The phosphorylation and cell adhesion properties of myelin-associated glycoprotein (MAG) isoforms were investigated. MAG is a member of the immunoglobulin supergene family and is thought to mediate interactions between myelinating glial cells and neurons in the central and peripheral nervous systems. Two isoforms of MAG exist, L-MAG and S-MAG, which exhibit differential expression patterns. Using retroviral infection and transfection of the cDNAs encoding L-MAG and S-MAG, L cell fibroblasts and NIH3T3 cell lines that express either isoform were generated. The expression of both isoforms on the cell surface of L cells induced the cells to aggregate in a MAG-dependent fashion. The adhesion phenomenon was determined to be calcium and temperature independent. A critical level of cell surface MAG expression was required for cell aggregation to occur. The adhesion was found to be heterotypic in nature, signifying the presence of a MAG receptor on the cell surface of L cells. The MAG isoforms are identical in their extracellular and transmembrane domains. They share a common region in their cytoplasmic domain, but are distinct at their carboxyl termini. The unique carboxyl tails are comprised of 54 amino acids in L-MAG and 10 residues in S-MAG. Both isoforms were determined to be phosphorylated in fibroblasts. L-MAG exhibited phosphorylation on serine, threonine and tyrosine residues. The phosphorylation on tyrosine was augmented by treatment of cells with ammonium vanadate. S-MAG phosphorylation occurred mainly on serine residues, but some phosphotyrosine was also detected. The phosphorylation of S-MAG, however, was relatively insensitive to vanadate treatment. Tryptic digest analysis showed that two serine and the major tyrosine phosphorylation site in MAG were identical in L-MAG and S-MAG. The major tyrosine phosphorylation site in MAG was, thus, identified as tyrosine 558. This tyrosine residue is homologous to the major tyrosine phosphorylation site in the fibronectin receptor, integrin. A similar phosphorylation pattern of L-MAG was observed in primary rat oligodendrocytes. Determination of the stoichiometry of phosphorylation revealed that phosphorylation of L-MAG was at least one order of magnitude greater than S-MAG, especially with respect to tyrosine phosphorylation. This result indicates the presence of a carboxyl terminal sequence unique to L-MAG that activates the phosphorylation of tyrosine 558. This region may represent docking sites for protein tyrosine kinase binding, making L-MAG-kinase interactions more efficient. Two populations of L-MAG molecules were discovered: those that are phosphorylated on tyrosine residues, and those that exhibit serine/threonine phosphorylation. In effect, tyrosine phosphorylation precludes serine/threonine phosphorylation, suggesting that the alternatively phosphorylated L-MAG molecules perform different functions. Increasing the tyrosine phosphorylation of MAG had no detectable effect on the cell adhesion behaviour of the cells. Tyrosine phosphorylation of L-MAG, however, induced its capacity to bind the SH2 domain of phospholipase C-gamma. Therefore, L-MAG has the potential to interact with a variety of signalling molecules via its cytoplasmic domain. While L-MAG may participate in signal transduction pathways, S-MAG may function in a more restricted manner and perform only as a cell adhesion molecule. Thus, the differential regulation of MAG isoform expression may serve to increase the repertoire of MAG functions at specific times during development.

Here we report on the attempt to isolate nodule-specific cDNA clones from a cDNA library that was derived from an actinorhizal species, Alnus incana, nodule poly A+ enriched RNA. mRNA corresponding to these cDNA clones could be detected in RNA isolated from the actinorhizal nodule but, could not be detected in RNA isolated from the root. The pattern of expression of four genes corresponding to the cDNA clones was investigated at various time points following infection. It appears the nodulin clones could be divided into two categories of expression; one showing a maximal expression prior to the onset of nitrogen fixation, and the other expressing a maximal level after the onset of nitrogen fixation. Sequence characterization of these clones eventually revealed that the cDNA clones were ribosomal in nature of both plant and bacterial origin. However, further analysis revealed that one cDNA clone of bacterial origin had a unique 105 bp region. PCR analysis suggests that it may be possible to survey various bacterial species for the presence or absence of this insert in order to classify and investigate the evolution of 23S rRNA within this region of the gene. (Abstract shortened by UMI.)

We have isolated both mitochondrial and total genomic DNA from Larix species to examine the physical and genetic complexity of mitochondrial DNA and to determine mitochondrial inheritance patterns in Larix hybrid crosses, using restriction endonuclease and Southern hybridization analyses with specific wheat mitochondrial gene probes. In addition, these techniques have been used to identify restriction fragment length polymorphisms in somatic embryogenic cell cultures and corresponding regenerated hybrid Larix trees. Evaluation of mitochondrial gene hybridization patterns of two Larix species and their reciprocal hybrids demonstrated a maternal mode of inheritance of mitochondrial DNA gene sequences. Conifers appear to have evolved mechanisms for organelle exclusion during fertilization, resulting in a predominantly uniparental mode of inheritance of their organelles (Owens and Morris, 1990). Maternal inheritance of the mitochondrial genome together with paternal inheritance of the chloroplast genome in Larix hybrid crosses will be useful in the determination of maternal and paternal contributions in hybrid and introgressed Larix populations, and will increase the precision in determining phylogenies of this species. (Abstract shortened by UMI.)

