The synapse is the primary locus of cell-cell communication in the nervous system. The elaboration of a functional synapse requires both a specialized structure and an efficient communication system. For my thesis work, I studied proteins implicated in each of these functions: the structural molecules dystroglycan and dystrophin, and the signaling elements Insulin Receptor Substrate p58/53 and insulin receptor. The α/β-dystroglycan complex, believed to be the heart of cellmatrix adhesion in muscle and other tissues, provides a link between dystrophin, a cytoskeletal protein at the base of the muscle cell's Dystrophin Associated Protein Complex, and the extracellular matrix. In addition, dystrophin is found at central synapses, tightly associated with the postsynaptic density. The absence of dystrophin and the secondary loss of its associated proteins causes the genetic disease Duchenne Muscular Dystrophy. DMD affects both muscle and brain, causing a severe muscular dystrophy and lower IQs than control groups. In the first portion of my thesis work, I sought to determine the role of dystroglycan, dystrophin's peripheral partner, at central synapses. I probed Northern blots of brain regions to delineate the distribution of brain β-dystroglycan mRNA and to uncover any β-dystroglycan-related transcripts in brain. Then, using subcellular brain fractions, and cultured hippocampal neurons, I determined that whereas α-dystroglycan is associated with central synapses, β-dystroglycan is not. This discovery is surprising, and differs from the finding that dystrophin and α- and β-dystroglycan colocalize at the presynaptic membrane of retinal photoreceptors. In the course of the above mentioned work, using the anti-β-dystroglycan antiserum Ab98, I discovered a pair of proteins that were tightly associated with the postsynaptic density. These polypeptides of 58 kDa and 53 kDa (p58/53) were highly enriched in postsynaptic density (PSD) fractions from rat cerebral cortex, hippocampus, and cerebellum. In pursuit of a potential synapse-specific dystroglycan relative, I purified p58 and p53 by a combination of hydrophobic interaction chromatography and two-dimensional gel electrophoresis. Mass spectroscopy and peptide microsequencing revealed that p58/53 is identical to the insulin receptor tyrosine kinase substrate p58/53 (IRSp53). Whereas IRSp58/53 has no significant homology to β-dystroglycan other than the one span of peptides that confers its antibody cross-reactivity, its localization to the PSD newly implicates insulin signaling at synapses. Analysis of IRSp58/53 mass profiles, peptides, and mRNA indicated that IRSp58 and IRSp53 are the product of the same coding sequence. Immunolocalization showed that IRSp58/53 is expressed in the synapserich molecular layer of the cerebellum. Immunostaining of cultured hippocampal neurons showed that both IRSp58/53 and insulin receptor are highly concentrated at synapses. Like IRSp58/53, insulin receptors are a component of the PSD fraction. Together, these data suggest that the synapse is a specialized site for insulin signaling in the brain.

