Reoviruses are double-stranded (ds) RNA viruses containing a genome of 10 dsRNA segments. A unique feature of reovirus replication is that each progeny virus contains one, and only one, of each of the 10 segments of dsRNA. However, the genetic elements that control the sorting and assembly of genome segments were not known. The M1 gene encodes the $\mu$2 protein which is a minor component of the virus core. Genetic studies have associated the M1 gene with cytopathic effect in cultured cells and pathogenesis in infected animals. The genetic basis for this function was not clear either. Because of the segmented nature of reovirus genome, coinfection of cells with two different serotypes can produce progeny viruses that contain genome segments from both serotypes (reassortants). Type 1 and type 3 reassortants that possess the type 3 L2 segment and type 1 M3 segment generate M1 segment deletion (smaller M1 segments) on serial passage (Brown et al., 1983). By analyzing reovirus M1 deletion mutants, it was found that both termini of the M1 gene were conserved. The results indicate that these two terminal regions contain all the genetic signals sufficient for the replication and assembly of the M1 gene and that 344 nucleotides could be the minimum length required for packaging into virions. Fifty-one nucleotide substitutions and 10 amino acid substitutions were identified between type 1 and type 3 M1 segments. These substitutions must be responsible for the phenotypic differences in cytopathic effect and pathogenesis that map to the M1 gene by genetic analyses. The M1 gene was further characterized by expression in mammalian cells. One $\mu$2-expressing L929 cell line was tested for its ability to support the growth of a reovirus temperature-sensitive (ts) mutant with a defect in its M1 gene. The restricted replication of the ts mutant at nonpermissive temperature was complemented by the $\mu$2-expressing cell line, demonstrating the feasibility of cultivating reoviruses with defective M1 genes. Mutagenesis of the M1 gene-containing DNA constructs and expression studies also clarified the translation initiation site of the M1 gene. It was found that translation of the M1 gene initiates from the first AUG codon starting at nucleotide 14 and that if there is initiation from AUG$\sb{161}$ as in 5$\sp\prime$-terminus-truncated constructs, a protein about 5 kDa smaller than $\mu$2 protein is produced. With the information about the M1 gene obtained from this study, preliminary reverse genetics work was initiated in attempts to further characterize the genetic signals within the conserved M1 termini and to develop reovirus as an expression vector. (Abstract shortened by UMI.)

MyoD is a member of the myogenic regulatory family of transcription factors which play a pivotal role in the formation of skeletal muscle. Stable expression of MyoD in the P19 embryonal carcinoma cell line leads to enhanced muscle formation only after aggregation of these cells. To study how MyoD and cell aggregation cooperate to affect muscle differentiation, a myc epitope tagged version of the protein was generated which functioned like the native MyoD protein. The production of an expression construct containing an internal ribosome entry site (IRES) was required to achieve efficient expression of mycMyoD using a drug resistance gene. With this expression construct it was determined that the level of mycMyoD protein did not increase after aggregation. Similarly no changes in cellular localization were observed as MycMyoD was located in the nucleus both before and after aggregation. There was no apparent change in dimerization partners upon aggregation as mycMyoD was found bound to E2A proteins in both the non-aggregated and aggregated cells. In a DNA binding assay mycMyoD from both cell types was able to bind to an E-box containing oligonucleotide. Thus, the effect of cell aggregation is not upon mycMyoD but rather is believed to be either at the point of DNA accessibility by mycMyoD:E2A protein heterodimers or at the stage of transcriptional activation. In vivo, these means of regulation are likely to ensure that myogenesis occurs only when proper cell contacts are achieved.

To elucidate the roles of the junctional elements in HPIV3 transcription, cDNAs were constructed containing CAT and luciferase reporter genes flanked by sequences representing the HPIV3 termini necessary for transcription, replication and packaging. Mutations to the gene end sequence abolished expression of the upstream and downstream genes. Deleting the gene start sequence at the junction resulted increased expression of the upstream gene, but abrogated downstream gene activity. Alterations in the length of the intergenic trinucleotide resulted in decreased expression of both upstream and downstream genes. Mutations in the sequence of this nontranscribed trinucleotide resulted in decreased activity of the upstream gene but no change in expression of the downstream gene. The gene end sequence does not appear to contain the only signals for termination of transcription. The purine trinucleotide intergenic region is important for termination, but only the presence of three nucleotides appears to be necessary and sufficient for expression of the following gene. Results obtained from assaying reporter activity could often be interpreted in several ways. For example, the data could not distinguish between polymerase readthrough and premature termination. Two RNA detection methods were investigated and show promise as means for detecting and analyzing specific RNA species in transfected cells. (Abstract shortened by UMI.)

