Activation of the basic helix-loop-helix (bHLH) gene TAL1 is the most common genetic event seen in both childhood and adult T cell acute lymphoblastic leukemia (T-ALL). Despite recent success in treating T-ALL patients, TAL1 patients do not respond well to current therapies. In hopes of leading the way to better therapies for these patients, we have sought to determine the mechanism(s) of Tal1 induced leukemia in mice. By generating a DNA-binding mutant Tal1 transgenic mouse we have determined that the DNA binding activity of Tal1 is not required to induce leukemia. We have also shown that Tal1 expression in the thymus affects thymocyte development and survival. We demonstrate that Tal1 heterodimerizes with the class I bHLH proteins E47 and HEB in our mouse models of TAL1 induced leukemia. Severe thymocyte differentiation arrest and disease acceleration in Tal1/E2A+/- and Tal1/HEB+/- mice provides genetic evidence that Tal1 causes leukemia by inhibiting the function of the transcriptional activators E47 and HEB which have been previously shown to be important in T cell development. In pre-leukemic Tal1 thymocytes, we find the co-repressor mSin3A/HDAC1 bound to the CD4 enhancer, whereas an E47/HEB/p300 complex is detected in wild type thymocytes. Furthermore, mouse Tal1 tumors are sensitive to pharmacologic inhibition of HDAC and undergo apoptosis. These data demonstrate that Tal1 induces T cell leukemia by repressing the transcriptional activity of E47/HEB and suggests that HDAC inhibitors may prove efficacious in T-ALL patients that express TAL1.

High frequencies of CTL recognizing laboratory strains of HIV-1 are present in HIV-1 infected adults as early as preseroconversion. The presence of HIV-1 specific CTL during primary infection has been correlated with better control of early viremia and a more delayed onset of CD4 lymphocyte loss. Previous experiments in our laboratory have demonstrated that, unlike HIV-1 infected adults, the majority of vertically infected infants lack CTL which recognize laboratory strains of HIV-1 within the first year of life. ADCC antibody responses against laboratory strains of HIV-1 env gene products are also delayed until at least two years of age. As a possible correlate, disease progression is also more rapid in vertically infected infants. We hypothesized that HIV-1-specific CTL are type-specific in early infancy and that the use of target cells expressing laboratory strain gene products might limit the detection of HIV-1-specific CTL. To address this hypothesis, HIV-1 env genes from early isolates of four infants were PCR amplified, cloned, and used to generate recombinant vaccinia vectors (vv). The frequencies of CTL precursors (CTLp) recognizing env gene products from autologous isolates and the IIIB strain of HIV-1 were measured at time points from early infancy to 19 months using limiting dilution analysis (LDA). ADCC titers were also measured against autologous and IIIB env gene products at 4 time points spanning 2 months to 2 years of age. CTL precursors from 3 of 4 of these patients were specific only for autologous HIV-1 env gene products during the first 6 to 12 months of age. A pattern of CTL responsiveness was observed in these 3 patients in which type-specific CTL precursors observed in early infancy were replaced by cross-reactive, group-specific CTL by 6 to 12 months of age. CTL precursors from a fourth patient at 12 months of age recognized IIIB env and 1 out of 2 envs derived from 2 autologous viral isolates. High titers titers of ADCC antibodies against autologous env were detected in two infants prior to the detection of ADCC antibodies to IIIB. In two other infants, group specific ADCC antibody responses were detected in late infancy. Our results demonstrate that young infants can mount HIV-1 specific CTL and ADCC responses. The ability of young infants to mount cellular immune responses to HIV-1 also provides support for the concept of perinatal vaccination to prevent HIV-1 transmission. Furthermore. the lack of broadly-reactive CTL in early infancy suggests that the use of vaccines based on laboratory strains of HIV-1 may not afford protection from vertical infection.

