might suggest a role for these proteins in histone gene regulation. In addition, Ki-67 has been implicated transcriptional control of ribosomal genes, although no role of HiNF-P in this function has been identified. HiNF-P is a known regulator of histone gene expression via a functional interaction with the global histone gene regulator and cyclin E/CDK2 substrate, p220. This thesis demonstrates HiNF-P directly interacts with the N-terminus of p220. This interaction requires multiple regions within the N-terminus including a LisH-like domain known to function in protein-protein interactions, a region (aa 121-145) known to be required for histone gene transactivation, and another uncharacterized region (209-318). In addition a phylogenically conserved region within the C-terminus of HiNF-P, the HiNF-P Specific Conserved Region (PSCR) is necessary for this interaction. Mutational analysis of these regions abrogates this interaction. HiNF-P and p220 co-localize at specific foci within the cell corresponding to Cajal bodies, which are known sites of histone gene clusters. This work shows that this interaction is necessary for histone gene transcriptional activation and HiNF-P dependent recruitment of p220 to histone H4 gene promoters. In addition HiNF-P as well as p220 interact with the Stem Loop Binding Protein (SLBP) and co-localize in situ. SLBP is a necessary factor for histone pre-mRNA processing events which also occur at Cajal bodies. These interactions provide evidence of the coupling of transcription and processing of histone genes and the involvement of common factors in both processes. This would allow for rapid production of abundant histone proteins which is needed during S phase. This thesis has identified multiple candidate interacting proteins of HiNF-P. These proteins establish HiNF-P as a protein involved in many cellular processes and mechanisms beyond transcriptional control of cell cycle dependent histone genes.

The function of GTP-binding proteins (G-proteins) in diverse intracellular pathways depends on their ability to switch between two forms, a GDP-bound (inactive) form and a GTP bound (active) form in a highly regulated GTPase cycle. The inactivation step of this cycle is regulated by GTPase-activating proteins (GAPs) which increase the intrinsic rate of hydrolysis of bound GTP; the activation step is regulated by a diverse family of GDP/GTP exchange factors (GEFs). A unique model system, which consists of the 13 kDa GEF Mss4 and the monomeric G protein Rab3A involved in presynaptic neurotransmission, was chosen to study the mechanism of G-protein regulation. Structure of Rab3A at high resolution The 2.0 Å crystal structure of Rab3A, bound to a non-hydrolyzable GTP-analog (GppNHp), enables a detailed description of the structural determinants that stabilize the active conformation and regulate GTPase activity within the Rab family. Although the overall structure is similar to that of GppNHp-bound Ras and other GTPases, localized but significant differences are observed in the vicinity of the conformational switch regions and the α3/β5 loop. The active conformation is stabilized primarily by extensive hydrophobic contacts between the switch I and II regions. Novel interactions with the γ phosphate, mediated by serine residues in the P-loop and switch I region, impose stereochemical constraints on the mechanism of GTP hydrolysis and provide a structural explanation for the broad range of GTPase activities within the Rab family. Residues implicated in interactions with effectors and regulatory factors map to a common face of the protein. The asymmetric distribution of charged and non-polar residues suggests a plausible orientation with respect to vesicle membranes that would position predominantly hydrophobic surfaces to interact with membrane-associated effectors and regulatory factors. Thus, the structure of Rab3A establishes a framework for understanding the molecular mechanisms underlying the function of Rab proteins in vesicle trafficking. High resolution structure of Mss4 and structure-based mutagenesis Activation of monomeric Rab GTPases, which function as ubiquitous regulators of intracellular membrane trafficking, requires the catalytic action of guanine nucleotide exchange factors. Mss4, an evolutionarily conserved Rab exchange factor, promotes nucleotide release from exocytic but not endocytic Rab GTPases. Chapter III describes the results of a high resolution crystallographic and mutational analysis of Mss4. The 1.65 Å crystal structure of Mss4 reveals a network of direct and water mediated interactions that stabilize a partially exposed structural sub-domain derived from four highly conserved but non-consecutive sequence elements. The conserved sub-domain contains the invariant cysteine residues required for Zn2+ binding as well as the residues