This dissertation addresses the role for Janus Kinase 3 (Jak3) in CD4+ T cell homeostasis and function. Jak3 is a protein tyrosine kinase whose activity is essential for signals mediated by the γc dependent cytokines IL-2, -4, -7, -9, -15, and -21. Previous data have demonstrated that peripheral CD4+ T cells from Jak3-deficient mice have a memory phenotype and are functionally impaired in both proliferative and IL-2 responses in vitro. Interestingly, Jak3/γc activity has been previously shown to play a role in the prevention of T cell anergy. These studies were initiated to more precisely define the role for Jak3/γc cytokines in the prevention of T cell anergy and the maintenance of functional CD4+ T cell responses. We began to address this question by assessing global gene expression changes between wild type and Jak3-/- CD4+ T cells. These data indicate that Jak3-/- CD4+ T cells have an increase in gene expression levels of inhibitory surface receptors as well as immunosuppressive cytokines. Further analyses confirmed that Jak3-deficient T cells express high levels of PD-1, secrete a Trl-type cytokine profile following direct ex vivo activation, and suppress the proliferation of wild type T cells in vitro. These characteristics indicate that CD4+ Jak3-/- T cells share properties with regulatory T cell subsets that have an important role in peripheral tolerance and the prevention of autoimmunity. We next addressed whether these regulatory characteristics were T cell intrinsic or rather the result of expanding in a Jak3-deficient microenvironment characterized by a number of immune abnormalities and a disrupted splenic architecture. Jak3-/- CD4+ T cells proliferate in vivoin a lymphopenic environment and selectively acquire regulatory T cell characteristics in the absence of any additional activation signals. While the precise mechanism by which Jak3-deficient T cells acquire these characteristics remains unclear, our data indicate that one important component is a T cell-intrinsic requirement for Jak3 signaling. These findings indicate several interesting aspects of T cell biology. First, these studies, demonstrate that the homeostatic proliferation of CD4+ T cells is not dependent on signaling via γc-dependent cytokine receptors. And, second, that the weak activation signals normally associated with homeostatic expansion are sufficient to drive Jak3-/- T cells into a non-conventional differentiation program. Previous data indicate that, for wild type T cells, signaling through both the TCR as well as γc-dependent cytokine receptors promote the homeostatic proliferation of T cells in lymphopenic hosts. Since Jak3-/- T cells are unable to receive these cytokine signals, their proliferation is likely to be wholly dependent on TCR signaling. As a consequence of this TCR signaling, Jak3-/- T cells proliferate, but in addition, are induced to up regulate PD-1 and to selectively activate the IL-10 locus while shutting off the production of IL-2. Since this fate does not occur for wild type T cells in a comparable environment, it is likely that the unique differentiation pathway taken by Jak3-/- T cells reflects the effects of TCR signaling in the absence of γc-dependent cytokine signaling. Interestingly, wild type T cells undergoing homeostatic expansion in lymphopenic hosts show many common patterns of gene expression to freshly-purified unmanipulated Jak3-/- T cells. For instance, micro array analysis of gene expression in wild type CD4+ T cells after lymphopenia induced homeostatic expansion show a similar pattern of upregulation in surface markers (PD-1 and LAG-3), and cytokine signaling molecules (IL-10 and IFN-γ cytokine, receptors, and inducible gene targets) to that of Jak3-/- CD4+ T cells immediately ex vivo. These data suggest that the process of homeostatic proliferation normally induces immune attenuation and peripheral tolerance mechanisms, but that full differentiation into a regulatory T cell phenotype is prevented by γc-dependent cytokine signals. Taken together these data suggest that Jak3 plays an important role in tempering typical immune attenuation mechanisms employed to maintain T cell homeostasis and peripheral tolerance.