Cells have evolved different pathways in order to tolerate damage produced by different cytotoxic agents. Each agent reacts differently with DNA causing formation of different types of adducts, each eliciting the SOS stress response to induce different cellular repair pathways. One such type of substrate generated by cytotoxic agents is the DNA double strand break (DSB). The main pathway to repair such damage in the cell is through a process of recombination. In this thesis, I specifically examined the anti-cancer therapeutic agent cisplatin, which forms single- and double-strand breaks in DNA, and methylating agents, which are proposed to also be capable of forming such breaks. Neither type of agent can directly form these breaks; however, they leave a signature type of damage lesion which is recognized by different repair processes. The mismatch repair (MMR) status of a mammalian cell or an Escherichia coli dam mutant relates directly to the sensitivity of the cells to the agents mentioned above. As the dam gene product plays an important role in this pathway and in other processes in the cell, when mutated, dam cells are more sensitive to methylating agents and cisplatin than wildtype. A combination of dam and either mutS or mutL restores resistance to the same agents to wild type levels. Therefore, mismatch repair sensitizes dam bacteria to these agents. The rationale for this comes from examining the viability of dam mutants, as dammutants are only viable because they are highly recombinogenic. The presence of MMR-induced nicks or gaps results in the formation of DSBs that require recombination to restore genomic integrity. Mismatch repair proteins inhibit recombination between homeologous DNA. Homeologous recombination (recombination between non-identical, but similar, DNA sequences) is only possible when the MMR proteins, MutS and MutL, are absent. It is postulated that this is because MutS recognizes the homeologous DNA and subsequently slows down or aborts recombination completely. The double mutant, dam mutS/L shows wild type levels of sensitivity to cisplatin because mismatch repair is no longer recognizing the adducts and recombinational repair is allowed to continue. Human cells behave in an analogous fashion to the bacterial dam mutant, showing sensitivity to cisplatin and methylating agents. When an additional mutation in a mismatch repair gene is present, the cells become as resistant as wild type. Therefore, the E. coli dammutant is a useful model system to study this mechanism of drug resistance. DNA containing cisplatin adducts or lesions resulting from methylation are substrates for other types of repair processes such as nucleotide excision repair and base excision repair; however they have also been implicated as substrates for MMR and recombinational repair. The goal of the work in this thesis was two-fold. The first was to identify the gene products and mechanism necessary for repair of cisplatin damage by recombination. The second was to examine the mechanism of cisplatin toxicity, and specifically how MMR proficiency aids in the cytotoxicity of this drug by preventing recombination. Using the duplicated inactive lac operon recombination assay, we were able to determine the requirements for spontaneous and cisplatin-induced recombination, the RecBCD and RecFOR pathways. We were also able to further postulate that the cisplatin- induced signature damage recognized by recombination was the double strand break, likely formed from fork stalling and regression or a subsequent collapse during DNA synthesis, thus requiring these pathways for repair. This observation led to the experiments involving examination of the mechanism of cisplatin toxicity and where MMR could inhibit specific steps of recombination with DNA containing cisplatin lesions. Low levels of cisplatin lesions slowed the rate of RecA-mediated strand transfer in vitro, likely due to its ability to form a large bend in the DNA. MutS bound to cisplatin lesions in the DNA during heteroduplex formation in the RecA strand exchange step of recombination, inhibiting branch migration, and aborting the reaction. In order for MutS to inhibit recombination with cisplatin lesions, the results in the work in Chapter IV, show that binding to the lesion requires the C-terminus of MutS to be present, possibly due to a requirement for tetramerization of the protein, a domain contained in the C-terminus of MutS. This antirecombination function is different than the mutation avoidance function of MutS, as binding of mismatches requires only dimers. This differential sensitivity for cisplatin versus a mismatch was further exemplified in Chapter V, the experiments with dna mutants, where the greatest difference in sensitivity was observed for a dnaE mutant (catalytic subunit of polIII), which was as sensitive to cisplatin as a dam mutant, but fairly resistant to treatment with MNNG. This is indicative of the potency of a cisplatin adduct to block polymerase progression, versus a mismatch which poses little problem to synthesis. Recombination is invoked to repair DSBs caused by the cisplatin lesions through the RecBCD and FOR pathways after fork regression or collapse. A main conclusion from these studies is that a cisplatin lesion is processed differently than a mismatch. The mechanism of how a cisplatin lesion is processed, forming the DSB which invokes recombinational repair is still unclear and continues to be investigated.