implicated in the interaction with Rab GTPases. A strictly conserved DΦΦ motif, consisting of an invariant aspartic acid residue (Asp73) followed by two bulky hydrophobic residues (Met74 and Phe75), encodes a prominently exposed 310 helical turn in which the backbone is well ordered but the side chains of the conserved residues are highly exposed and do not engage in intramolecular interactions. Substitution of any of these residues with alariine dramatically impairs exchange activity towards Rab3A, indicating that the DΦΦ motif is a critical element of the exchange machinery. In particular, mutation of Phe75 results in a defect as severe as that observed for mutation of Asp96, which is located near the zinc binding site at the opposite end of Rab interaction epitope. Despite severe defects, however, none of the mutant proteins is catalytically dead. Taken together, the results suggest a concerted mechanism in which distal elements of the conserved Rab interaction epitope cooperatively facilitate GDP release. The basis for selective recognition of exocytic Rab family GTPases by Mss4 Rab3A is involved in Ca2+ -dependent exocytosis and neurotransmitter release. Mss4, an evolutionarily conserved Rab exchange factor, promotes nucleotide release from exocytic RabGTPase (Rab1, Rab3A, Rab8, and Rab10, Sec4 and Ypt1) but not endocytic Rab GTPases (Rab2, Rab4, Rab5, Rab6, Rab9 and Rab11). To understand the basis for selective recognition of exocytic Rab family GTPases by Mss4, a structure based mutagenesis study of Rab3A was conducted. Three residues in Rab3A (Phe51, Val61 and Thr89) were found to be critical for interaction with Mss4. Phe51 is located at the N- terminus of the switch region, adjacent to the Mg2+ and nucleotide binding site. Val61 in the β2 strand and Thr89 in the switch II region flank a triad of hydrophobic residues that is conserved in the Rab family. These residues comprise critical determinants underlying the broad specificity of Mss4 for exocytic Rab family proteins. In addition to determining the high resolution crystal structures of Rab3A and Mss4, the experiments described above identify critical structural determinants for the exchange activity of Mss4 and provide insight into the selective recognition of Mss4 by exocytic Rab GTPases.

Cytokinesis is the final event of the cell division cycle and results in physical and irreversible separation of a mother cell into two daughter cells. Cytokinesis must only occur after chromosomes have segregated during mitosis to ensure each daughter cell receives the proper complement of genetic material. Failure to execute normal cytokinesis can result in aneuploidy and/or polyploidy, a hallmark of many cancers. Cytokinesis occurs mechanically through constriction of an actin-myosin based contractile ring, while initiation of ring constriction is temporally and spatially mediated by complex signaling networks. It is absolutely crucial that cytokinesis is tightly coordinated with the cell cycle in order to preserve the fidelity of cell division. We hypothesized that to achieve such tight control of cytokinesis, cells may utilize both promotional and inhibitory signals, however how cells maintained this control was poorly understood. The goal of this thesis was to characterize how cells regulate signaling of cytokinesis, both positively and negatively, during cell division using the fission yeast Schizosaccharomyces pombe as a model organism. Two approaches were employed. (1) We first sought to characterize the positive timing mechanism that signals cytokinesis though a detailed investigation of Sid1p, a protein kinase essential for activation of ring constriction. (2) Secondly, we sought to define how cells negatively regulate cytokinesis through investigation of Dma1p, a spindle checkpoint protein implicated in inhibition of cytokinesis. Our results reveal a conserved signaling network, termed the Septation Initiation Network (SIN), of which Sid1p is an intermediate component, that controls temporal and spatial regulation of cytokinesis. We found Sid1p is additionally controlled by Cyclin Dependent Kinase activity, uncovering an important link between mitotic events and initiation of cytokinesis. Furthermore, we found that aberrant SIN activation can override a microtubule-damage-induced spindle checkpoint arrest. This effect is counteracted by Dma1p, which normally inhibits the SIN during checkpoint activation to preserve cell viability until damage is repaired. We conclude that signaling cytokinesis is tightly coordinated with mitosis in S. pombe by positive signals acting through Sid1p and the SIN, and under certain conditions, negative signals acting through Dma1p. Considering the conservation of cell cycle regulators in the eukaryotic kingdom, it is likely that similar mechanisms to control cytokinesis exist in humans.

