Multiple myeloma is a plasma cell malignancy that localizes to the bone. It is diagnosed in 15,000 new patients per year, making it the second most common hematologic malignancy. The major source of morbidity in these patients is due to bone destruction induced by the myeloma cells leading to severe bone pain and pathologic fractures. Bone destruction in myeloma is mediated by an increase in osteoclast activity, the cells that normally resorb bone, with a concomitant decrease in osteoblast number and function, the cells that normally rebuild bone. The cause of the decrease in osteoblasts is not well understood. Interleukin-3 (IL-3) is upregulated in myeloma compared to normal controls and can mediate osteoclast activation in myeloma. This dissertation investigates the potential role of IL-3 as an osteoblast inhibitor in myeloma. First, IL-3 blocked osteoblast differentiation in a primary murine osteoblast culture system in a dose-dependant manner. Importantly, IL-3-mediated osteoblast inhibition occurred at IL-3 levels present in bone marrow samples from patients with myeloma. IL-3 did not inhibit osteoblast differentiation in cell lines, indicating that the IL-3 effects were not direct. Conversely, IL-3 caused proliferation in CD45+ hematopoietic cells present in the primary murine cultures, and depletion of CD45+ cells from these cultures resulted in a loss of IL-3 inhibition of osteoblast differentiation. Reconstitution of cultures with CD45+ cells resulted in restoration of the ability of IL-3 to inhibit osteoblasts. These CD45+ cells were shown to be CD11b+ and in the monocyte/macrophage lineage. Further studies were conducted into the mechanism of IL-3-mediated osteoblast inhibition. Cell-to-cell contact was required between osteoblasts and CD45+ hematopoietic cells, and separation of the cell population by transwell cultures abolished IL-3 inhibition of osteoblasts. Transcript levels of several integrins expressed on osteoblasts were not increased by treatment with IL-3, indicating that increased binding of CD45+ cells to osteoblasts is not the mechanism required for osteoblast inhibition. Contact between CD45+ cells and osteoblasts could result in increased expression of a juxtacrine factor that mediates IL-3 inhibition of osteoblasts. In myeloma, IL-3 can mediate proliferation of malignant cells, stimulation of osteoclast activity, and inhibition of osteoblast activity, which ultimately leads to exacerbation of lytic lesions in these patients. Thus, IL-3 is a potential target for myeloma therapy.

Elevated levels of the molecular adaptor protein p21waf1/cip1 (p21) and of the IL-3 receptor alpha chain are correlated with chemoresistance and poor prognosis in acute myeloid leukemia (AML). p21 is a core regulator of many biological functions including cell cycle control, apoptosis and differentiation. Our laboratory has demonstrated a decrease in p21 expression levels during cytokine-induced granulocytic differentiation, leading us to hypothesize that p21 antagonizes granulopoiesis. The proliferative cytokine IL-3 has been shown to prevent granulocytic differentiation of murine and human myeloid progenitor cells. We also hypothesized that IL-3 inhibition of differentiation is mediated in part by p21, and tested this in murine 32Dcl3 myeloblasts that are used to model granulopoiesis. Our findings demonstrated that p21 antagonized differentiation by promoting apoptosis of cells exposed to the differentiation inducer G-CSF. We also showed that p21 prevented premature expression of primary granule proteins and contributed to maintenance of the myeloblast phenotype. Furthermore, p21 knockdown accelerated morphologic differentiation of 32Dcl3 cells stimulated to differentiate with G-CSF. We then determined how IL-3 maintains p21 expression in myeloblast cells. We showed that IL-3 stabilized p21 mRNA in myeloblasts leading to high levels of p21 protein. This effect mapped to the 3' untranslated region (UTR) of the p21 transcript. p21 transcript stabilization by IL-3 was independent of PI3-kinase and ERK pathway signaling. In vitro binding assays provided evidence that distinct sets of RNA:protein interactions occur within the proximal 303 nucleotides of the p21 3' UTR and are regulated by IL-3 and G-CSF signaling. Association of a 60-65 kDa protein with p21 riboprobes correlated with IL-3 mediated p21 mRNA stabilization, whereas binding by a 40-42 kDa protein was associated with destabilization of p21 transcripts in 32Dcl3 cells undergoing G-CSF-induced differentiation. These findings provide the first evidence for IL-3-mediated stabilization of mRNA transcripts in myeloid progenitor cells. The finding that p21 antagonized granulopoiesis is also novel. Because high levels of the IL-3 receptor and high p21 expression have separately been linked to poor outcomes in AML, IL-3 mediated p21 mRNA stabilization may contribute to differentiation blockade during AML pathogenesis.