Nuclear hormone receptors (NHR) constitute a superfamily of ligand inducible transcriptional activators that enable an organism to regulate development and homeostasis through switching on or off target genes in response to stimuli reflecting changes in environment as well as endocrine. NHRs include classical steroid hormone receptors (GR, AR, ER and MR) and retinoid, thyroid hormone receptors. One long-term goal of our lab is to understand the molecular mechanisms through which the transcriptional activity of NHRs is regulated. Extensive studies in the past few years have revealed that in addition to the dependence on ligand availability, the transcriptional activity of NHRs is also regulated by two types of proteins: co activators and corepressors. In the absence of ligand, many NHRs, including TR and RAR can actively repress target gene transcription with the help of corepressors, proteins that physically interact with both NHRs and histone deacetylases (HDACs). Functional interactions between NHRs and corepressors therefore lead to tightly compact and transcriptionally non-permissive chromatin structures after the removal of obstructive acetyl groups from histone tails by HDACs. On the other hand, ligand binding stabilizes NHRs in a conformation that favors interaction with proteins other than corepressors; many of these proteins are able to potentiate the transcriptional activity of NHRs through various mechanisms, such as histone acetylation, chromatin remodeling and recruitment of basal transcription machinery and are collectively termed coactivators. Two highly related corepressors, SMRT (silencing mediator of retinoid and thyroid hormone receptors) and N-CoR (nuclear receptor corepressor), have been cloned. This research in corepressor SMRT started by a systematic study of its subcellular localization. We found that SMRT predominantly forms a specific nuclear punctuate structure that does not appear to overlap with any other well-known subnuclear domains/speckles. Although our searching for specific sequence signals that may determine the specific speckle localization of SMRT did not yield conclusive results, we discovered the colocalization of unliganded RAR and certain HDACs, including HDAC1, 3,4 and 5, in the SMRT nuclear speckles. Moreover, SMRT is likely to be the organizer of such speckles since it appears to be able to recruit other proteins into these speckles. The presence of HDAC1 in the SMRT speckles suggests a direct association between these two proteins, which has not been detected by previous biochemical analyses. Interestingly, HDAC1 point mutants that are completely defective in deacetylase activity failed to locate to SMRT nuclear speckles, while another partially active mutant maintained the colocalization. These discoveries may indicate SMRT nuclear speckles as novel nuclear domains involved in transcriptional repression. More physiologically relevant support for this hypothesis arises from study of HDAC4 and 5. HDAC4 and 5 are potent inhibitors of transcriptional activator MEF2C. Nuclear presence of HDAC4/5 can block the activation of MEF2C, which is required during muscle differentiation. Normally, HDAC4 is predominantly located in cytoplasm. However, we found that in the presence of SMRT overexpression, HDAC4 was found mostly in SMRT nuclear speckles. This accumulation enhanced HDAC4 mediated inhibition on MEF2C transcriptional activity in a transient transfection assay. SMRT overexpression also resulted in accumulation of HDAC5 in the SMRT nuclear speckles compared to the nuclear diffuse distribution in the absence of SMRT. Again, this accumulation of HDAC5 in nuclear speckles correlated with enhanced inhibition of MEF2C. Taken together, our study suggested that instead of being merely a corepressor for NHRs, SMRT might function as an organizer of a nuclear repression domain, which may be involved in a broad array of cellular processes. In contrast to the limited number of corepressors, numerous co activators have been identified; the SRC (or p160) family is relatively well studied. This family includes three highly related members, SRC-1, TIF2/GRIP1, RAC3/AIB1/ACTR/p/CIP. Similar domain structures are shared among these factors, with the most highly conserved region, the bHLH-PAS domain found within the N terminal ~400 amino acid residues. This study of RAC3 aims to identify the function of the highly conserved N terminal bHLH-PAS domain by isolating interacting proteins through yeast two-hybrid screening. One candidate gene isolated encodes the C terminal fragment of the human homologue of the yeast protein MMS19. Functional studies of this small fragment revealed that it specifically interacted with human estrogen receptors (ERs) and inhibited ligand induced transcriptional activity of ERs in the transient transfection assay. Then we cloned the full-length human MMS19 cDNA and characterized the hMMS19 as a weak coactivator for estrogen receptors in the transient transfection assay. Furthermore, when tested on separate AF-1 or AF-2 of ERs, hMMS19 specifically enhanced AF-1 but had no effect on AF-2. These results identified hMMS19 as a specific coactivator for ER AF-1.

