Epstein-Barr virus (EBV) is associated with a number of human diseases and does not have a vaccine. It is believed that neutralizing antibodies are an important immune effector for an EBV vaccine, but it is unknown whether serum neutralizing antibodies can alter EBV infection through the oral mucosa. The studies presented in this dissertation were designed 1) to adapt the rhesus macaque animal model to allow testing of neutralizing antibodies in a biologically relevant system and 2) to better define the neutralizing antibody response of EBV infected humans. Infection of rhesus macaques with the EBV related lymphocryptovirus, rhLCV, provides an accurate model system for studying EBV, but there were two hurdles that needed to be overcome before neutralizing antibodies could be tested in this model. First, there are no neutralizing antibodies specific to rhLCV and we found that a potent EBV neutralizing antibody, 72A1, did not cross-neutralize rhLCV. Second, murine monoclonal antibodies are inherently immunogenic in macaques and induce anti-antibody responses that limit their utility. To create a virus sensitive to 72A1 neutralization, the major membrane glycoprotein of rhLCV with replaced with EBV gp350. The data presented here show that this chimeric virus can use EBV gp350 to support entry into macaque cells in vitro and following oral inoculation of rhesus macaques. To reduce the immunogenicity of the murine antibodies, “rhesusized” antibody variants were generated and shown to retain antigen specificity. The combination of this novel, chimeric virus and the “rhesusized” antibodies will now allow testing of neutralizing antibodies in the macaque model. Although multiple EBV glycoproteins have been shown to induce neutralizing antibodies in mice, studies of the human neutralizing antibody response have been narrowly focused on a single antibody binding epitope on gp350. Here we show that antibodies binding to this epitope do not represent all EBV neutralizing activity in human sera. Additionally, these data suggest that the neutralizing response is much broader than appreciated, with multiple glycoproteins inducing EBV neutralizing antibodies. Accurately defining the repertoire of viral glycoproteins targeted by human neutralizing antibodies can inform us of the naturally immunogenic proteins that may make good vaccine immunogens. / Biology, Molecular and Cellular

The distribution of lentiviruses among primates reflects a history of interspecies transmission and emergence of new virus-host relationships. The degree to which viruses must adapt to the genetic environment of new host species, and how adaptations to the new host initially affect viral fitness are two understudied elements of emergence. The simian immunodeficiency virus (SIV) of rhesus macaques (SIVmac) emerged as the result of a cross-species transmission of SIV from the sooty mangabey monkey (SIVsm) into rhesus macaques, and comparing cohorts of SIVmac- and SIVsm-infected macaques provides an opportunity to examine a lentivirus evolving during the early stages of emergence. Using archived samples from four cohorts of macaques, we compared evolution of “established” macaque-adapted viruses (SIVmac239, SIVmac251) to incompletely-adapted, “emerging” viruses (SIVsmE543, SIVsmE660). Longitudinal samples included the inoculum for each cohort, as well as acute and chronic plasma samples for each animal. Samples were processed for deep sequencing, and consensus sequences of complete viral coding regions were assembled de novo. Computational and manual analysis of the sequences revealed a set of loci that diverged considerably only in the SIVsm-infected animals, suggesting that adaptations at these loci are important for emergence of SIVsm in rhesus macaques. These candidate adaptations included known adaptations to overcome restriction by macaque TRIM5α. In order to quantify the impact of these candidate adaptations on viral replication, each mutation was introduced into SIVsmE543 (forward mutations, reflecting adaptation to the macaque host) and SIVmac239 (reversions to the ancestral residue). These were then tested in a deep sequencing-based fitness assay, in which changes in the frequencies of mutant and parental sequences replicating in cell culture were used to calculate differences in relative fitness. Substitutions in the coding sequences for the Matrix, Capsid, and Vif proteins were found to enhance fitness of SIVsm in rhesus cells, confirming the hypothesis that they represent species-specific adaptations. Together, these studies represent a novel approach to the identification and functional characterization of viral determinants of cross-species transmission. / Medical Sciences