The thesis research was designed to study the characteristics of two important physiological processes in smooth muscle: Ca2+ transport mediated by the plasmalemmal Ca2+-ATPase and muscarinic receptor-G protein interactions. In resting smooth muscle, several Ca2+ extrusion or sequestration processes offset the passive inward leak of Ca2+. Although biochemical evidence suggests that the plasmalemmal Ca2+ pump plays a key role in this process, the precise role of this enzyme could not be proven until a reliable estimate of the inward Ca2+ leak was measured. Recent studies using dispersed smooth muscle cells from the toad stomach provided an estimate of the basal transmembrane Ca2+ flux rate; thus, we examined the transport capacity of the plasmalemmal Ca2+pump in this tissue. Gastric smooth muscle tissue was disrupted by homogenization and nitrogen cavitation. Membranes enriched 20 fold for plasma membrane markers were obtained using differential centrifugation and purification by flotation on discontinuous sucrose gradients. The membrane vesicles exhibited an ATP-dependent 45Ca uptake that was insensitive to azide or oxalate but sensitive to stimulation by calmodulin or inhibition by orthovanadate and the calmodulin antagonists trifluoperazine (TFP) or calmidazolium (CMZ). 45Ca accumulated in the presence of ATP was rapidly released by Ca2+ ionophore but not by agents that stimulate Ca2+ release from the sarcoplasmic rettculum (caffeine, inositol trisphosphate, GTP). However, both CMZ and TFP evoked a Ca2+ release that was comparable to that observed in the presence of Ca2+ ionophore, suggesting that these compounds have profound effects on membrane Ca2+permeability. 45Ca transport exhibited a high affinity for Ca2+ (KD 0.2 μM) and a high transport capacity, producing a > 12,000-fold gradient for Ca2+and a transmembrane flux rate at least 3-fold greater than that observed in resting smooth muscle cells. As a first step toward understanding the biochemical basis for the diversity of muscarinic cholinergic actions on smooth muscle, we examined the distribution of muscarinic receptor subtypes and coupling to guantne nucleotide-binding (G) proteins in airway and gastric smooth muscle. Receptor subtypes were classified in membranes prepared from bovine trachea and toad stomach based on the relative abilities of the selective antagonists pirenzepine (M1), AF-DX 116 (M2) and 4-DAMP (M3) to displace the binding of nonselective antagonist [3H]QNB (quinuclidinyl benzilate). Based on the binding profiles for these antagonists, it was concluded that both smooth muscle types contain a mixture of M2 and M3 subtypes. In trachea the majority of receptors (86%) were M2, whereas in stomach the majority of receptors (88%) were M3. The displacement of [3H]QNB binding by the agonist oxotremorine indicated a mixed population of high affinity (KD = 4 nM) and low affinity (KD = 2-4 μM) binding sites. The addition of GTPγS abolished all high affinity agonist binding, suggesting that coupling of the receptors to G proteins may confer high affinity. Reaction of membranes with pertussis toxin in the presence of [32P]NAD caused the [32P]-labelling of a ~ 41 kD protein in both gastric and tracheal smooth musc1e. Pretreatment of the membranes with pertussis toxin and NAD completely abolished high affinity agonist binding in gastric smooth muscle, but produced little if any decrease in high affinity agonist binding in trachea. We conclude that, although muscarinic receptor activation leads to the elevation of intracellular Ca2+ and to contraction of both airway and gastric smooth muscle, the dissimilar distributions of receptor subtypes and distinct patterns of coupling to G proteins may indicate that each smooth muscle type uses different receptor-G protein interactions to regulate intracellular signalling pathways.

This thesis investigates the role of B cells as antigen-specific antigen-presenting cells (APC) in self tolerance to low concentrations of soluble self proteins and in acquired tolerance to low doses of soluble foreign protein antigens. Experiments were performed in normal and B cell-deficient animals, and tolerance induction was measured by T cell proliferation assays. T cell proliferation was reduced in B cell-deficient mice, indicating that B cells may be involved in efficient activation of naive T cells in response to protein antigen both in vivo and in vitro. To study acquired tolerance induced by low doses of soluble foreign protein antigen, normal and B cell-deficient adult mice were injected intravenously with repeated low doses (10 μg) of deaggregated ovalbumin (OVA), and then challenged with OVA in complete Freund's adjuvant. In animals treated with deaggregated OVA, the in vitro proliferative responses of LN T cells to OVA were significantly reduced, and production of the Th1 cytokine, IFN-γ, in response to OVA was lost. This occurred in both normal and B cell-deficient treated animals, indicating that B cell antigen presentation was not required for this phenomenon. B cells were also unnecessary for self tolerance of T cells to the transgenic self antigen, hen egg lysozyme (HEL), in a transgenic mouse strain with very low serum lysozyme concentration. Partial low zone tolerance induced by deaggregated, low-dose OVA was selective for the Th1 response, as measured by in vitro proliferation and IL-2 and IFN-γ production, because antibody responses of normal mice to this T cell-dependent antigen were largely unaffected. Both treated and untreated animals produced equivalent titers of anti-OVA antibodies, predominantly of the IgG1 and IgG2b isotypes, following challenge with OVA in complete Freund's adjuvant. Tolerance to low levels of the transgenic HEL self protein in mice expressing different MHC molecules was also addressed. Transgenic mice that were H-2b/b in the class II region were not tolerant to the transgenic self protein, whereas transgenic mice of the H-2b/k were tolerant.