The Human Cytomegalovirus, a member of the Herpesvirus family, is a relatively non-pathogenic virus except in those individuals whose immune systems have not fully developed (newborns) or are immunosuppressed (transplant patients) or immunodeficient (individuals with AIDS). A rapid test to detect the virus in clinical samples is needed in order to assess the status of an HCMV infection and to monitor the virus in an individual over time. Faced with the alternative detection systems such as viral culture, immunofluorescence assays and DNA probing the polymerase chain reaction is an attractive choice because of its rapidity and unrivalled sensitivity. Nested polymerase chain reaction for HCMV was used in the present work in order to determine how useful it is in the detection of the virus in peripheral blood samples. Two nested assays were evaluated and optimized using blood samples from HIV-1 seronegative blood donors, HIV-1 seropositive asymptomatic individuals and individuals with AIDS. The results show that the assay is rapid, sensitive and specific. Furthermore, the virus was readily detectable by PCR in individuals with AIDS while it is infrequently found in the other two groups. Finally, the PCR results were positive at 14 and 7 months, respectively, in two individuals with AIDS prior to the development of overt HCMV mediated disease suggesting that PCR is a useful monitoring tool in individuals at high risk for HCMV disease. PCR has many applications in basic and applied research and potentially in the detection of infectious diseases. The results here suggest that its use may help to predict who may and may not develop the potentially sight and life-threatening diseases caused by HCMV.

RAR$\beta$ has been shown to be expressed in the developing heart at the 8 somite stage. Differentiation of EC cells with DMSO produces a mixture of embryonic cardiac, skeletal, endodermal and other mesodermal derivatives. The RAR$\beta$-2 isoform is predominantly expressed in the differentiated cardiac muscle cells. RT-PCR with isoform specific primers has been used to identify RAR$\beta$-2 as the predominant isoform. These results show, that like RA-treated EC cells, expression of the RAR$\beta$-2 isoform in DMSO differentiated P19 cells predominates. Stable transformants of the RARE$\beta$-2-CAT reporter gene construct were pooled and differentiated with either RA or DMSO. The results obtained demonstrate that RA induced CAT activities increased to a maximum of 188 fold by day 3. The increase in RAR$\beta$-2 mRNA levels was not due to another enhancer element as evidenced by nuclear run-on analysis. RAR$\beta$2 mRNA levels seem to be regulated by a post-transcriptional mechanism. The stability of the DMSO induced RAR$\beta$2 message was determined using the transcriptional inhibitor actimomycin-D on day 4 and 7 of differentiation. The half life of the message was found to be approximately equal on both days. We postulate that stabilization of the message occurs at an earlier time point, and that accumulation of the message occurs up until Day 7 of differentiation. Simultaneous administration of RA and DMSO to aggregating P19 cells blocked the up-regulation of Brachyury expression in these cells. No beating cardiac muscle was formed, however neuroectoderm was found to be the predominant cell type, indicating that the DMSO differentiating effect was usurped by the presence of RA. (Abstract shortened by UMI.)

Nuclear receptors and POU transcription factors have been shown to regulate gene transcription through complex cis-acting elements located in upstream promoter regions of target genes. Regulation has been shown to occur through both synergistic and inhibitory mechanisms. Specifically, glucocorticoid dependent transcriptional synergism has been observed within complex regulatory elements on the mouse mammary tumor virus Long Terminal Repeat (MMTV LTR) that contain binding sites for both the glucocorticoid hormone nuclear receptor (GR) and the ubiquitous POU factor, octamer transcription factor-1 (Oct-1). By contrast, on promoters like that of the histone H2B gene that contains regulatory elements only for Oct-1, glucocorticoids actually appear to inhibit the activation of gene transcription by Oct-1. Previous experimental observations were suggestive of a direct physical interaction between the DNA binding domain of GR and the POU DNA binding domain of Oct-1. In this work I have examined the requirements for the binding of in vitro translated, radiolabeled GR to the POU domain of Oct-1 expressed as a glutathione-S-transferase (GST) fusion protein and immobilized on glutathione sepharose. (Abstract shortened by UMI.)

Plateau-phase normal human skin fibroblasts (AG1522) pre-exposed to low-dose-rate ionizing radiation (IR) became less susceptible to the lethal effect of a subsequent acute challenge dose of radiation. A threshold and an optimum adapting dose were observed. This adaptive response (AR) at the survival level was accompanied by a decreased number of chromosomal breaks due to the challenge dose, as indicated by a reduction in the frequency of micronucleus formation. The frequency of micronucleus formation was further reduced when an incubation period at 37$\sp\circ$C separated the challenge dose and the adapting dose delivered at either low- or high-dose-rate. The rate of reduction of micronucleus formation was higher for the low dose-rate as compared to the high dose-rate adapting dose, suggesting that the ability of human cells to adapt to radiation increases with decreasing rates of damage. A certain amount of damage per unit time with which the cell can cope appears to be necessary to trigger the AR to IR. The rate of repair of DNA double-strand breaks, as indicated by the frequency of micronucleus formation, was higher in adapted cells, suggesting that the mechanism of adaptation could include increased repair capacity and/or an increased ease of access of repair enzymes to the lesion. Adapted cells also showed a much longer delay in reaching the binucleate state than non-adapted cells, suggesting a second mechanism of adaptation which may increase the time available for DNA repair. The analysis of RNA from adapted cells showed a decreased level of cyclin A and cyclin B transcripts consistent with a mechanism leading to a delay in the progression of the cells in the cell cycle. The transcript levels of other genes possibly involved in the cellular response to IR were also altered. Rodent C3H 10T${1\over2}$ cells showed a similar adaptation when assayed for micronucleus formation. The adapted cells were also protected against transformation to malignancy by a subsequent high dose of radiation. Transformation frequency was reduced about two-fold by low-dose-rate adapting doses ranging from 0.1 to 1.5 Gy. Flow cytometric measurements showed that the cell cycle distribution of the plateau phase cells used was unaltered during the various treatments, indicating that the observed AR cannot be attributed to selection of cells at a radioresistant stage of the cell cycle.