Elicitation of neutralizing antibodies (NAbs) remains a major goal in the generation of a successful vaccine against human immunodeficiency virus type 1 (HIV-1). One of the major obstacles in generating such NAbs is the vast sequence diversity of their target, the HIV-1 envelope (Env) protein. This dissertation will describe multiple vaccination strategies utilizing a soluble form of Env (gp140) to address the sequence diversity of HIV-1 Env and to improve the NAb responses elicited through vaccination in the guinea pig model. A bioinformatically optimized mosaic gp140, designed as a single sequence to cover global HIV-1 sequence diversity, was utilized in combination with a clade C gp140, resulting in a complementary repertoire of NAbs when compared to vaccination with each single protein. Additionally, a quadrivalent mixture of four novel clade C gp140s with distinct antigenic properties was found to elicit a greater magnitude of NAbs when compared to any single protein in the mixture. Furthermore, longitudinal, heterologous prime/boost vaccination regimens resulted in binding antibodies to a greater number of distinct sequences within a single epitope of variable loop 2, and NAb of a greater magnitude, than a homologous prime/boost regimen. Finally, a panel of rationally designed variable loop 2 and 3 sequence modified trimers was designed. When utilized in a mixture, the variable loop 2 modified trimers elicited a greater magnitude and breadth of NAbs than the single, unmodified wild type protein alone. These data show that multivalent vaccination strategies with HIV-1 Env trimers result in improved NAb responses compared to single Env protein vaccinations and may help to guide the development of future vaccination regimens. / Medical Sciences

No description available.

This thesis describes three phases of our studies on the biology of human coronavirus, 229E: the development of a plaque assay system, studies on the effect of Actinomycin D on the replication of this virus, and the isolation and preliminary characterization of an internal component of the virion. A medium was developed which permitted the production of plaques by 229E on monolayers of L 132 cells in closed containers which could be incubated in regular incubators, a result which we were not able to duplicate using published formulations. This assay system exhibited a linear dose-response curve over a wide range of virus concentrations and was applicable to the assay of this virus in several cell lines; in addition, it could be used for the assay of other viruses. Several other features of the assay are also discussed briefly. The replication of 229E in L 132 cells was found to be inhibited in the presence of Actinomycin D, a potent inhibitor of DNA-dependent RNA synthesis. While the drug has adverse effects on the host cell, several arguments have been brought forth which argue in favor of a specific effect on virus replication; the replication of vesicular stomatitis virus was unimpeded at 10 times the 50% inhibitory level of AMD for 229E, and largely unaffected at 100 times the mean inhibitory dose; in addition the cells were not themselves greatly affected and the time course of drug sensitivity appeared to indicate a temporally discrete period of AMD susceptibility. The sensitivity of AMD was shown to be dependent on the virus input used to initiate the replicative cycle. Using disruption by the detergent Nonidet P-40, we were able to isolate, in sucrose gradients, a subviral particle. This particle, a nucleoprotein, has a density of 1.27 and appears linear in the electron microscope. It is suggested that this structure probably represents the viral nucleocapsid. Some further aspects of its relation to virus morphology are discussed.

Both pseudorabies virus and the various strains of avian sarcoma and leukosis viruses are economically important pathogens. In Canada, pseudorabies is an exotic virus which primarily infects swine. The disease is rapidly fatal in most species although adult pigs may become carriers. Because pseudorabies virus (PrV) is exotic to Canada, imported animals are rigorously tested for the presence of antibody titers to the virus. The leukosis caused by Rous associated virus (RAV) has a major impact on the health of poultry flocks. Different strains of the avian leukosis viruses cause damage of differing severity but the general characteristics of infection include decreases in both egg and meat production. There would be significant benefits if non-infectious, antigenic fractions from both of these viruses could be produced in bacterial cells. These proteins could be used in serum antibody testing to make the detection of both viruses considerably simpler than the currently used ELISA and virus neutralization tests. As well, the production of large quantities of RAV env protein could simplify studies of the mechanisms of host specificity and provide additional information about chicken cell receptors. I have attempted to produce an antigenically recognizable protein from both pseudo-rabies virus and Rous associated virus. Attempts were made to express such a protein from pseudorabies virus by cloning either cDNA or genomic DNA fragments into plasmid vectors followed by colony immunodetection using a polyclonal antibody against PrV prepared in pig. Although several recombinants produced positive immunological results, further examination showed that none contained PrV DNA sequences. Standard screening methods were modified and improved during the course of this work. Subsequently, other laboratories identified significant difficulties in using unpurified polyclonal antibodies prepared in swine which provided an explanation for the results I observed. A cDNA library constructed from RNA prepared from RAV-1 was screened using a probe prepared from env sequences in pSR-XD2, a plasmid containing much of the RSV sequence. Transformants containing RSV homologous DNA were identified. During purification of these clones and subsequent screenings many of the plasmids identified as containing RSV homologous sequence lost their inserts. Additional studies demonstrated that plasmids which exhibited the strongest homology with the RSV probe and which contained the longest inserts, lost these inserts at a high rate. A second cDNA cloning using a different bacterial host and a different viral strain (RAV-2) produced several positive clones, with insert lengths ranging from 250 by to 950 bp. These inserts were sequenced and compared to known RSV and RAV sequences. A high degree of homology was observed. The longest RAV-2 cDNA fragments included most of the env gp37 polypeptide coding region. In order to test for expression of an antigen, inserts from two independently isolated recombinants were subcloned into lambdagt11 in each of the three reading frames. A polyclonal antibody prepared against Rous sarcoma virus was used to test for the expression of any recognizable antigenic determinants. No antigen production was detected even though nucleic acid plaque hybridizations demonstrated that the subcloning had been successful in introducing RAV-2 sequences into lambdagt11. Possible explanations for this, as well as a discussion of the composition of the RAV-2 sequences and their placement on the RSV genetic map are considered.