The chicken hepatoma cell line, LMH, was evaluated with respect to its usefulness for studies of the regulation of heme metabolism. Levels of δ-aminolevulinate synthase mRNA arid accumulation of porphyrins were used to evaluate the heme biosynthetic pathway. Regulation of heme oxygenase-1 by known inducers was used as a measure of heme degradation. The induction of heme oxygenase-1 by sodium arsenite was characterized. AP-1 transcription factor elements and MAP kinase signal transduction pathways that modulate expression of endogenous heme oxygenase-1 and transfected heme oxygenase-1 reporter gene constructs in response to arsenite were delineated. In initial studies, the drug glutethimide was used alone or in combination with ferric nitrilotriacetate to induce δ-aminolevulinate synthase mRNA. Levels of porphyrins, intermediates in the heme biosynthetic pathway, and levels of δ-aminolevulinate synthase mRNA were increased by these treatments in a manner similar to those previously observed in the widely used model system, primary chick embryo liver cells. The iron chelator, deferoxamine, gave a characteristic shift in the glutethimide induced porphyrin accumulation in primary hepatocytes, but was found to have no, effect on LMH cells. Heme mediated repression of δ-aminolevulinate synthase mRNA levels was similar among primary hepatocytes and LMH cells. Heme oxygenase-1 was regulated by heme, metals, heat shock, and oxidative stress-inducing chemicals in LMH cells. Heat shock induction of heme oxygenase-1 mRNA levels was observed for the first time in primary chick embryo liver cells. These data supported the further use of LMH cells to elucidate mechanisms responsible for modulating heme oxygenase-1 gene expression in response to inducers. The remainder of the studies focused on the role of heme oxygenase-1 as a stress response protein. The oxidative stress inducer, sodium arsenite was used to probe the cellular mechanisms that control the expression of heme oxygenase-1. A series of promoter-reporter constructs were used to search the heme oxygenase-1 promoter for arsenite responsive elements. Several activator protein-1 (AP-1) transcription factor binding elements were identified by computer sequence analysis. Three of these sites, located at -1578, -3656, and -4597 base pairs upstream of the transcription start site, were mutated. The arsenite responsiveness of the reporter constructs containing mutated AP-1 elements was less than that of the same constructs containing wild type AP-1 elements. At least part of the arsenite-mediated induction of heme oxygenase-1 required the activity of AP-1 transcriptional elements. The MAP kinase signal transduction pathways and heme oxygenase-1 are activated by similar stimuli, including cellular stress. MAP kinases have been shown to exert control over gene expression through effects on the AP-1 family of transcription factors. The MAP kinases ERK, JNK, and p38 were activated by arsenite in LMH cells. Constitutively activated components of the ERK and p38 pathways increased expression of heme oxygenase-1 promoter-luciferase reporter constructs. Arsenite-mediated induction of heme oxygenase-1 was blocked by dominant negative ERK or p38 pathway components, and by specific inhibitors of MEK (upstream ERK kinase) or p38. In contrast, reporter gene expression was unchanged in the presence of constitutively activated JNK pathway components. Dominant negative JNK pathway components had no effect on arsenite induced heme oxygenase-1 gene activity. In summary, LMH cells were characterized as a new model system for the study of heme metabolism. This cell line was then used to delineate promoter elements and signaling pathways involved in the arsenite responsiveness of heme oxygenase-1 gene expression. Three AP-1 transcription factor binding sites in the heme oxygenase-1 promoter region were required for responsiveness to arsenite. The MAP kinases ERK and p38 were shown to play an integral role in arsenite-mediated induction of heme oxygenase-1. These studies elucidate one facet of heme oxygenase-1 regulation, and provide tools that will be useful in delineating additional regulatory mechanisms.