The NOD mouse is a widely studied model of type 1 diabetes. The loss of self-tolerance leading to autoimmune diabetes in NOD mice involves at least 27 genetic loci. Curing type I diabetes in mice and humans by islet transplantation requires overcoming both allorejection and recurrent autoimmunity. This has been achieved with systemic immunosuppression, but tolerance induction would be preferable. In addition to their genetic defects in self-tolerance, NOD mice resist peripheral transplantation tolerance induced by costimulation blockade using donor-specific transfusion and anti-CDl54 antibody. Failure has been attributed to the underlying autoimmunity, assuming that autoimmunity and resistance to transplantation tolerance have a common basis. Hypothesizing that these two abnormalities might be related, we investigated whether they had a common genetic basis. Diabetes-resistant NOD and C57BL/6 stocks congenic for various reciprocally introduced Idd loci were assessed for their ability to be tolerized. Surprisingly, in NOD congenic mice that are almost completely protected from diabetes, costimulation blockade failed to prolong skin allograft survival. In reciprocal C57BL/6 congenic mice with NOD-derived Idd loci, skin allograft survival was readily prolonged by costimulation blockade. Unexpectedly, we observed that (NOD x C57BL/6)F1 mice, which have no diabetes, nonetheless resist induction of tolerance to skin allografts. Further analyses revealed that the F1 mice shared the dendritic cell maturation defects and abnormal CD4+ T cell responses of the NOD but had lost its defects in macrophage maturation and NK cell activity. Finally, using a genome wide scan approach, we have identified four suggestive markers in the mouse genome that control the survival of skin allografts following DST and anti-CD154 mAb therapy. We suggest that mechanisms controlling autoimmunity and transplantation tolerance in NOD mice are not completely overlapping and are potentially distinct, or that the genetic threshold for normalizing the transplantation tolerance defect is higher than that for preventing autoimmune diabetes. We conclude that resistance to allograft tolerance induction in the NOD mouse is not a direct consequence of overt autoimmunity and that autoimmunity and resistance to costimulation blockade-induced transplantation tolerance phenotypes in NOD mice are not under identical genetic control.

Mature αβ and γδ T cells arise from a common precursor population in the thymus. Much debate has focused on the mechanism of T cell lineage choice made by these multi-potential precursor cells. It is widely believed that the decision of these precursor cells to commit to the γδ or αβ T cell lineages is regulated primarily by a specific instructive signal relayed through the appropriate T cell receptor. Contrary to this model, we present evidence for a TCR-independent lineage commitment process. Comparison of global gene expression profiles from immature αβ and γδ lineage thymocytes identified Sox13, an HMG-box transcription factor, as a γδ T cell-specific gene. Unlike other HMG-box transcription factors such as TCF1, LEF1 and SOX4, that are critical for proper αβ T cell development, Sox13 expression is restricted to early precursor subsets and γδ lineage cells. Importantly, SOX13 appears to influence the developmental fate of T cell precursors prior to T cell receptor expression on the cell surface. Transgenic over-expression of Sox13 in early T cell precursors strongly inhibits αβ lineage development, in part, by inhibiting precursor cell proliferation and concomitantly, leading to increased cell death among αβ lineage subsets. Steady-state γδ T cell numbers, however, appear unaffected. Strikingly, the DP αβ lineage cells that do develop in Sox13 transgenic mice are imprinted with a γδ- or precursor-like molecular profile, suggesting that SOX13 plays an active role in the lineage fate decision process or maintenance. Sox13-deficient mice, on the other hand, have selectively reduced numbers of γδ thymocytes, indicating that SOX13 is essential for proper development of γδ T cells. We present additional data demonstrating that SOX13 is a canonical WNT signaling antagonist modulating TCF1 activity, raising a strong possibility that WNT signals, and their modulators, are at the nexus of γδ versus αβ T cell lineage commitment.