Abstract not available.

The natural signal sequence of HIV-1 gp120 contains an unusually long hydrophobic domain and five positively charged amino acids. When the gp120 gene was cloned into a baculovirus expression vector under the control of the baculovirus polyhedrin gene promoter, it exhibited an extremely low level of secretion. However, deletion of the signal sequence resulted in the production of large quantities of a nonglycosylated form of gp120 and fusion of honeybee melittin or murine interleukin 3 signal sequences, which contain only one or no positively charged residues, respectively, resulted in a high level of expression as well as glycosylation and secretion. Four charge-altered signal mutants were generated by oligonucleotide-directed mutagenesis. Positively charged amino acids in the natural signal sequence were substituted with neutral amino acids. The results of these experiments showed that the expression and secretion of gp120 was progressively increased by decreasing the positive charge in a stepwise fashion from + 5 to + 3, + 2, and + 1. However, elimination of all five positive charges (leaving a net negative charge of -1 at the NH 2 terminus) caused accumulation of large amounts of a nonglycosylated form of gp120 but decreased the amounts of glycosylated forms of gp120. These signal peptide mutants clearly demonstrate that the positively charged amino acids in the natural signal sequence of HIV-1 gp120 are key factors determining its poor expression and secretion in insect cells. Analysis of intracellular transport and folding of gp120 further indicates that the highly charged uncleaved signal peptide rather than disulfide bond formation is an important factor limiting transport of gp120 from the rough endoplasmic reticulum (RER) to the Golgi apparatus; its presence affects gp120 folding and slows its rate of transport to the cell surface. The requirement for carbohydrate on HIV gp120 in CD4 binding has been the subject of much debate. There have been conflicting reports regarding the role of gp120 glycans in binding to CD4. An important question is whether the carbohydrate itself plays an important role in this interaction. Nonglycosylated and glycosylated forms of gp120 from HIV-1 and HIV-2 were produced using the baculovirus expression system and their CD4 binding properties were determined. The nonglycosylated forms of gp120 generated by either deletion of the signal sequence or synthesized in the presence of tunicamycin failed to bind to CD4. In contrast, highly mannosylated recombinant gp120 bound well to soluble CD4. Enzymatic removal of carbohydrate chains from glycosylated gp120 by endoglycosidase H (endo H) or by a mixture of endoglycosidase F and N-glycanase (endo FNG) in the presence or absence of SDS had little or no effect on the ability of gp120 to bind CD4. The data indicate that carbohydrate chains per se do not play a significant role in interaction between gp120 and CD4 molecules but that N-linked glycosylation is required for correct protein folding that provides the proper conformation for CD4 binding. Analysis of intracellular folding of gp120, using its ability to bind CD4 as a functional assay for overall conformation, further supports the hypothesis that N-linked glycosylation of HIV gp120 plays an essential role in promoting either the correct folding of the protein or in its stabilization.