Measles (MV) virions, like those of other enveloped viruses, enter cells by fusing their lipid membranes with those of the target host cells. Additionally, infected tissues often possess giant multinucleate cells, known as syncytia, which are formed by fusion of infected cells with uninfected neighbors. Expression of both the MV attachment (H) and fusion (F) proteins is required for membrane fusion. MV H mediates receptor binding in order to bring the two membranes into close proximity prior to F activation and is thought to trigger F activation through a specific interaction between the two proteins. Although measles H and F are efficiently transported to the cell surface when expressed independently, evidence has been reported in support of an intracellular interaction between the two proteins that can be detected using an ER co-retention approach. However, it was not determined if the putative co-retention was specific to the two measles glycoproteins, as is their ability to complement each other for efficient fusion promotion. Thus, in this thesis, the formation of an intracellular complex between MV H and F was re-examined. Consistent with the formation of an intracellular complex, cell surface expression and receptor binding of untagged wt MV H is slightly reduced by co-expression of an excess of ER-tagged MV F compared to co-expression with wt F. However, the reduction in surface expression is non-specific in that it can also be induced with heterologous proteins of NDV, which lack significant homology with those of MV. Although this approach did not detect a specific intracellular interaction between MV H and F, it cannot be ruled out that there is a weak association of the proteins that is undetectable by this method. This led to the use of an alternative approach to investigate the cellular site(s) of interaction between the measles H and F proteins. Consistent with a cell surface interaction between MV H and F, the combination of surface biotinylation and co-immunoprecipitation detects formation of a virus-specific H-F complex. Approximately, 21% of the total amount of MV H at the cell surface can be captured with MV F using an antibody against the latter protein. Two complementary approaches were used to address the relationship between this cell surface interaction and receptor recognition by MV H. First, the proteins were co-immunoprecipitated from the surface of Chinese hamster ovary (CHO) cells, which do not express either MV receptor, CD46 or CD150. Similar levels of MV H can be co-immunoprecipitated with F from the surfaces of parental CHO cells and stably transfected cells that express, human CD46 (CHO-CD46), indicating that binding to CD46 is not the trigger for the H-F interaction. Second, MV H proteins, carrying mutations that dramatically reduce CD46 binding, were shown to co-immunoprecipitate efficiently with F from the surface of HeLa cells. Significantly, these results indicate that MV H and F interact in the absence of, and thus prior to, receptor binding. This is in direct contrast to the NDV HN-F cell surface interaction, which is thought to be triggered by receptor binding. Identification of the domains of the para myxovirus attachment and fusion proteins that mediate membrane fusion activities is an essential part of understanding the mechanism of fusion. As a result of the H-F interaction prior to receptor binding, MV H attachment to its cellular receptor must result in conformational changes that trigger activation of the F protein. Site-directed mutagenesis analyses of two regions of MV H indicate that a HR domain in the stalk of the attachment protein is essential to the ability of H to activate F. However, either it is not the only region of H that interacts with F or it is indirectly involved in F activation because mutations in the HR do not disrupt MV H-F complex formation at the cell surface. Additionally, the functional interaction between MV H and F may be mediated, at least in part, by Loop 1 of the amino terminus of the C-rich region of the fusion protein. However, the exact role of this region of the F protein in fusion promotion remains to be determined. Importantly, the cell surface interaction between MV H and F proteins appears to be mediated by more that one region of each protein. In contrast to NDV, in no case has a definitive link between any single amino acid difference in MV H or F and an inability to form the cell surface H-F complex been established. In conclusion, the data presented in this dissertation support a model of measles membrane fusion in which the Hand F proteins form a complex prior to receptor recognition. This complex may hold F in its meta-stable pre-fusion state until binding of H to receptors at the cell surface triggers dissociation of the complex, releasing F to assume its fusogenic form. Importantly, these data also indicate that, although paramyxoviruses may all use the same general process. for promotion of membrane fusion, the mechanism may vary in multiple aspects. A more complete understanding of the means by which measles promotes membrane fusion may direct the development of specific strategies aimed at interfering with the early stages of infection.

