Murine Clk1 (also known as Sty) was identified as a dual-specificity kinase capable of phosphorylating serine, threonine and tyrosine residues when expressed in bacteria. Expression of Clk1 is deveolpmentally regulated at the level of RNA expression. Embryonic cells express two mRNAs of similar size, whereas differentiated cells express two additional transcripts of larger size. Little is known about the bochemical and biological activity of Clk1 in mammalian cells. We demonstrate that the two embryonically expressed mRNAs are derived through alternative splicing of a unique exon (exon EB) within the Clk1 precursor mRNA. These mRNAs encode full-length catalytically active Clk1 and a truncated inactive polypeptide, Clk1$\rm\sp{T}.$ Larger incompletely spliced Clk1 transcripts accumulate in differentiated cells. When expressed in mammalian cells Clk1 possesses dual-specificity kinase activity and is capable of forming complexes with other molecules of Clk1 and Clk1$\rm\sp{T}.$ The regions involved in binding map to the amino-terminal non-catalytic domain of Clk1. Phosphorylation sites map to the amino acids encoded by the alternatively splice exon EB. Clk1$\rm\sp{T}$ and catalytic mutants of Clk1 co-localise with splicing factors in intranuclear speckles, whereas catalytically active Clk1 causes a redistribution of these factors within the nucleus. This activity requires the presence of amino acids encoded by exon EB. These results suggest a role for Clk1 and Clk1$\rm\sp{T}$ in the regulation of RNA splicing. Splicing of a Clk1 mini-gene, encompassing exon EB, in vivo is regulated by Clk1 and Clk1$\rm\sp{T}.$ Catalytically active Clk1 stimulates exclusion of EB leading to the production of Clk1$\rm\sp{T}$ mRNA. In contrast, Clk1$\rm\sp{T}$ promotes EB inclusion leading to production of Clk1 mRNA. Two Clk1-related human kinases, hClk2 and hClk3, also exhibit dual-specificity kinase activity and cause the redistribution of nuclear splicing factors. Similar to Clk1$\rm\sp{T},$ the hClk truncated isoforms, hClk2$\rm\sp{T}$ and hClk3, co-localise with splicing factors in nuclear speckles. hClk2 and hClk3 are able to influence the splicing pattern of a murine Clk1 mini-gene in vivo, indicating that they can also regulate precursor mRNA splicing. Taken together these results imply a role for the Clk family of kinases in the regulation of gene expression at the level of RNA processing.

Human cytomegalovirus (CMV) is a herpesvirus that is responsible for significant morbidity and mortality in the setting of the acquired immunodeficiency syndrome (AIDS). The objective of this work was to devise a sensitive assay for the early diagnosis of CMV infection in HIV-infected individuals at highest risk of developing clinical CMV disease based on their CD4$\sp+$ T cell count and clinical status. A novel PCR assay was developed which can be applied to detect both latent and active CMV DNA in appropriate clinical specimens. Sensitivity and specificity were established in comparison with conventional diagnostic tests, including cell culture and serology. Other herpesvirus DNA (EBV, HSV-1, HSV-2) and bacterial (Neisseria gonorrhoeae) DNA were used to confirm the specificity of the assay. A sensitive, specific and semi-quantitative PCR protocol was applied to detect both latent and active CMV DNA in appropriate clinical specimens. We hypothesized that selective PCR in neutrophils could allow the specific detection of actively replicating virions. Blood samples routinely submitted for CMV culture were shown to be acceptable for CMV PCR. Of 25 patients studied (HIV seropositive, CD4 100/$\mu$l), 9 had CMV DNA in their circulating neutrophils, while only one had a positive CMV blood culture. Six out of nine (including the culture positive individual) had CMV retinitis, while other 3 had a clinical illness compatible with CMV. None of the 16 patients with negative neutrophil CMV PCR results had any evidence of clinical disease attributable to CMV. Follow-up testing of 6 patients with positive baseline PCR result and treated appropriately for CMV showed negative PCR results at 1-3 months. This PCR test may be useful in the diagnosis and follow-up of HIV-infected individuals at risk of CMV disease. (Abstract shortened by UMI.)