To explore whether rabies viral variants co-localize with discrete bat host populations (sub-populations), both the host genome and rabies virus of Eptesicus fuscus (big brown bats), Myotis lucifugus (little brown bats) and other Myotis specimens, collected during 1989 to 2004 from diagnostic submissions from across the country, were genetically characterized. Bat species population analysis was performed by nuclear DNA genotyping, scored by variation of several microsatellite loci, and through phylogenetic analysis of Cox-1 (cytochrome oxidase subunit I) gene sequences located on mitochondrial DNA. Microsatellites are relatively short DNA stretches consisting of tandem repeats of one to five nucleotides which exhibit high levels of allelic variation. Cox-1 gene sequence analysis provides accurate species level of identification for Myotis lucifugus specimens. Two hundred and ninety five DNA samples of Eptesicus fuscus were examined at 9 microsatellite loci, and 126 DNA samples of Myotis lucifugus were examined at 7 microsatellite loci, both datasets were analyzed by a series of genetic population analysis softwares. Phylogeny of Cox-1 gene sequences with 552 nucleotides by using 106 DNA samples of bats Myotis lucifugus was analyzed as an alternative strategy of the microsatellite gene marker for population structure determination of Myotis lucifugus . Consequently, two populations---East (group I) and West (group II) were determined for Eptesicus fuscus bats in Canada. No population structure was identified for Myotis lucifugus bats. Rabies viral variants were identified by nucleotide sequencing of the central divergent portion of the P (phosphoprotein) gene---a region previously proved to be a sensitive target for molecular epidemiology analysis. Viral RNA, isolated from 231 samples of Eptesicus fuscus and Myotis species was sequenced over a 597 by region. Phylogenetic analysis of these data identified six rabies viral variants circulating in Eptesicus fuscus and one rabies viral variant circulating in Myotis species. Based on the results of genetic characterization, the spatial distribution of the bat host subpopulations and their associated rabies virus variants were determined. Rabies variants I and II circulate in Eptesicus fuscus population I (East); rabies variants III, IV and V circulate in Eptesicus fuscus population II (West); and rabies variant VI circulates in both populations. A distinct rabies virus variant VII was associated with Myotis species. No population substructure would be identified for Myotis lucifugus bats.

Inhibition of the interferon-mediated antiviral response is a major determinant of virulence in influenza A virus. The NS1 protein of influenza A virus has been identified as the IFN antagonist. However, the specific mechanisms for IFN antagonism are not known. NS1 binds both the PKR as well as single and double stranded RNA to inhibit activation of PKR. Adaptation of human influenza virus to the mouse lung may likewise involve mutations that affect IFN antagonism. The prototype human influenza A virus A/HK/1/68 (H3N2 subtype) was utilized for this project. Six mouse-adapted variants possessing four different point mutations on NS gene were obtained from independently derived mouse-adapted variants. The mutations were located in the RNA binding domain and several sites in an 8 amino acid region from as 98 to 106. Mutant HKMA20c was of interest because it possessed a mutation in common with highly pathogenic avian influenza virus H5N1 (Leucine on AA 103 on NS1 protein). Base on these phenomena, a hypothesis was brought out that this region may encode a site of interaction with a host or viral factor and that mutation(s) on it may enhance the ability of NS1 protein to function as an IFN antagonist. The approach was to first characterize the IFN resistance and IFN induction properties of HK mouse-adapted mutants as well as pathology in mice lung, followed by the generation of defined recombinant viruses possessing desired mutation(s), then analysis of these recombinant viruses for IFN induction and IFN resistance. Rescuing of the HK wild type NS gene and those of HK mouse-adapted NS mutant genes into the backbone of parental HK genome discovered that NS20, NS20c and NS411 produced attenuating phenotypes on their own. Finally, all these mutant NS genes were inserted into backbone of WSN (lab adapted strain, A/WSN/33, H1N1) genome to construct recombinant viruses. Using recombinant viruses that differ due to individual NS1 gene showed that all of the NS1 mutations increased resistance to IFN in mouse cells. The extent of IFN resistance due to individual mutation was influenced by cell types: epithelium versus fibroblast, as well as host type: mouse versus human. Interestingly most of the NS1 mutations attenuated growth of virus which suggests that resistance to IFN involves changes in host interaction that are not optimal for growth in the absence of IFN. Infection assay showed mouse-adapted mutant NS genes conferred resistance to mouse IFN and vulnerability to human IFN. And mutant NS1 protein NSMA20c had the most potent ability to resist human and mouse IFN. IFN assay showed that mutant NS1 protein NSMA20c had the highest ability to induce mouse IFN. Furthermore, all synthetic recombinant viruses induced low amounts of IFN in human cells. Immunopathology of infected lungs showed that mouse adapted progeny virus HKMA20C had acquired a crucial ability to spread to and infected alveoli which may be influenced by IFN resistance. Mouse-adapted variants possess mutations that increase IFN resistance that in some instances leads to higher IFN induction. One of the key discussions was that most of these mutations cluster in a small region that has been previously mapped to involve a region of host protein (eIF4GI) interaction.