Myotonic dystrophy (DM) is an autosomal dominant genetic disease which affects approximately 1 in 8000 individuals globally. This is a multisystemic disorder which primarily targets muscle tissues. The genetic defect underlying DM is a highly unstable trinucleotide CTG repeat sequence located in the 3$\sp\prime$ untranslated region of a gene encoding a protein with serine/threonine protein kinase activity. The number of CTG repeats in non-DM individuals ranges from 5 to 35, whereas in DM individuals it can range from 50 to over 2000. The mechanism of disease and the role of the kinase are currently unknown. The main characteristics of the mutation are its expanded length and high instability. The instability usually leads to an increase in the number of CTG repeats as the mutant allele is transmitted from one generation to the next. This is seen concurrently with an increase in the severity of the clinical phenotypes through successive generations. The mechanism(s) underlying the unstable behaviour of this mutation have been unknown. I studied several possible elements that might affect the observed instability. These studies were performed on a large number of intergenerational transmissions of the mutant allele in DM patients, as well as on different tissues from the same patients. In vivo studies revealed that both the gender of the transmitting parent and the size of the mutation had a significant effect on its intergenerational dynamics. Instability was also seen in cells that underwent mitosis and meiosis, as well as in patients' cells with very low mitotic activity. In addition, in vitro studies showed instability of this sequence in cells from DM patients which were grown in tissue culture. This instability was seen to not necessarily be associated with the cell-cycle-coupled DNA replication in those cells. Since my previous analyses had ruled out recombination between homologous chromosomes as a major element involved in the instability I focused my studies on tests for the possibility that DNA repair was associated with the instability of this mutation. These analyses revealed that there were breaks or gaps occurring specifically within the mutation. Since breaks and gaps are DNA repair intermediates this suggested that the high instability of this mutation in DM patients was due to repair attempts on the structure adopted by the long arrays of CTG repeats. The results of the previous studies on intergenerational transmissions of the mutation, as well as in vivo and in vitro studies of the unstable properties of this sequence are also compatible with a DNA repair model of repeat instability. These data point the way to a more focused approach to the identification of specific mechanistic pathway's underlying the expansion of this highly unstable CTG repeat in DM. The results of this research may also help in defining the mechanism(s) of expansion of the mutations of an increasing number of human genetic diseases found to be caused by highly unstable trinucleotide repeats.

Dual specificity kinases are known to have important roles in cellular regulation. The Sty dual specificity kinase has been shown to segregate with, and phosphorylate, SR family splicing factors, but the consequences of this are not known. Additionally, it is not known if the kinase has any other regulatory functions. In an attempt to determine the function of this protein in a mammalian system, Sty deficient mice were generated. This was accomplished through gene targeting, in ES cells, with a promoterless IRES-$\beta Geo$ based vector. Using this efficient system, we obtained an 82% recombinant frequency. This is in contrast to a PGK-neo based vector, targeting the same locus, for which none of 377 clones were homologous recombinants. The knockout mice generated exhibit no overt phenotype, though more detailed analysis is still in progress. A possible explanation for the lack of a phenotype is a potential redundancy between Sty and closely related family members, one of which has recently been identified.

The Clk (cdc2-like kinase) family of kinases has been well conserved throughout evolution and includes members from yeast to humans. In mammalian systems, four members have been identified: Clk1, Clk2, Clk3, and Clk4. Although Clk kinases have been implicated in the regulation of alternative splicing, their exact biological function is unknown. Our lab has used a gene targeted approach to ascertain loss-of-gene function of murine Clk kinases. The work presented in this thesis describes the cloning and gene targeted disruption of murine Clk2 in J1 embryonic stem cells. Two lambda phage clones were isolated from a D3 genomic library, using full length human Clk2 cDNA as a probe. Employing primers derived from the genomic sequence, a 581 bp Clk2 partial cDNA fragment was cloned from P19 embryonal carcinoma cells by RT-PCR. Using DNA fragments isolated from the genomic clones, a promoter trap targeting vector was constructed. (Abstract shortened by UMI.)