Herpes Simplex Virus (HSV) is a pervasive human pathogen that can establish both symptomatic productive infections and asymptomatic latent infections. During infection, the HSV genome undergoes physical changes that are regulated by cellular and viral proteins. These changes lead to either a template for genome replication during the productive cycle or a persistent stable genome configuration during latency. Changes in viral genomes and events leading to these changes during a particular life cycle are not clearly understood. Using both Gardella gel analysis for circular HSV genomes and restriction enzyme analysis for end-to-end linkages of HSV genomes, we show that HSV genomes circularize only in the absence of the HSV immediate early gene ICP0. In the presence of ICP0 genome circularization is inhibited. Although HSV replication has been previously thought to initiate by a theta mechanism from a circular genome template, these results suggest that HSV replication initiates from a linear genome template due to the presence of ICP0 during lytic infection. We also show that circular genomes are the stable form during long-term persistent infections that model latency. Because HSV genomes begin as linear, double-stranded DNA during infection, host cells may treat the ends of incoming genomes as DNA double strand breaks (DSB) and subsequently repair these ends by circularization of genomes. However, it is unclear if/how these DSB repair pathways contribute to the manipulation of HSV genome configurations during infection and how viral proteins, in particular ICP0, may alter/counteract this repair response to form a template for replication during the productive cycle. Here we show that the cellular DSB repair mechanism, non-homologous end-joining (NHEJ), is the major mechanism by which HSV genomes are circularized. ICP0 not only inhibits HSV genome circularization but also affects the abundance of proteins involved in NHEJ and the distribution of other repair proteins during infection. The study presented here begins to uncover how the interplay between host cell repair responses and the virus' reply to these responses contributes to forming either a genome template for replication during the productive cycle or a persistent stable configuration during latency.

The Gaussian network model (GNM) can be used as a first approximation for describing the fluctuation dynamics of proteins, the limits of applicability and the range of validity of the model parameters need to be established. A systematic analysis of the GNM predictions is done within the scope of this thesis, and the potential utility of GNM for elucidating structure-dynamics-function relations in enzymes is explored. The application of the GNM to a set of 183 non-homologous proteins shows that it can predict the X-ray crystallographic temperature factors more precisely than full-atomic normal mode analysis (NMA) does. Furthermore, the application to 1250 non-redundant proteins indicates that the GNM predictions agree better with NMR solution data, than X-ray crystallographic, and measurements taken at high diffraction temperatures. A systematic study of 98 enzymes that belong to different enzyme classes (EC) shows that catalytic residues are distinguished by their restricted mobilities in the global modes. The amplitudes of their fluctuations rank in the lowest 7% range amongst the rank-ordered mobilities of all residues. Catalytic residues also bear more restricted mobilities than their 4 flanked neighbors in sequence and this feature holds for more than 70% of the examined catalytic residues, suggesting a communication between chemical activity and molecular mechanics. The observed restricted mobility of catalytic residues is used as a criterion for identifying active sites of enzymes in a newly developed algorithm (COMPACT). The method shows a high sensitivity and a moderate-to-low specificity for a set of representative monomeric enzymes. All the false-positives predicted by COMPACT are found to be highly conserved, suggesting that their finely tuned dynamics results from evolutionary pressure. These particular sites are proposed to serve as alternative drug binding targets. We have implemented this tool in iGNM, a database of protein dynamics. Protein dynamics stored in iGNM or computed from the online calculation server (oGNM) have assisted in identifying possible silver ion binding residue in creatinase and describing the loop mobilities of low-fidelity DNA polymerase. Over all, this dissertation supports the view that protein structures have been designed to undergo conformational changes that are required for their biological functions.