Recombinant vaccinia viruses VF and VHN were shown to express glycoproteins with molecular weights similar to authentic HPIV3 F and HN proteins that were transported to the cell surface and recognized by monoclonal antibodies specific for HPIV3 F and HN. The HN glycoprotein in VHN-infected cells exhibited both hemagglutinin and neuraminidase activities. Both cleaved and uncleaved forms of the F glycoprotein were immunizers from VF-infected cell lysates. Fusogenic activity of the F protein was demonstrated only upon coinfection of HeLa T4$\sp+$ cells with VF + VHN and resulted in syncytium formation and hemolysis. Recombinant adenoviruses AdF and AdHN, expressed the HPIV3 F and HN proteins on the surface of infected HeLa T4$\sp+$ cells, respectively. AdHN-infected cells produced HN protein that was glycosylated and exhibited both hemagglutinin and neuraminidase activities. AdF infection led to the synthesis of glycosylated F$\sb0$ that was proteolytically cleaved. As was observed for vaccinia virus recombinants, syncytium formation was only observed upon coinfection of HeLa T4$\sp+$ cells with AdF + AdHN The F and HN proteins expressed by recombinant adenoviruses were recognized by monoclonal antibodies specific for the HPIV3 F or HN protein and were immunogenic in mice. Sera from AdF-or AdHN-inoculated mice immunoprecipitated appropriate glycoproteins from lysates of HPIV3-infected cells and neutralized HPIV3. Amino acid substitutions were introduced into three regions of the HPIV3 F protein to identify sequences involved in fusogenic activity. Substitution of uncharged amino acids for positively charged residues within the cytoplasmic tail, or introduction of a charged residue within the transmembrane domain, did not appear to affect fusogenic activity. Several different mutations designed to disrupt the leucine zipper or the heptad repeat motifs resulted in mutant F proteins that not only exhibited decreased fusogenic activity but concomitantly lost the ability to react with monoclonal antibody c128, which recognized an epitope in antigenic site B of the HPIV3 F glycoprotein. An association between the heptad repeat and leucine zipper domains, therefore, appears to be required for the HPIV3 F glycoprotein to adopt a conformation that is recognized by antibody c128 and that appears to be important for fusogenic activity.

The purpose of this study was to introduce a synthetic reoviral gene into the reovirus particle. The gene was introduced in the form of (+) sense ssRNA transcripts. Reovirus type I infected cells were transfected with a synthetic deleted version of the T3 M1 gene of reovirus. Alternatively virions and cores modified by various chemical and physical treatments were incubated with synthetic (+) sense ssRNA copies of the M1 gene in replication assays. Neither of these methods resulted in successful replication of a synthetic gene or its introduction into reovirus. Studies examining the fate of transcripts showed that synthetic gene or its introduction into reovirus. Synthetic RNA was not recognized by viral proteins or particles when transfected into reovirus infected cells. Efforts were made to select lethal M1 mutants by passage of T1, T3 and G2 reovirus in $\mu$2 protein (product of M1 gene) expressing cells. These mutants could be used later as a selection system for reoviruses which rescued a synthetic M1 gene. In this system only reovirus with the rescued synthetic M1 gene should be able to grow on normal L929 cells. However, lethal M1 mutants were not obtained. Reovirus particles that had been subjected to physical and chemical treatments were analyzed by RNase A digestion and electron microscopy and by transcription assay. Modified core particles maintained or increased transcription activity relative to native cores. The most significant finding was the demonstration of active soluble transcriptase derived from viral cores that, previously, were thought to be the smallest component of the virus exhibiting transcriptase activity. This transcriptase complex is composed of dsRNA and $\lambda$3 protein. The problem of restricted access of exogenous RNA to the transcriptase can be overcome using this form of transcriptase and may be applicable to introduction of a synthetic gene into reovirus. (Abstract shortened by UMI.)