DNA damage response pathways are a complicated network of proteins that function to remove and/or reverse DNA damage. Following genetic insult, a signal cascade is generated, which alerts the cell to the presence of damaged DNA. Once recognized, the damage is either removed or the damaged region is excised, and the original genetic sequence is restored. However, when these pathways are defective the cell is unable to effectively mediate the DNA damage response and the damage persists unrepaired. Thus, the proteins that maintain the DNA damage response pathway are critical in preserving genomic stability. One essential DNA repair protein is the Breast Cancer Associated gene, BRCA1. BRCA1 is essential for mediating the DNA damage response, facilitating DNA damage repair, and activating key cell cycle checkpoints. Moreover, mutations in BRCA1 lead to a higher incidence of breast and ovarian cancer, highlighting the importance of BRCA1 as a tumor suppressor. In an effort to better understand how BRCA1 carried out these functions, researchers sought to identify additional BRCA1 interacting proteins. This led to the identification of several proteins including the BRCA1 Associated C-terminal Helicase, BACH1. Due to the direct interaction of BACH1 with a region of BRCA1 essential for DNA repair and tumor suppression, it was speculated that BACH1 may help support these BRCA1 function(s). In fact, initial genetic screenings confirmed that mutations in BACH1 correlated not only with hereditary breast cancer, but also with defects in DNA damage repair processes. The initial correlation between BACH1 and cancer predisposition was further confirmed when mutations in BACH1 were identified in the cancer syndrome Fanconi anemia (FA) (complementation group FA-J), thus giving BACH1 its new name FANCJ. These findings supported a previously established link between the FA and BRCA pathways and between FA and DNA repair. In particular, we demonstrated that similar to other FA/BRCA proteins, suppression of FANCJ lead to a substantial decrease in homologous recombination and enhanced both the cellular sensitivity to DNA interstrand cross-linking agents and chromosomal instability. What remained unknown was specifically how FANCJ functioned and whether these functions were dependent on its interaction with BRCA1 or other associated partners. In fact, we identified that FANCJ interacted directly with the MMR protein MLH1. Moreover, we found that the FANCJ/BRCA1 interaction was not required to correct the cellular defects in FA-J cells, but rather that the FANCJ/MLH1 interaction was required. Although both the FA/BRCA and MMR pathways undoubtedly mediate the DNA damage response, there was no evidence to suggest that these pathways were linked, until recently. Our findings not only indicate a physical link between these pathways by protein-protein interaction, but also demonstrated a functional link.

The studies presented in this thesis addressed mainly two aspects of Piwi-interacting RNA (piRNA) biology in the Drosophilagermline. We investigated the role of the piRNA pathway in embryonic axis specification. piRNAs mediate silencing of retrotransposons and the Stellate locus. Mutations in the Drosophila piRNA pathway genes armitage and aubergine disrupt embryonic axis specification, triggering defects in microtubule polarization and asymmetric localization of mRNA and protein determinants in the developing oocyte. Mutations in the ATR/Chk2 DNA damage signal transduction pathway dramatically suppress these axis specification defects, but do not restore retrotransposon or Stellatesilencing. Furthermore, piRNA pathway mutations lead to germline-specific accumulation of γ-H2Av foci characteristic of DNA damage. We conclude that piRNA based gene silencing is not required for axis specification, and that the critical developmental function for this pathway is to suppress DNA damage signaling in the germline. We have also identified a new member of the piRNA pathway. We show that mutations in rhino, which encodes a rapidly evolving Heterochromatin Protein 1 (HP1) chromo box protein, lead to germline specific DNA break accumulation, trigger Chk2 kinase dependent defects in axis specification, and disrupt germline localization of Piwi proteins. Mutations in rhino and the piRNA pathway gene armitage disrupt silencing of all major transposon families, but do not alter expression of euchromatic or heterochromatic protein coding genes. Deep sequencing studies show that rhino mutations significantly reduce or eliminate anti-sense piRNAs derived from the majority of transposable elements in the Drosophila genome, and lead to a dramatic reduction in piRNAs derived from major piRNA production clusters on chromosomes 2R and 4. Rhino protein localizes to distinct nuclear foci, and associates with the chromosome 2R and 4 clusters by chromatin immunoprecipitation. The Rhino HP1 homologue is therefore required for piRNA biogenesis, transposon silencing, and maintenance of germline genome integrity.