This body of work details the exploitation of an incredibly powerful neural culture system, which enables the in vitrostudy of events involved in peripheral nervous system (PNS) development. Using a myelinating dorsal root ganglion (DRG) explant culture system, node of Ranvier formation and maintenance and the associated generation and maturation of myelin segments was examined. In addition, Schwann cell (SC) development, dynamics, and migration were extensively studied. First, in characterizing these cultures, the discrete axonal localization of specific ankyrin isoforms was revealed. Ankyrins are peripheral membrane proteins that immobilize classes of integral membrane proteins to the spectrin based-membrane skeleton. Ankyrins interact with proteins such as the voltage-dependent/gated sodium channel (vgsc) and members of the L1 family of cell adhesion molecules. These interactions are physiologically relevant to the formation of membrane specializations involved in axon guidance and the initiation and propagation of action potentials. We examined ankyrinB and ankyrinG expression in cultured DRG explants, which allowed visualization of individual axons. AnkyrinB and ankyrinG exhibited differential localizations to specific axonal populations. This was evident as early as one day in vitro and persisted over time. In mature pre-myelinated cultures, axons having an apparent diameter of less than 1 µm predominantly expressed ankyrinB, whereas axons having a diameter greater than or equal to 1 µm predominantly expressed ankyrinG (based on immunocytochemical reactivity). When myelination was induced, ankyrinGwas appropriately localized to sites of nodal development flanked by myelinating glial processes in the large caliber axons. These observations suggest that axons destined for myelination may express a distinct complement of peripheral, and perhaps integral, membrane proteins as compared to those observed in non-myelinated axons. These distinguishing features may play a role in the selection of axons for myelination. This work was followed with defining the role axo-glial interactions play in organizing domains along the axon being myelinated. Nodes of Ranvier are specialized, highly polarized axonal domains crucial to the propagation of saltatory action potentials. In the PNS, axon-glial cell contacts have been implicated in SC differentiation and the formation of nodes of Ranvier. SC microvilli establish axonal contact at mature nodes, and their components have been observed to localize early to sites of developing nodes. However, a role for these contacts in node formation remains controversial. Using the myelinating explant culture system, we observed that SCs reorganize and polarize microvillar components, such as the ezrin-binding phosphoprotein 50kDa (EBP50)/regulatory co-factor of the sodium-hydrogen exchanger isoform 3 (NHERF-1), actin, and the activated ezrin, radixin, and moesin (ERM) family of proteins, concomitant with myelination in response to inductive signals. These components were targeted to the SC distal tips where live cell imaging revealed novel, dynamic growth cone-like behavior. Further, localized activation of the Rho signaling pathway at SC tips gave rise to these microvillar component-enriched “caps” and influenced the efficiency of node formation. Extending these findings, a more profound examination of SC dynamics was undertaken. This was a particularly important experimental transition, as SC motility is crucial in the development and regeneration of the PNS. The seemingly equivalent bipolarity of mature SCs represents a conundrum in terms of directed motility. Fluorescence time-lapse microscopy of transfected SCs within the myelinating DRG explants revealed a novel cycling of these cells between static, bipolar and motile, unipolar morphologies via asymmetric process retraction and extension. Concentrations of PIP2 (phosphatidylinositol (4,5)-bisphosphate), activated ERMs, and EBP50 delineated the transitory asymmetry associated with the generation and neuron-like migration of the unipolar cell. EBP50 over-expression enhanced unipolar SC migration, suggesting a new role for this adaptor protein in cell motility. Further, the ERMs themselves were found to be essential to both motility and process dynamics with ERM disruption yielding a dysfunctional, multipolar SC phenotype. We propose this novel form of motility may be associated with the correct alignment and spacing of SCs along axons prior to elaboration of the myelin sheath. These compiled studies present significant advances in understanding and examining axo-glial interactions in the PNS. This work establishes the foundation for further, novel exploration of normal PNS development and the regeneration and repair mechanisms involved in PNS injury and disease states.