Nucleotide composition bias and codon usage bias are commonly observed in a wide range of organisms, but the causes of both are still under debate. One view is that codon usage bias and base composition bias are both the result of a directional mutation pressure in favor of AT or GC; others argue that codon usage bias is a result of natural selection acting to mold the codon choice to match the frequency of the corresponding tRNAs. A number of recent studies have indicated that nucleotide bias is especially marked in gene families that are undergoing concerted evolution. In this thesis, I use the Drosophila trypsin gene family as a model system to study the causes and consequences of concerted evolution in Drosophila, and to investigate the evolutionary forces that may be responsible for the observed nucleotide bias. From each of the two related Drosophila species, D. melanogaster and D. erecta, a 12kb genomic region was sequenced and eight trypsin genes were identified. Some members of this gene family have been evolving independently while the others have been evolving in a concerted fashion. In both species, the nonsynonymous codon positions have a moderate GC content, while the synonymous sites are very GC-rich. For the genes that have been undergoing concerted evolution, due to rapid gene conversion, their synonymous G + C content is much higher than that of the independently evolving genes. In addition, these genes are characterized by an elevated frequency of pyrimidines (C or T) at synonymous sites on the coding strand. A combination of selective constraints, directional DNA mutation pressure, and DNA repair bias could have resulted in the base composition pattern observed in these trypsin genes.

Bacillus thuringiensis (Bt) produces a proteinaceous parasporal crystalline inclusion which is pathogenic to insect larvae. The insecticidal activities of crystals from different subspecies of Bt are highly specific for larvae within a particular insect order and in some cases for specific larvae within a family. Lepidopteran-active Bt crystals are composed of 130 kDa $\delta$-endotoxin (protoxin molecules, covalently linked by disulfide bridges). On ingestion by susceptible larvae, the $\delta$-endotoxin is converted to a 58-68 kDa toxin by the action of the high pH gut juice (pH $\sim$ 10-11) containing proteolytic enzymes and other proteins which associate with the toxin. One of the objectives of the research was to determine the role of enzymes/proteins in the spruce budworm (Choristoneura fumiferana) gut in determining the activity/specificity of a toxin. The gut juice of the spruce budworm larvae was found to contain a single trypsin-like serine protease (CFT-1) which was responsible for the activation of the $\delta$-endotoxin. CFT-1 was purified from gut juice by a combination of size exclusion and ion exchange chromatography. Some characteristics, such as molecular mass, N-terminal and active site sequences, charge and substrate specificity were similar to mammalian pancreatic trypsins. However, differences were observed in the burst kinetics and pH dependence of the catalysis where the apparent ionization constant of the active site histidine was over 1 pH unit higher than in the mammalian serine proteases. Another significant difference was that CFT-1 was stable for long periods of time at pH values as high as 11 whereas mammalian trypsins rapidly autolyse at pH values greater than 8. Inhibition of CFT-1 in gut juice prevented activation of the $\delta$-endotoxin. However, gut juice activation produced a slightly smaller toxin than the CFT-1 alone. It was observed during the digestion of $\delta$-endotoxins with gut juices that toxin was precipitated. This precipitate was shown to be caused by a protein in the gut juice. This protein has a molecular mass of 75 kDa and anionic character based on ion-exchange chromatography. A weak elastase-like activity was also shown to be associated with the precipitating protein. When the aggregate of toxin and gut juice protein was dissociated, it was observed that smaller toxin fragments were generated indicating that the toxin had been further proteolysed. The precipitation and degradation of toxin by gut juice may explain the observation that, when gut juice is used to activate the $\delta$-endotoxin, the in vitro recovery of toxin is much lower than expected. Therefore precipitation may play a role in determining the toxicity towards various insect larvae. Evidence had been obtained that DNA was associated with the $\delta$-endotoxin and toxin in vitro after the usual purification procedures and this DNA appeared to play an important role in the activation of the $\delta$-endotoxin. A study was undertaken to determine whether DNA found in purified Bt crystal preparations was also associated with crystals in vivo. The results obtained from photomicrography of Bt during its stages of development showed that DNA condenses in the region where the parasporal crystal forms. These results provided further evidence that the DNA found to be associated with purified crystals in vitro was not artifactual.