Human Papillomaviruses (HPVs) require terminal differentiation of the host cell to produce infectious virions. The process of viral maturation involves a variety of changes in the expression/activity of host proteins that lead to high-level replication of the viral genome and expression of the late viral genes. Although the late promoter regions of HPV-16 are still not fully characterized, differentiation-dependent regulation of viral genes is thought to involve changes in the binding of host cell transcription factors to the viral promoter and regulatory regions. Currently, little is known about specific cellular transcription factors involved in this process. We have used the Panomics TransSignal Protein/DNA array to identify changes in the levels of cellular transcription factors during methylcellulose-induced differentiation of W12 (20863) cells containing HPV-16. We then identified the differentially expressed transcription factors that specifically bind to HPV-16 promoters. We have validated the results obtained from the Panomics array by Western blot analysis and with chromatin immunoprecipitation. This approach identified approximately thirty transcription factors, many of which represent novel viral DNA-host protein interactions. At present, no treatments exist that effectively target and eliminate papillomaviruses (PVs) from infected cells or prevent its replication. We are employing a strategy to prevent virus replication in PV-infected cells through the conditional expression of the herpes simplex virus type 1 thymidine kinase (TK) gene. Expression of TK in this system is expected to be triggered by a homologous recombination event between the endogenous PV genome and a nonexpressing TK gene cassette, which is expected to change the nonexpressing cassette into a form that expresses TK. Various constructs were generated to express the TK in the above manner. Transfection of cell lines with a TK nonexpressing plasmid did not result in TK production. However, cotransfection of cell lines with PV plasmids along with the above TK construct containing PV sequences resulted in TK expression as shown by Northern and Western blot analyses. We also developed a TK expression cassette utilizing an adeno-associated virus (AAV) vector. Delivery of the cassette by AAV to PV-infected cells resulted in TK expression, and ganciclovir treatment resulted in efficient killing of these cells.

A pathogenicity locus of Pseudomonas syringae pv. syringae identified by Tn5 mutagenesis was investigated. The mutant strain PS9024 is attenuated for disease expression in its host, Phaseolus vulgaris, but produces the hypersensitive reaction (HR) in the nonhost, tobacco (Nicotina tabacum). A cosmid clone carrying 16 kilobases (kb) of contiguous genomic DNA partially complements this mutant. Altered growth of the mutant in planta was also partially restored. Marker exchange mutagenesis with Tn3- HoHo1 at two other sites within this locus results in mutants with attenuated and severely reduced pathogenicity. The locus is complex and contains repetitive DNA sequences. Northern analysis reveals that this locus is expressed in planta, but is not expressed in a rich growth medium, and the transcript is larger than 10 kb, suggesting that the locus is transcribed as a polycistronic mRNA. Comparison of total cellular protein profiles of R32 and PS9024 using SDS-PAGE analysis further reveals that at least nine protein bands ranging from approximately 100 kD or above in size are present in the wild type strain R32, but absent from the mutant. Additionally, a protein of approximately 45 kD is absent from the mutant. The site of Tn5 insertion has been partially sequenced. The initial search of the data banks suggested a gene or genes related to the ornithine biosynthetic pathway map to this locus. Further study strongly suggest a gene that encodes a membranceassociated protein and under the control of a promoter identical to appA gene promoter maps at this site and it is involved in the process of pathogenesis. / Graduation date: 1991