Glioblastoma multiforme (GBM) is a grade IV astrocytoma and is the most aggressive malignant primary brain tumor in adults. Without treatment, patients are expected to survive an average of three months. Conversely, current treatment regimens only extend survival to 12-14 months. Characteristically, GBM tumors are highly proliferative, invasive and stop responding to treatments relatively fast due to therapy resistance. Interestingly, GBM also exhibits high metabolic activity but manages to maintain a low level of reactive oxygen species (ROS). These ROS neutralization capabilities are sustained by system Xc–, a sodium-independent, electro neutral transporter that is found in the plasma membrane of GBM cells. System Xc– is composed of a regulatory heavy subunit (4F2hc) linked to a 12 transmembrane domain catalytic light chain subunit (xCT) that mediates the uptake of L-cystine into the cell, and L-glutamate out of the cell, at a 1:1 ratio. Imported cystine is quickly reduced to L- cysteine, the rate limiting substrate in glutathione (GSH) synthesis. Glutathione is a major antioxidant in the central nervous system that is responsible for maintaining intracellular redox homeostasis by neutralizing ROS by direct and indirect methods. The function of chemo and radiation therapy is to generate significant levels of ROS that tigger the cell to undergo apoptosis. High intracellular GSH levels in cancer cells are associated with drug resistance and detoxification of alkylating agents such as temozolomide (TMZ). Therefore, system Xc– represents a potential target to reduce glioma cell survival and reduce tumor progression. Sulfasalazine is an FDA approved drug in the treatment of arthritis and Crohne’s disease and has been shown to inhibit system Xc–. In vitro SASP studies demonstrated a strong antitumor potential in preclinical mouse models of malignant glioma. However, two clinical trials using sulfasalazine with standard chemo and radiation therapy to treat GBM patients were terminated due to off-target effects. Both results showed high toxicity and no change in the overall survival of patients. These studies demonstrate the need for a more effective inhibitor of system Xc–. To further elucidate the role of system Xc– in GBM survival, stable xCT knock-down and over-expressing U251 glioma cells were generated. These lines were characterized for survival, proliferation, apoptosis and resistance to oxidative and genotoxic insult. As expected xCT-knockdown cells exhibited lower GSH levels, increased intracellular ROS and markers for apoptosis after oxidative and genotoxic insult. The xCT-over-expressing cells displayed higher levels of GSH, increased resistance to hydrogen peroxide and various chemotherapy drugs including TMZ. An interesting unforeseen result of xCT over-expression in glioma cells was an increase in the metabolic activity as a result of increased mitochondria. Using xCT-modified glioma lines stably, we demonstrate for the first time that system XC– over-expression not only promotes survival under oxidative stress but may also decreases sensitivity to chemotherapy treatment and increase metabolic properties. Therefore, therapeutic manipulation of this transporter either alone or in combination with other treatments may improve clinical outcome in patients diagnosed with GBM.

The ability of incorporation non-canonical amino acids (ncAAs) using translation offers researchers the ability of extend the functionality of proteins and peptides for many applications including synthetic biology, biophysical and structural studies, and discovery of novel ligands. Here we describe the three projects where the addition of ncAAs to in vitro translation systems creates useful chemical biology techniques. In the first, a fluorinated histidine derivative is used to create a novel affinity tag that allows for the selective purification of peptides from a complex mixture of proteins. In the second, the high promiscuity of an editing-deficient valine-tRNA synthetase (ValRS T222P) is used to demonstrate ribosomal translation of 13 ncAAs including those with novel side chains, α,α disubstitutions, and cyclic β amino acids. Lastly, a couple of these amino acids are integrated into the powerful ligand discovery tool of mRNA display for the discovery of helical peptide ligands.

Ligands of the human monoamine transporters encompass a wide range of both illicit and therapeutic drugs that act upon neural circuitry related to reward, motivation, and the processing of salient stimuli. The present study utilizes two methods for analyzing transporter substrates and inhibitors in order to characterize activity and assess potency. The first measures transient changes in intracellular calcium as a surrogate for transporter activity by harnessing the electrical coupling of monoamine transporters and L-type calcium channels. This is used to analyze novel chimera of the strong hDAT inhibitors methylphenidate and 𝛼-PPP in order to assess the contribution of specific moieties to potency. The observed reduction in potency suggests that methylphenidate may bind to the transporter in a manner distinct from 𝛼-PPP, as chimera would otherwise be expected to show similar activity to parent compounds. These results highlight the importance of 𝛼-carbon substituents and the relatively small contribution of beta-carbon groups to inhibitor potency at hDAT, while the lack of activity at hSERT suggests potency is not strongly influenced by beta-carbon or N-alkyl substituents. In order to further characterize drug-transporter interaction, a method was developed to analyze the kinetics of binding and unbinding using both known and novel hNET ligands, including a series of N-alkyl derivatives of 4-methylamphetamine. The study emphasizes the importance of both association and dissociation kinetics to affinity and sets up a methodological framework with two ways for determining Kd, with notable advantages over current models. The results indicate that lengthening the N-alkyl chain of 4-methylamphetamine leads to a decrease in potency and a shift in activity from substrate to blocker, with the results of N-propyl 4-methylamphetamine in particular indicating the potential existence of multiple low-affinity binding sites, each with distinct on and off kinetics. The implications of these results help elucidate the mechanism of action of transporter ligands and set up a framework for future studies that can more specifically classify the interaction between transporters and inhibitors or releasing agents.