Knowledge of how a polypeptide folds from a space-filling random coil into a biologically-functional, three-dimensional structure has been the essence of the protein folding problem. Though mechanistic details of DNA transcription and RNA translation are well understood, a specific code by which the primary structure dictates the acquisition of secondary, tertiary, and quarternary structure remains unknown. However, the demonstrated reversibility of in vitroprotein folding allows for a thermodynamic analysis of the folding reaction. By probing both the equilibrium and kinetics of protein folding, a protein folding mechanism can be postulated. Over the past 40 years, folding mechanisms have been determined for many proteins; however, a generalized folding code is far from clear. Furthermore, most protein folding studies have focused on monomeric proteins even though a majority of biological processes function via the association of multiple subunits. Consequently, a complete understanding of the acquisition of quarternary protein structure is essential for applying the basic principles of protein folding to biology. The studies presented in this dissertation examined the folding and assembly of two very different multimeric proteins. Underlying both of these investigations is the need for a combined analysis of a repertoire of approaches to dissect the folding mechanism for multimeric proteins. Chapter II elucidates the detailed folding energy landscape of HIV-1 protease, a dimeric protein containing β-barrel subunits. The folding of this viral enzyme exhibited a sequential three-step pathway, involving the rate-limiting formation of a monomeric intermediate. The energetics determined from this analysis and their applications to HIV-1 function are discussed. In contrast, Chapter III illustrates the association of a coiled coil component of L. terrestriserythrocruorin. This extracellular hemoglobin consists of a complex scaffold of linker chains with a central ring of interdigitating coiled coils. Allostery is maintained by twelve dodecameric hemoglobin subunits that dock upon this scaffold. Modest association was observed for this coiled coil, and the implications of this fragment to linker assembly are addressed. These studies depict the complexity of multimeric folding reactions. Chapter II demonstrates that a detailed energy landscape of a dimeric protein can be determined by combining traditional equilibrium and kinetic approaches with information from a global analysis of kinetics and a monomer construct. Chapter III indicates that fragmentation of large complexes can show the contributions of separate domains to hierarchical organization. As a whole, this dissertation highlights the importance of pursuing mulitmeric protein folding studies and the implications of these folding mechanisms to biological function.

In humans, Bloom’s Syndrome is caused by a mutation of the RecQ helicase BLM. Patients with Bloom’s Syndrome exhibit a high amount of genomic instability which results in a high incidence of cancer. Though Bloom’s Syndrome has been intensively studied, there are still many questions about the function of BLM which need to be answered. While it is clear that loss of BLM increases genomic instability, the other effects of genomic instability on the organism aside from cancer such as a potential effect on aging, have yet to be elucidated. In Chapter II, I identify new phenotypes in the C. elegans ortholog of BLM, him-6. him-6 mutants have an increased rate of cell death, a mortal germ line phenotype, and an increased rate of mutations. Upon further examination of the mutator phenotype, it was determined that the increased rate of mutations was caused by small insertions and deletions. The mutator phenotype identified in him-6 mutants closely mimics the cellular phenotype seen in Bloom’s Syndrome cells. This indicates that HIM-6 may behave in a similar fashion to BLM. In addition to the mutator phenotype, it was found that loss of him-6causes a shortened life span. This may provide evidence that there is a link between genomic stability and aging. In Chapter III, I identify a new role for the transcription factor DAF-16. DAF-16 in C. elegans has been intensively studied and regulates a wide variety of pathways. In this chapter, I demonstrate via the well established unc-93 assay that loss of daf-16 causes a subtle mutator phenotype in C. elegans. This indicates that DAF-16 may play a role in suppression of spontaneous mutation. When I examined other classic genomic instability phenotypes, I found at 25°C, the number of progeny in the DAF-16 mutants was significantly reduced compared to wild type worms. Additionally, I demonstrate daf-16(mu86)has a cell death defect. This study identifies several new phenotypes caused by a loss of him-6. These phenotypes provide further evidence that loss of him-6 causes genomic instability. In addition, this study also demonstrates that him-6 has a shortened life span which may be due to genomic instability. Secondly, this study identifies a new role for DAF-16 in preventing the occurrence of spontaneous mutations. This may indicate a novel function for DAF-16 in maintaining genomic stability.