During virus infections, activation of CD8 T cells takes place in secondary lymphoid organs including spleen and lymph nodes. The kinetics of the T cell response in lymphoid tissues has been clearly studied. However, a large number of virus-specific T cells disseminate into various nonlymphoid tissues. As reservoirs for effector and memory cells, nonlymphoid organs play an important role for defending against infections. T cell responses in nonlymphoid organs may differ from lymphoid organs. T cell repertoire usage in lymphoid and nonlymphoid tissues was studied in an acute lymphocytic choriomeningitis virus (LCMV)-infected murine model. The hierarchy of CD8 T cell specificities was examined with cytotoxic T lymphocyte (CTL) sodium 51 chromate (51Cr) release assays and intracellular interferon (IFN)γ assays. T cell receptor (TCR) repertoire usage was determined by complementarity determining region (CDR)3 length spectratyping analysis. Both T cell specificity and TCR repertoire usage revealed some similarities and differences between several organs. Within an epitope-specific CD8 T cell population, the TCR repertoire usage was similar in different organs of the same mouse, but highly heterogeneous between individual mice with genetically identical backgrounds. A very restricted CD4 TCR repertoire was observed in BALB/c mice after secondary respiratory syncytial virus (RSV) infection. Most of the CD4 T cells of BALB/c mice pre-immunized with RSV glycoprotein (GP) predominantly express Vβ14 TCR with discrete oligoclonal CDR3 regions. Depletion of Vβ14 CD4 T cells dramatically reduced immunopathology. The apoptotic phenotype of LCMV-specific CD8 T cells was studied in various lymphoid and nonlymphoid tissues during acute and memory stages of infections. Peripheral tissues (peritoneal cavity (PEC), fat pad, and lung) reacted with a much lower frequency with the early apoptotic marker Annexin V than those in spleen and lymph nodes. This was not due to a TCR-based selection because similar TCR spectratypes were seen in different organs. Activated lymphoid and nonlymphoid T cells from LCMV GP33 transgenic mice, which have identical TCR α and β chains on all T cells, had differential Annexin V binding. When incubated shortly in vitro, most Annexin V+ T cells rapidly fragmented their DNA and became terminal transferase-mediated dUTP nick end-labeling positive (TUNEL+), while much fewer Annexin V- cells became TUNEL+. Therefore, those Annexin-V+ cells were truly in a pre-apoptotic stage. The differential spontaneous apoptosis in different tissues is independent of several death/survival-related molecules, including Fas/Fas ligand (FasL), turner necrosis factor (TNF)α, interleukin (IL-15), perforin, B cell lymphoma (Bcl)-2 and independent of virus tropism. I further investigated the significance of the high Annexin V reactivity of lymphoid T cells. Pre-apoptotic cells were prevented from fragmenting their DNA by anti-CD3 or IL-2 stimulation in vitro. However, this pre-apoptotic phenotype precluded generation of memory. Annexin V reactive cells did not give rise to long-lived memory after being transferred into naïve hosts. The pre-apoptotic phenotype is also an intrinsic property of the epitope. Different proportions of apoptotic cells were found in LCMV effector and, memory T cells specific to two different epitopes, nucleoprotein (NP)396 and GP33. Higher Annexin V reactivity of NP396-specific CD8 T cells was independent of virus tropism and duration of encounter with antigen. Higher expression of IL-7R was found in peripheral, Annexin V- and GP33-specific CD8 T cells, indicating that IL-7-dependent signals may inhibit apoptosis. Nonlymphoid T cells were more resistant than lymphoid T cells to activation-induced cell death (AICD). When stimulated with anti-CD3 in vitro for 40 hours (hr), a significantly reduced number of splenic transgenic T cells were recovered with much higher frequency of Annexin V reactivity and TUNEL staining than transgenic T cells from PEC. Consistent with the finding that Fas and FasL regulates AICD, a much lower expression of Fas and FasL was observed in PEC and lung transgenic T cells than spleen and lymph nodes after short time stimulation. FasL blockage largely increased cell-number recovery and reduced Annexin V and TUNEL staining of spleen transgenic T cells. Interestingly, the leukocyte environment played an important role of deciding the fate of transgenic T cells. When placing activated spleen transgenic T cells with excess infected PEC cells, spleen transgenic cells rapidly reduced their Annexin V staining and TUNEL staining and were recovered with greater number after stimulation. Vice versa, PEC transgenic T cells became Annexin V and TUNEL positive with lower numbers of cells recovered when placed with excess splenocytes. Less detection of Annexin V+ cells in peripheral tissues was not due to rapid phagocytosis by macrophages, because Cytochalasin D, which can inhibit phagocytosis, did not induce equal amount of pre-apoptotic cells in spleen and PEC. This reduced death in the periphery may contribute to the long-term maintenance of nondividing nonlymphoid memory T cells, enabling them to efficiently function without being driven into apoptosis. Overall, this study characterizes in detail the different T cell repertoire usage and apoptosis of virus-specific T cells based on their organ localization and specificities and helps to better understand T cell immunity after infections and vaccine design.