DNA damage can cause genome instability, which may lead to human cancer. The most common form of DNA damage is DNA base damage, which is efficiently repaired by DNA base excision repair (BER). Thus BER is the major DNA repair pathway that maintains the stability of the genome. On the other hand, BER mediates DNA demethylation that can occur on the promoter region of important tumor suppressor genes such as Breast Cancer 1 (BRCA1) gene that is also involved in prevention and development of cancer. In this study, employing cell-based and in vitro biochemical approaches along with bisulfite DNA sequencing, we initially discovered that an oxidized nucleotide, 5’,2-cyclo-2-deoxyadenosine in DNA duplex can either cause misinsertion by DNA polymerase β (pol β) during pol β-mediated BER or inhibit lesion bypass of pol β resulting in DNA strand breaks. We then explored how a T/G mismatch resulting from active DNA demethylation can affect genome integrity during BER and found that pol β can extend the mismatched T to cause mutation. We found that AP endonuclease 1 (APE1) can use its 3'-5' exonuclease to remove the mismatched T before pol β can extend the nucleotide preventing a C to T mutation. The results demonstrate that the 3'-5' exonuclease activity of APE1 can serve as a proofreader for pol β to prevent mutation. We further explored the effects of exposure of environmental toxicants, bromate and chromate on the DNA methylation pattern on the promoter region of BRCA1 gene with bisulfite DNA sequencing. We found that bromate and chromate induced demethylation of 5-methylcytosines (5mC) at the CpG sites as well as created additional methylation at several unmethylated CpG sites at BRCA1 gene in human embryonic kidney (HEK) 293 cells. We further demonstrated that the demethylation was mediated by pol β nucleotide misinsertion and an interaction between pol β and DNA methyltransferase 1 (DNMT1) suggesting a cross-talk between BER and DNA methyltransferases. We suggest that DNA base damage and BER govern the interactions among the environment, the genome and epigenome, modulating the stability of the genome and epigenome and disease development.

Mis-folded proteins and their associated aggregates are a contributing factor in some human diseases. In this study we used the protein lysozyme as a model to define aggregation structures under denaturing conditions. Sasahara et al. (2007), Frare et al. (2009, 2006), and Rubin et al. (2008) observed conditions where heat denatured lysozyme formed fibril structures that were observed to be 8-17 nanometers in diameter under the electron microscope. Even though the crystal structure of lysozyme is known, the denatured form of this protein is still unknown. Therefore, we used Rosetta++ protein folding and blind docking software to create in silico models of the protein at denaturing temperatures and subsequently docked them into aggregates. Here we compare those structures and select forms consistent with the fibril structure from the previous papers. The next step is to be able to use the predicted models of the fibrilar forms of denatured lysozyme to help us understand the exact conformation of fibril structures. This will let us confirm the docking interactions during the fibril aggregation process. The ultimate goal is to use the validated denatured structures to model interactions with heat shock proteins during the dis-aggregation process.

Enterohemorrhagic Escherichia coli O157:H7 (EHEC) is an attaching and effacing pathogen that upon attachment to host cells, induce characteristic attaching and effacing lesions and formation of F-actin rich pedestals beneath sites of bacterial attachment. EHEC harbors a Type III secretion system through which it delivers dozens of effectors into the host cell. The two secreted effectors critical for EHEC-mediated actin pedestal formation are the translocated intimin receptor (Tir) and EspFU. EspFU consists of an N-terminal secretion signal and a C-terminus containing six tandem 47-residue proline-rich repeats, each of which can bind and activate the actin nucleation promoting factor N-WASP. Structural and functional analyses described here have identified the mechanism of N-WASP activation by EspFU and the minimal domains and specific residues required for this activity. While EspFU and Tir are the only bacterial effectors required for F-actin pedestal formation, recruitment of EspFU to Tir is mediated by an unidentified putative host factor. To identify the host factor responsible for linking these two effectors, a combination of in vitro and functional assays were used to identify the host factor, IRTKS and the residues required for these interactions were defined. Further, the presence of at least two 47-residue repeats in all characterized clinical isolates of canonical EHEC strains led us to address the minimal requirements for EspFU functional domains to promote recruitment to Tir and N-WASP activation. Here we show that two proline-rich elements of EspFU are required for recruitment of EspFU by IRTKS to sites of bacterial attachment. Furthermore, once artificially clustered at the membrane, a single N-WASP binding element of EspFU can induce actin pedestal formation.