Tolerance-based stem cell transplantation using sub-lethal conditioning is being considered for the treatment of human disease, but safety and efficacy remain to be established. In order to study these two issues, we first established that mouse bone marrow recipients treated with sub-lethal irradiation plus transient blockade of the CD40-CD154 costimulatory pathway develop permanent hematopoietic chimerism across allogeneic barriers. Our conditioning regimen of 6 Gy irradiation, a short course of anti-CD154 mAb and 25 million fully allogeneic BALB/c bone marrow cells consistently produced long-term, stable, and multilineage chimerism in C57BL/6 recipients. Furthermore, chimeric mice displayed donor-specific transplantation tolerance, as BALB/c skin allografts were permanently accepted while third-party CBA/JCr skin allografts were promptly rejected. We next determined both the safety and efficacy of this protocol by infecting chimeric mice with lymphocytic choriomeningitis virus (LCMV) either at the time of transplantation or at several time points afterwards. Infection with LCMV at the time of transplantation prevented engraftment of allogeneic, but not syngeneic, bone marrow in similarly treated mice. Surprisingly, infected allograft recipients also failed to clear the virus and died. Post-mortem study revealed hypoplastic bone marrow and spleens. Hypoplasia and death in these mice required the combination of 6 Gy irradiation, LCMV infection on the day of transplantation, and an allogeneic bone marrow transplant but did not require the presence of anti-CDl54 mAb. Allochimeric mice infected with LCMV 15 days after transplantation were able to survive and maintain their bone marrow graft, indicating that the deleterious effects of LCMV infection on host and graft survival are confined to a narrow window of time during the tolerization and transplantation process. The final section of this thesis studied the mechanisms of graft rejection and death in sublethally irradiated recipients of allogeneic bone marrow and infection with LCMV at the time of bone marrow transplantation. Infection of interferon-α/β receptor knockout mice at the time of transplantation prevented the engraftment of allogeneic bone marrow, but the mice survived. Therefore, IFN-αβ is involved in the development of marrow hypoplasia and death, whereas a second mechanism is involved in blocking the development of chimerism in these mice. Through the use of depleting mAb's and knockout mice we demonstrate that three types of recipients survived and became chimeric after being given sublethal irradiation, anti-CD154 mAb, an allogeneic bone marrow transplant and a day 0 LCMV infection: mice depleted of CD8+ T cells, CD8 knockout mice, and TCR-αβ knockout mice. Our data indicate that the mediator of bone marrow allograft destruction in LCMV-infected mice treated with costimulatory blockade is a radioresistant CD8+ NK1.1- TCRαβ+ T cell. We conclude that a non-cytopathic viral infection at the time of transplantation can prevent engraftment of allogeneic bone marrow and result in the death of sub-lethally irradiated mice treated with costimulation blockade. The abrogation of allogeneic bone marrow engraftment is mediated by a population of CD8+ NK1.1- TCRαβ+ T cells and the mediator of hypoplasia and death is viral induction of IFN-αβ.

p53, known as the “guardian of the genome”, is the most well-characterized tumor suppressor gene. The central role of p53 is to prevent genome instability. p53 is the central node in an incredibly elaborate genome defense network for receiving various input stress signals and controlling diverse cellular responses. The final output of this network is determined not only by the p53 protein itself, but also by other p53 cooperating proteins. p300 and CBP (CREB-Binding Protein) act as multifunctional regulators of p53 via acetylase and ubiquitin ligase activities. Prior work in vitro has shown that the N-terminal 595 aa of p300 encode both generic ubiquitin ligase (E3) and p53-directed E4 functions. Analysis of p300 or CBP-deficient cells revealed that both coactivators were required for endogenous p53 polyubiquitination and the normally rapid turnover of p53 in unstressed cells. Unexpectedly, p300/CBP ubiquitin ligase activities were absent in nuclear extracts and exclusively cytoplasmic. In the nucleus, CBP and p300 exhibited differential regulation of p53 gene target expression, C-terminal acetylation, and biologic response after DNA damage. p300 activated, and CBP repressed, PUMA expression, correlating with activating acetylation of p53 C-terminal lysines by p300, and a repressive acetylation of p53 lysine-320 induced by CBP. Consistent with their gene expression effects, CBP deficiency augmented, and p300 deficiency blocked, apoptosis after doxorubicin treatment. Subcellular compartmentalization of p300/CBP’s ubiquitination and transcription activities reconciles seemingly opposed functions—cytoplasmic p300/CBP E4 activities ubiquitinate and destabilize p53, while nuclear p300/CBP direct p53 acetylation, target gene activation, and biological outcome after genotoxic stress. p53 is a prominent tumor suppressor gene and it is mutated in more than 50% of human tumors. Reactivation of endogenous p53 is one therapeutic avenue to stop cancer cell growth. In this thesis, we have identified S5as a critical regulator of p53 degradation and activity. S5a is a non-ATPase subunit in the 19S regulatory particle of the 26S proteasome. Our preliminary data indicates that S5a is required for p53 instability and is a negative regulator of p53 tranactivation. As a negative regulator of p53, S5a may therefore also represent a new target for cancer drug development against tumors that specifically maintain wild type p53.