DNA immunization, inoculation with an antigen-expressing plasmid DNA, is a new method for generating an antigen-specific immune response. At the time these investigations began, very little was known about the immune response produced by DNA vaccines. Thus, the first aim of our studies was to perform a detailed examination of the antibody response generated by DNA immunization with an influenza hemagglutinin (HA)-expressing DNA in BALB/c mice. Using several different routes and methods of DNA immunization, we observed a number of findings. Although all three forms of DNA immunization elicited strong anti-HA antibody responses, i.m. and i.d. saline DNA immunization required approximately 100 times more DNA than a gene gun DNA immunization to raise an equivalent titer of anti-HA antibody. Indeed, as little as one inoculation and one boost by gene gun of 0.0004 μg of DNA produced a measurable antibody response in 50% of mice. Unexpectedly, we found the isotype of the antibody response differed among groups of mice immunized by different forms of DNA immunization. Intramuscular and i.d. saline DNA immunization produced predominantly an IgG2a anti-HA antibody response, whereas gene gun DNA immunization elicited mostly an IgG1 anti-HA antibody response. Considering that IgG2a and IgG1 antibody isotypes were known to correlate with Th1 and Th2 immune responses, respectively, we analyzed the type of immune responses produced by i.m., i.d., and gene gun DNA immunization. We found that i.m. and i.d. saline DNA immunization produced a Th1 predominant cellular immune response. In contrast, gene gun DNA immunization produced a Th2 cellular immune response. The differences in the type of immune responses were found to be due to the method of DNA immunization, and not due to the route of DNA inoculation. A gene gun DNA immunization of muscle produced the same IgG1, Th2 immune response as a gene gun DNA immunization of skin, while a saline DNA immunization of muscle and skin produced mostly an IgG2a, Th1 immune response. Each method of DNA immunization created good memory Th cell responses. The type of immune response created by an initial DNA immunization remained fixed even after multiple boosts with the identical method of DNA immunization, following a boost with the alternative method of DNA immunization, or after a viral challenge. The differentiation of naive Th cells into Th1 or Th2 cells depends on a variety of factors. We performed many experiments to elucidate which factors played a role in the generation of Th1 or Th2 immune responses following saline DNA immunization and gene gun DNA immunization. DNA dose response studies revealed the use of different doses of DNA between groups of saline DNA and gene gun DNA immunized mice did not account for the differentiation of distinct Th cell subsets. Cytokine production inducible by a number of factors inherently associated with either saline DNA or gene gun DNA immunization did not affect Th differentiation. For instance, contamination of plasmid DNA with lipopolysaccharide did not account for differences in the immune response. Immunostimulatory CpG sequences did not affect Th differentiation following DNA immunization, but they did enhance the IgG2a antibody response to coinoculated HA protein. Finally, cotransfection of IFNγ or IL-4 expressing plasmids with an HA-expressing plasmid by gene gun inoculation or as a saline DNA injection did not shift the type of immune response in a Th1 or Th2 direction, respectively. Thus, it appeared that increased cytokine stimulation was not responsible for selective Th subset differentiation. One factor related to the method of DNA immunization did seem to correlate with Th1 differentiation. Deposition of plasmid DNA extracellulary by saline DNA injections (as opposed to intracellular DNA delivery by gene gun) may have stimulated Th1 immune responses. Manipulating a gene gun DNA immunization to deliver DNA to the dermis (and thus extracellularly) shifted the immune response from that of a Th2 type to a mixed Th1/Th2 type. Furthermore, evidence was gathered demonstrating that pDNA can interact with cell surface molecules and that specific sequences in pDNA can act as a ligand and bind to molecules. Taken together, our data led us to propose a new model for Th1 differentiation following saline DNA immunization. We believe extracellular pDNA binds to an APC cell surface molecule which activates the cell. The activated APC preferentially stimulates naive Th cells to differentiate into Th1 cells. Finally, studies using a variety of mice differing in their genetic backgound and MHC genotype demonstrated the generality of our findings regarding i.m. saline DNA inoculations of an HA-expressing pDNA. Saline DNA immunization produced IgG2a, Th1-predominant immune responses independent of the genetic background and MHC genotype of the mice. In contrast, the type of immune response elicited by a gene gun DNA immunization was dependent on the MHC genotype of mice. Thus the type of immune response produced by gene gun DNA immunization probably depends on the specific antigen (and its effect on MHC-peptide/TcR interaction and signaling) and is less likely due to any inherent feature associated with the process of gene gun DNA delivery.