Newcastle disease virus (NDV) is a member of the genus Avulavirus of the Paramyxoviridaefamily of enveloped negative-stranded RNA viruses. The virus causes respiratory, neurological, or enteric disease in many species of birds, resulting in significant losses to the poultry industry worldwide. Strains of the virus are classified into three pathotypes based on the severity of disease in chickens. Avirulent strains that produce mild or asymptomatic infections are termed lentogenic, whereas virulent strains are termed velogenic. Strains of intermediate virulence are termed mesogenic. The envelope of NDV virions contains two types of glycoproteins, the hemagglutinin-neuraminidase (HN) and fusion (F) proteins. HN mediates three functions: 1) virus attachment to sialic acid-containing receptors; 2) neuraminidase activity that cleaves sialic acid from progeny virions to prevent self-aggregation; and, 3) complementation of the F protein in the promotion of fusion. Though it is widely accepted that cleavage of a fusion protein precursor is the primary determinant of NDV virulence, it is not the sole determinant. At least two other proteins, HN and the V protein, contribute to virulence. The V protein possesses interferon (IFN) antagonistic activity. The long-range goal of these studies is to understand the roles of HN and V in the differential virulence patterns exhibited by members of the NDV serotype. The first aim is to compare the IFN antagonistic activity of the V protein from a lentogenic and a mesogenic strain of the virus. The results of this study demonstrate that the V protein of the mesogenic strain Beaudette C (BC) exhibits greater IFN antagonistic activity than that of the lentogenic strain La Sota. Hence, the IFN antagonistic activities of the two V proteins correlate with their known virulence properties. Comparison of the C-terminal regions of La Sota and BC V proteins revealed four amino acid differences. The results demonstrate that the IFN antagonistic activity of La Sota V increases when any one of these residues is mutated to the corresponding residue in BC V. Conversely, the IFN antagonistic activity of BC V decreases when any one of these four residues is mutated to the corresponding residue in La Sota V. However, no single residue accounts for the difference in IFN antagonistic activity between the two V proteins. Also, analysis of La Sota V and BC V proteins with multiple mutations in these positions revealed that the four residues are collectively responsible for the difference in the IFN antagonistic activity of the two V proteins. Finally, characterization of chimeric La Sota/BC V proteins showed that the N-terminal region also contributes to the IFN antagonistic activity of V. Contrary to an earlier report, results described here demonstrate that the NDV V protein does not target STAT1 for degradation. However, both La Sota and BC V proteins target interferon regulatory factor (IRF)-7 for degradation and promote the conversion of full-length IRF-7 to a lower molecular weight form (IRF-7*). This is the first demonstration that IRF-7 is targeted by a paramyxovirus V protein. The amount of IRF-7* decreases in a dose-dependent manner in the presence of a proteasome inhibitor, suggesting that IRF-7* is a degradation product of IRF-7. Furthermore, the BC V protein promotes complete conversion of IRF-7 to IRF7*, whereas the La Sota V protein does so less efficiently. Again, this is consistent with the difference in IFN antagonistic activity of the two V proteins, and in turn, with their virulence. The second aim is to characterize an HN-specific monoclonal antibody called AVS-I. A previous study suggested that AVS-I recognizes an epitope that is conserved in lentogenic strains and raises the possibility that this epitope may colocalize with a determinant of virulence in HN. To further characterize antibody AVS-I and the epitope it recognizes, we (i) determined its specificity for several additional strains of the virus, (ii) mapped its binding to HN in competition with our own antibodies, (iii) determined its functional inhibition profile, and (iv) isolated and sequenced an AVS-I escape mutant. The results demonstrate that AVS-I binds to a conformational epitope at the carboxy terminus of HN. This suggests that this region of HN may define a determinant of virulence. However, it was also shown that AVS-I, which was previously thought to be specific for avirulent strains of NDV, actually recognizes individual mesogenic and velogenic strains. In conclusion, the data presented in this dissertation contributes to a greater understanding of the molecular basis for NDV virulence and may aid in development of antiviral strategies and generation of recombinant NDVs suitable for use in cancer and gene therapy.