Nuclei from groups of twelve pooled left ventricles of rats were isolated and DNA, RNA and protein contents of the nuclear suspension were determined. RNA polymerase activity was assayed in the presence of either Mg++ or Mn++ and (NH4)2SO4. Electronmicroscopic examination of normal nuclear preparations revealed undamaged nuclei with very little contamination. Cardiac enlargement was produced by constricting the abdominal aorta in experimental animals. The operated animals were sacrificed at different intervals after operation between 4 hours and two weeks The "pure" nuclear suspensions of the myocardium of these animals was isolated for examination. Sham operated animals were used as control up till the third day, after this time normal animal served as control, since the difference between sham operated and normal animal did not persist beyond 3 days. An increase in RNA, DNA and protein content of the nuclear preparation was observed in aortic constricted animals. DNA showed a delayed increase compared to the time course of changes in RNA and protein. RNA polymerase activity showed a biphasic increase following constriction in both Mg ++ and Mn++-(NH4)2SO4 activated systems. In rats killed soon after constriction (4 hours in this study) the Mn++-(NH4)2SO4 system showed the greater percentage increase and on the 3rd-4th day after aortic constriction the increase was considerably greater in the Mg ++ activated system. An increase in RNA polymerase activity as early as 4 hours after aortic constriction was demonstrated in this study, among other findings. This early increase suggests that the stimulation of RNA synthesis - one of the first biochemical events in hypertrophy may play a role in translating the mechanical stress caused by increased outflow resistance into tissue growth.

This study involved the in vivo phenotypic characterization of the Drosophila melanogaster gene tango (tgo) through classification of both its zygotic-effect during CNS development and maternal-effect during early embryonic patterning. Particularly, this study involved characterization of the prd-repeat domain of the Tango protein, or its allelic equivalent, tgo 3. tgo has been previously isolated as a bHLH-PAS nuclear transcription factor required in the regulation of CNS development through its dimerization of the CNS master transcriptional regulatory gene, single-minded. Having performed an inter-allelic analysis, I grouped the CNS zygotic effects of tgo into various phenotypic classifications. The phenotypes exhibited a pleiotropy of CNS mutant effects and were classified as fused commissure, neurogenic, stalled or ambiguous nervous system defects. These phenotypes may possibly be attributed to different but partially overlapping functions of the various Tgo domains. Additionally, I analyzed the maternal-effect of tgo during precellular blastoderm development. Germline clonal analysis of the tgo3 allele revealed that tgo has an important role during early embryogenesis. Moreover, maternal tgo3 phenotypes resemble those of the more well-characterized maternal coordinate and gap segmentation genes. tgo3 germline clones were embryonic lethal and were classified as severe or intermediate, depending on the extent of deleterious effects on the resultant cuticular phenotype. Germline clones that showed a deletion of the entire anterior end, abdominal segments Al to A5 and missing the Filzkorper and spiracular opening in the posterior end were classified as severe and are comparable to the cuticular phenotypes of amorphic bicoid mutants and Kruppel ( Kr) phenotypes. Less severe deleterious effects, such as improper development of various anterior and posterior end structures or aberrant denticle band formation, were classified as intermediate tgo 3 germline clones, comparable to weaker bicoid alleles. Subsequent stainings to observe the effect on embryonic patterning through gap and pair-rule protein localization revealed alterations in spatial expression patterns. Alterations in both Kruppel and Even-skipped localization patterns in embryos mutant for maternal tgo3 suggest a role for tgo during early embryonic patterning, perhaps through the activity of its prd-repeat domain. Ectopic expression studies, employing the GAL4-UAS system, involved heat shock trials eliciting the ubiquitous expression of a truncated C-terminal Tgo protein (UAS-tgoDeltaC ) and both the ubiquitous land targeted expression of UAS- tgoDeltac. This targeted misexpression analysis revealed tgo as both a possible activator and repressor of Kr gap gene expression, in addition to affecting the protein distribution pattern of the later-acting segmentation gene, engrailed. Taken together, these germline clonal and ectopic expression studies suggest that the role of Tgo during early embryonic patterning may function through combinatorial interactions with maternal-effect and/or early-acting segmentation gene products. Moreover, this role may depend on the function(s) of the C-terminal sequences encoded by Tgo, including the prd repeat domain. Ultimately, this early embryonic role of tgo seems to be distinct from its role as a nuclear transcription factor.