The sarcoglycanopathies are a subset of the limb girdle muscular dystrophies (LGMD) caused by mutations in the sarcoglycan genes (alpha, beta, gamma and delta). In collaborative studies, delta-sarcoglycan was delivered to deficient hamsters using a recombinant adeno-associated virus (AAV), which rescued muscle biochemically, histologically, and functionally. Murine knockouts for the other sarcoglycans permitted us to pursue AAV-mediated gene delivery. AAV-mediated gene delivery of beta-sarcoglycan to deficient mice provided long-term biochemical and histological rescue. AAV-mediated gene delivery of alpha-sarcoglycan to deficient mice showed initial rescue of biochemical and histological defects, although expression was not persistent. Severe Combined Immune-Deficient (SCID) mouse studies indicated that alpha-sarcoglycan over-expression leads to cytotoxicity. The apparent cytotoxicity can be interpreted with emerging models of sarcoglycan complex assembly. These studies show that AAV-mediated delivery of even closely related proteins can lead to different outcomes, and aspects of protein biochemistry can alter efficacy of gene delivery. Inherited muscle disorders typically have defined primary biochemical defects. However, there are likely secondary responses that mitigate gene delivery success. To dissect such variables, we studied the immunostimulatory properties of dystrophic muscle. We hypothesized that immune cell infiltrate accompanying degeneration/regeneration could be immunostimulatory, which could elicit an immune response to delivered transgenes, hampering the success of gene delivery. To study this, we tested antibody response to and persistence of, beta-galactosidase in normal and dystrophic muscle. Consistent with our hypothesis, dystrophic muscle showed increased immune surveillance and recognition of beta-galactosidase, evidenced by antibody titers and clearance of transduced cells. Furthermore, biochemical rescue of the dystrophy quenched the immune response. This indicated that dystrophic muscle is more prone to immune responses and that aspects of tissue pathology influence the persistence and efficacy of gene delivery. Our results suggest that full biochemical rescue will attenuate immunostimulatory effects. We also address a hurdle facing AAV-mediated gene therapy; namely, delivery methods. We developed an injection manifold, which was used to safely, accurately, and consistently deliver genes to 20 mm2 regions of muscle. Taken together, these results more clearly define barriers to gene delivery. Future research will finely tune regulation of transgenes and enable full rescue of biochemical defects.

The nef gene of HIV encodes a small myristylated protein that is required for viral replication and pathogenesis. Nef lacks intrinsic catalytic activity and has been postulated to function through binding to cellular proteins and altering signaling pathways. Nef binds to the macrophage-specific Src family member Hck through its SH3 domain with the highest affinity reported for an SH3-mediated protein-protein interaction. Studies presented in this dissertation characterized the interface constituting this unusually strong interaction by comparing the ability of different Nef alleles to bind and activate Hck in cells. By aid of molecular modeling and mutational analysis, we found the residues in the Nef hydrophobic pocket critically contributing to high affinity interaction with Hck. This is the first study to show that the HIV-1 Nef hydrophobic pocket is critical for SH3-mediated Hck activation in vivo and identified the pocket as a rational target for drug design to selectively disrupt Nef-Hck signaling in HIV-infected cells. We also investigated the role of HIV Nef in survival signaling in myeloid cells, which is a more relevant cell type for HIV infection. Recently, Nef has been demonstrated to be an important factor in promoting the survival of HIV-infected T cells. Since macrophages serve as HIV viral reservoirs, suppression of apoptosis of infected macrophages could have more impact on persistent virus infection in the host. In this study, we demonstrate that Nef protects the macrophage precursor cell line, TF-1 from apoptosis by upregulating the anti-apoptotic gene, Bcl-XL. In addition, the survival signal and Bcl-XL induction by Nef require Erk MAPK activation. This study provides the first evidence that Nef generates anti-apoptosis signals in cells of the myelomonocytic lineage and adds important evidence to the hypothesis that Nef may contribute to the establishment and maintenance of an HIV reservoir by conferring a survival advantage on HIV-infected macrophages