Stimulation of CD40 on APCs through CD40L expressed on helper CD4+ T cells activates and “licenses” the APCs to prime CD8+ T cell responses. While other stimuli, such as TLR agonists, can also activate APCs, it is unclear to what extent they can replace the signals provided by CD40-CD40L interactions. In this study, we used an adoptive transfer system to re-examine the role of CD40 in the priming of naïve CD8+ T cells. We find an approximately 50% reduction in expansion and cytokine production of TCR-transgenic T cells in the absence of CD40 on all APCs, and on dendritic cells in particular. Moreover, CD40-deficient and CD40L-deficient mice fail to develop endogenous CTL responses after immunization and are not protected from a tumor challenge. Surprisingly, the role for CD40 and CD40L are observed even in the absence of CD4+ T cells; in this situation, the CD8+T cell itself provides CD40L. Furthermore, we show that although TLR stimulation improves T cell responses, it cannot fully substitute for CD40. We also investigated whether CD40-CD40L interactions are involved in the generation, maintenance, and function of memory CD8+ T cells. Using a virus infection system as well as a dendritic cell immunization system, we show that the presence of CD40 on DCs and other host APCs influences the survival of activated effector cells and directly affects the number of memory CD8+ T cells that are formed. In addition, memory CD8+ T cell persistence is slightly impaired in the absence of CD40. However, CD40 is not required for reactivation of memory CD8+ T cells. It seems that CD40 signals during priming also contribute to memory CD8+ T cell programming but this function can be independent of CD4+T cells, similar to what we showed for primary responses. Altogether, these results reveal a direct and unique role for CD40L on CD8+ T cells interacting with CD40 on APCs that affects the magnitude and quality of primary as well as memory CD8+ T cell responses.

In response to external stimuli, many intracellular signaling proteins undergo dynamic changes in localization to the plasma membrane. Using the Saccharomyces cerevisiaemating pathway as a model, I investigated the molecular interactions that govern plasma membrane localization of signaling proteins, and how the plasma membrane compartmentalization of a signaling complex influences the overall signaling behavior of the pathway. Signaling proteins often consist of multiple interaction domains that collectively dictate their localization and function. Ste20 is a p21-activated kinase (PAK) that functions downstream of the Rho-type GTPase Cdc42 to activate several mitogen-activated protein (MAP) kinase pathways in budding yeast, including the mating pathway. I identified a short domain in Ste20 that directly binds to membrane lipids via electrostatic interaction. A mutation in this domain abolishes both the localization and function of Ste20. Thus, the previously known Cdc42 binding is necessary but not sufficient; instead, direct membrane binding by Ste20 is also critical. By replacing this domain with heterologous membranebinding domains, I demonstrated that phospholipid specificity is not essential in vivo. Functionally important short membrane-binding domains were also found in the Cdc42 effectors Gic1 and Gic2, indicating that generic membrane binding can work in concert with the CRIB domain to regulate activation of Cdc42 targets. These results underscore the importance of cooperation between protein-protein and protein-membrane interaction in achieving proper localization of signaling proteins at the cell cortex. At the system level, MAP kinase cascades can be graded or switch-like. The budding yeast mating pathway exhibits a graded response to increasing levels of pheromone. Previously the scaffold protein Ste5 was hypothesized to contribute to this graded response. To test this idea, I activated the pathway in a variety of ways and measured the response at the single cell level. I found that the graded response is not perturbed by the deletion of negative regulators of the pathway whereas the response became switch-like when the pathway was activated by a crosstalk stimulus that bypasses the upstream components. Interestingly, activation of the pathway in the cytoplasm using the graded expression of MAPKKK resulted in an ultrasensitive response. In contrast, activation of the pathway at the plasma membrane using the graded expression of membranetargeted active pathway components remained graded. In these settings, the scaffold protein Ste5 increased ultrasensitivity when limited to the cytosol; however, if Ste5 was allowed to function at the plasma membrane, signaling was graded. The results suggest that, in the mating pathway, the inherently ultrasensitive MAPK cascade is converted to a graded system by the scaffoldmediated assembly of signaling complexes at the plasma membrane. Therefore, the plasma membrane localization of Ste5 helps shape the input-output properties of the mating MAPK pathway in a manner that is suitable for the biology of mating. Taken together, this thesis underscores the importance of plasma membrane localization during mating pathway signaling in yeast. The examples described here provide further appreciation of how multiple interaction domains can function together to achieve specific targeting of the signaling proteins, as well as advances in understanding the role of scaffold proteins in modulating signaling behavior to promote graded signaling at the plasma membrane.