Plant mitochondrial genomes are known to undergo frequent DNA rearrangements during evolution. Such DNA rearrangements have been observed within the extreme 5' or 3' termini of plant mitochondrial genes. The present studies of the gene for NADH dehydrogenase subunit I (nad1) in wheat demonstrate that DNA rearrangements can even occur at internal sites. This work shows that the nad1 gene is comprised of five single-copy exons in wheat mitochondria. The predicted NAD1 amino acid sequence is closely related to that from non-plant and chloroplast counterparts. Somewhat surprisingly, these nad1 sequences are scattered at four distantly-separated sites in the wheat mitochondrial genome. The analysis of regions flanking nad1 coding segments revealed the presence of sequence motifs and helical structures that are hallmarks of group II introns. A model is proposed for nad1 gene expression in which one cis- and three trans-splicing events are necessary for the production of nad1 mRNAs. To investigate nad1 gene expression at the RNA level, transcripts arising from the four nad1 coding regions were analyzed. Northern blot hybridizations and cDNA sequence analysis show that stable transcripts contain all five correctly-linked nad1 exons. S1 nuclease analysis in the regions flanking the nad1 coding segments also revealed the presence of stable transcripts which would be large enough to contain group II intron structures of normal sizes (<3 kb). As has been observed for virtually all plant mitochondrial mRNAs, the maturation of nad1 transcripts also involves RNA editing events. Interestingly, one of these RNA editing sites converts an ACG codon to AUG to create an initiation codon, and this suggests that RNA editing at this site is obligatory for translation of nad1 mRNAs to proceed. RNA editing is also observed within the discontinuous nad1 a/b intron in wheat (which is of particular interest because it lacks several key features of group II introns) but any contribution to intron structure or splicing is as yet uncertain. The wheat nad1 gene organization seems to have resulted from DNA rearrangements that occurred within previously continuous introns. This organization of the nad1 gene in wheat, a monocot, differs from that of dicot nad1 genes and this suggests that some DNA rearrangements have occurred relatively recently during plant evolution. The unusual nad1 gene structure shows that DNA rearrangements can alter not only gene order but also gene structure, provided that scattered gene pieces are properly transcribed and spliced.

Microspectrophotometric measurements of nuclear DNA of various species of oats revealed substantial differences among the diploid species. The karyotypes had been divided into two distinct groups, A and C, on the basis of cytogenetic studies, and this was confirmed by the microspectrophotometric data. The tetraploid Avena barbata, with the genome AABB, has a lower DNA content than does autotetraploid A. strigosa (AAAA). This suggests that tetraploid oats originated via segmental allotetraploidy rather than via autotetraploidy. The DNA content of the B genome was about 20% lower than that of the A genome. This may serve as a marker in the search for the potential donor of the B genome. Autoradiographic studies of the average durations of mitotic phases in the root tip cells of three oats species with different amounts of nuclear DNA showed that the DNA synthetic phase (S) of diploid and autotetraploid strigosa were similar, suggesting that homologous genomes in Avena species replicate in synchrony. It is proposed that the origin of certain genomes in polyploid oats might be resolved by comparing the replication patterns of chromosomes between the diploid and the polyploid species.