The endoplasmic reticulum is the primary organelle in the cell which has the responsibility of properly folding proteins belonging to the secretory pathway. Secretory proteins are essential for a variety of functions within the body like metabolism, growth and survival. Hence, proper folding of the proteins in the ER is absolutely essential to maintain cellular and body function. The environment of the ER is substantially different from that of the cytoplasm and is primed essentially to provide the optimum conditions to fold newly synthesized polypeptides following translation by the ribosomes in the cytoplasm and on the surface of the ER. In order for secretory proteins to fold properly, ER homeostasis must be maintained. ER homeostasis is defined by the dynamic balance between the ER protein load and the ER capacity to process this load. The optimum environment of the ER, or ER homeostasis, can be perturbed by pathological processes such as hypoxia, glucose deprivation, viral infections, environmental toxins, inflammatory cytokines, and mutant protein expression, as well as by physiological processes such as aging. Disruption of ER homeostasis causes accumulation of unfolded and misfolded proteins in the ER. This condition is referred to as ER stress. Cells cope with ER stress by activating the unfolded protein response (UPR). The UPR is initiated by three ER transmembrane proteins: Inositol requiring 1 (IRE1), PKR-like ER kinase, and activating transcription factor 6 (ATF6). These three master regulators sense and interpret protein folding conditions in the ER and translate this information across the ER membrane to activate downstream effectors, spliced XBP1, phosphorylated eIF2α and ATF4, and cleaved active ATF6 respectively. These effectors have two distinct outputs, homeostatic and apoptotic. Homeostatic outputs are adaptive responses that function to attenuate ER stress and restore ER homeostasis. These responses include the attenuation of protein translation to reduce ER workload and prevent further accumulation of unfolded proteins, upregulation of molecular chaperones and protein processing enzymes to enhance the ER folding activity, and the increase in ER-associated degradation (ERAD) components to promote clearance of unfolded proteins. When ER stress reaches a point where the cells cannot tolerate the load of unfolded proteins any more, apoptosis sets in. One of the major secretory proteins in mammals, vascular endothelial growth factor VEGF, is essential for either normal or pathological angiogenesis (blood vessel development). VEGFA is the primary member of this family which is expressed in all endothelial cells and is responsible for sprouting and invasion of blood vessels into the interstitium and thus helps in supplying nutrients and oxygen to growing cells. Recent studies have indicated that cells suffering from insufficient blood supply experience ER stress. The ER needs energy and oxygen for the folding process, thus nutrient deprivation (low ATP production) and hypoxia caused by insufficient blood supply leads to inefficient protein folding and ER stress in cells, especially in cancer cells that grow and spread rapidly. This condition also occurs in the development of the mammalian placenta. The placenta is an essential tissue characterized by a lot of blood vessels. It is responsible for the exchange of nutrients and growth factors between maternal and fetal blood vessels and hence is essential for survival of the embryo. Nutrient deprivation and hypoxia stimulate the production of VEGFA and other angiogenic factors, leading to protection against ischaemic injury in both cancer cells as well as the developing placenta. In this dissertation, we report that the three master regulators of the UPR, IRE1α, PERK and ATF6α, mediate transcriptional regulation of VEGFA under ER stress in cancer cells. Inactivation of any of the three master regulators leads to attenuation of VEGFA expression under ER stress. We show that IRE1α is able to regulate VEGFA through its downstream transcription factor XBP1 which activates the VEGFA promoter. IRE1α mediated VEGFA regulation is also essential for normal development of labyrinthine trophoblast cells in the placenta. ATF6α also regulates VEGFA via its promoter. PERK is able to activate VEGFA by preferential activation of its downstream effector, ATF4, which binds intron 1 of the VEGFA gene. Thus our work reveals a twopronged differential regulatory action of the UPR sensors on VEGFA gene expression. This work suggests that a fully active UPR is essential for VEGFA upregulation under ER stress. All three regulators are required in cancer cells for normal VEGFA expression. This tight regulation of VEGFA by the UPR presents a wonderful opportunity for therapeutic intervention into angiogenic growth of tumors.