Dengue virus (DENV) infects 50-100 million people worldwide every year and is the causative agent of dengue fever (DF) and the more severe dengue hemorrhagic fever/dengue shock syndrome (DHF/DSS). There are four genetically and immunologically distinct DENV serotypes (DENV-1, DENV-2, DENV-3, and DENV-4). Evidence suggests that an increased risk for DHF/DSS during secondary infection with a heterologous DENV serotype is due to an immunopathological response mediated by serotype-cross-reactive memory T cells from the primary infection. Furthermore, epidemiological studies have shown that the sequence of infection with different DENV serotypes affects disease severity. Though much has been learned from human studies, there exist uncontrollable variables that are intrinsic in this system such as genetic factors and unknown infection histories. These factors can skew experimental results, making interpretations difficult. Therefore, a murine model to study the immunologic aspects of sequential dengue infections would be an asset to the field of dengue research. To examine the effect of sequential infection with different DENV serotypes on the CD8+ T cell response, we immunized Balb/c mice with a primary DENV infection on day 0 and subsequently challenged with a heterologous secondary DENV infection on day 28. We tested all possible sequences of infection with the four serotypes. We analyzed the T cell response to two previously defined epitopes on the DENV E (Ld-restricted) and NS3 (Kd-restricted) proteins. Using ELISPOT and intracellular cytokine staining, we measured the frequency of T cells secreting IFNγ and TNFα in response to stimulation with these epitopes during three phases: acute primary, acute secondary, and the memory phase after primary infection. We found that the T cell response in heterologous secondary infections was higher in magnitude than the response in acute primary infection or during the memory phase. We also found that the hierarchy of epitope specific responses, as measured by IFNγ secretion, was influenced by the sequence of infections. The adoptive transfer of immune serum or immune splenocytes suggested that memory T cells from the primary infection responded to antigens from the secondary infection. In vitroexperiments with T cell lines generated from mice with primary and secondary DENV infections suggested the preferential expansion of crossreactive memory T cells. In testing all of the different possible sequences of infection, we observed that two different sequences of infection (e.g., DENV-2 followed by DENV-1 versus DENV-2 followed by DENV-3) resulted in differential CD8+ T cell responses to the NS3 peptide even though both secondary infection serotypes contain the identical peptide sequence. To investigate this phenomenon, we examined the role of CD4+ T cell help on the memory CD8+ T cell response. We found that CD4+ T cell cytokine responses differ depending on the sequence of infection. In addition, it was also shown that crossreactivities of the CD4+ T cell response are also sequence-dependent. Moreover, denguespecific memory CD4+ T cells can augment the secondary CD8+ T cell response. Taken together, we demonstrated that this serotype sequence-dependent phenomenon is the result of differential help provided by cross-reactive memory CD4+T cells. The findings in this novel mouse model support the hypothesis that both CD4+ and CD8+ serotype-cross-reactive memory T cells from a primary dengue virus infection alter the immune response during a heterologous secondary dengue virus infection. These data further elucidate potential mechanisms whereby the specific sequence of infection with different dengue virus serotypes influences disease outcomes in humans.