Heterogenous nuclear ribonucleoproteins (hnRNPs) are predominantly nuclear and have been shown to be involved in multiple cellular processes. There are more than 20 members of hnRNP protein family and they all bind RNA. Some hnRNPs like the ones in hnRNP H family was shown to have sequence specific RNA binding activity. hnRNP H potein family includes hnRNP H1, H2. H3 and F. This study mapped the location of hnRNP H/F binding site on the SV40 Late (SVL) pre-mRNA relative to the polyA signal. The binding site is characterized down to five consecutive Guanines. Disruption of hnRNP H/F binding site on the SVL pre-mRNA had a negative effect on polyadenylation in vitro and this effect is translated to downregulated gene expression in Gfp based in vivo assays. The fact that this effect was seen with constructs having single polyA sites implicates a pronounced effect on alternative polyadenylation. The second part of this study involves microarray experiments with A20 B-cell lymphoma and AxJ Plasmacytoma cells. More than 400 genes were found to be differentially expressed in AxJ cells. While this group includes many known B-cell stage specific genes, there are also numerous novel differentially expressed transcripts. The list includes transcription, splicing and polyadenylation factors, too. Elongation factor RNA polymerase II 2 (ELL2) is induced in AxJ cells as assayed by microarray and RT-PCR experiments. The overexpression of hnRNP F was shown revert the B-cell clock, in so as far as secretory to membrane Immunoglobulin ratio is concerned. In an attempt to identify the genes underlying this mechanism, microarray experiments were preformed with hnRNP F overexpressing AxJ plasmacytoma cells and compared the gene expression profile with the one of mock transfected cells. There are about 30 differentially expressed transcripts upon forced expression of hnRNP F. Three B-cell stage specific genes were differentially expressed in hnRNP F overexpressing AxJ cells. ELL2 is among these genes. ELL2 is a known transcriptional elongation enhancer. A novel model regarding the secretory to membrane Immunoglobulin switch and involving ELL2 is proposed.

The predominant mechanism of herpes simplex virus type 1 (HSV-1) entry into permissive cells involves initial virus attachment to the cells by the interaction of envelope glycoproteins gC and gB with cell surface glycosaminoglycans (GAGs), binding of envelope glycoprotein D to one of several dissimilar co-receptors, and fusion of the virus envelope with the cell membrane requiring the combined essential functions of glycoproteins gD, gB and gH/gL. The binding of gD to its cognate receptor appears to result in emission of an activating signal to the fusion apparatus which minimally consists of the other essential glycoproteins. To gain a better understanding of gDs involvement in the fusion-activating process, we took the approach of separating gD from the virus envelope to determine whether a soluble form of gD (sgD) could mediate entry of gD-deficient virus. The results showed that sgD enabled entry of gD-deficient HSV-1 into CHO-K1 cells bearing the gD receptors HVEM or nectin-1. Using mutant forms of sgD that selectively bind to one or the other receptor, we demonstrated that entry by this mechanism is receptor specific. Investigation of the mechanism of sgD-mediated entry demonstrated that the presence of virus at the cell surface was required at the time of sgD-receptor binding, which could be explained in part by our observation that sgD rapidly dissociated from the receptor under our experimental conditions. In addition, entry was not eliminated instantaneously when receptor-bound sgD was exposed to 37ºC, suggesting that the active conformation of receptor-bound sgD is not highly unstable. sgD was not stabilized at the cell surface or internalized in the presence of gD-deficient virus. Using lysosomotropic agents as well as protease protection assays, we obtained no reproducible evidence that sgD-mediated entry takes place by endocytosis. Surprisingly, virus attachment to cell-surface GAGs was not required for sgD-mediated entry. Furthermore, gD-deficient virus attached to GAG-deficient cells in the absence of sgD, revealing a previously unknown binding interaction between the HSV virion and the cell. This interaction was shown to be of a less stable nature than the virus-GAG interaction, and may play a role in normal virus entry. Our results provide new tools and directions to unravel the still incompletely understood events set in motion by gD binding to its receptor.