Molecular characterization of the Seoul 80-39 virus, a prototype of the Seoul serotype, was carried out by cloning and sequencing. The large (L) genomic RNA segment is 6530 nucleotides long. The viral complementary-sense RNA contains a single open reading frame (ORF) from the AUG codon at nucleotide positions 37-39 to the UAA stop codon at nucleotide positions 6490-6492. This ORF could encode a polypeptide of 2151 amino acids (246,662 D) which likely corresponds to the L protein detected in purified viral particles (Elliot et al., 1984) and is assumed to be an RNA-dependent RNA polymerase molecule (Schmaljohn and Dalrymple, 1983). The M RNA segment is 3651 nucleotides long. A single ORF was detected in the viral complementary-sense RNA. This ORF could encode a polypeptide of 1133 amino acids which is cotranslationally processed into viral glycoproteins G1 and G2. The S genomic RNA segment was characterized from the 3$\sp\prime$ end to nucleotide 1332. One long ORF was detected from the AUG codon at nucleotide positions 43-45 to the UAA stop codon at nucleotide positions 1330-1332. This ORF has a capacity to encode a 48 kD protein which corresponds to the viral nucleocapsid (N) protein. A comparison of coding properties of the tripartite genomes of various hantaviruses allowed construction of the evolutionary tree for the Hantavirus genus. The second part of the study was focused on the maturation of Hantaan 76-118 virus glycoproteins G1 and G2 virus morphogenesis. Glycoproteins G1 and G2 form a complex. Experiments with the glycosylation inhibitor brefeldin A indicate that complex formation between the two glycoproteins occurs in the ER. Resistance of both G1 and G2 to endoglycosidase D suggests a structure with more than five mannose residues. Processing of a high mannose structures ((Man)$\sb8$(GlcNAc)$\sb2$ to (Man)$\sb7$(GlcNAc)$\sb2$ or (Man)$\sb6$(GlcNAc)$\sb2$) may take place in the cis-Golgi by the enzyme $\alpha$-mannosidase I. Virus budding was observed only in Hantaan virus-infected cells, in the trans-Golgi network. When the Golgi apparatus was destroyed by brefeldin A treatment, assembly of virus particles was not observed. The third project was to express nucleocapsid proteins of Seoul 80-39 virus and Hantaan 76-118 virus using baculovirus recombinants. Analysis of recombinant virus-infected Sf9 cell lysates by SDS-PAGE showed that N proteins were expressed in high levels and were identical to virion associated N proteins in size and antigenicity. (Abstract shortened by UMI.)

I have examined the expression of calcium channels in P19 cells to determine how their expression is altered during differentiation into cardiac muscle cells. Colonies of beating muscle cells formed from DMSO-induced differentiation of P19 cells were treated with low concentrations of specific L-type calcium channel blockers and all contractions ceased, implying the existence of functional L-type calcium channels in DMSO-differentiated P19 cells. The expression of the various subunits of the L-type calcium channel was examined by PCR (Polymerase Chain Reaction) and Northern blot analysis. The adult skeletal muscle isoform of the $\alpha$1 subunit was undetectable by PCR; however, Northern blot analysis using the skeletal muscle $\alpha$1 subunit probe revealed the presence of two possible isoforms in P19 cells, with mRNA sizes of 6.5 kb and 13 kb. The cardiac isoform of the $\alpha$1 subunit was found to be expressed in DMSO-treated cells, and the appearance of the cardiac $\alpha$1 subunit mRNA on Northern blots corresponded with the on-set of contractile activity. The $\alpha$1 subunit transcript observed was found to correspond to a 9 kb message which was similar to that identified in adult cardiac muscle. The expression of the other subunits of the L-type calcium channel was also investigated. The skeletal forms of the $\beta$ and gamma subunits were undetectable in P19 cells, whereas the $\alpha$2 subunit was present in DMSO-treated P19 cells as determined by PCR but was undetectable by Northern blot analysis. Northern blot analysis using the $\beta$ subunit probe indicated the presence of a weakly related transcript in P19 cells. These results indicate that unique subunits of the L-type calcium channel may be expressed in P19 cells. (Abstract shortened by UMI.)

The research described in this thesis focuses on glucose repression of gene expression in a higher eukaryotic organism, the fruit fly Drosophila melanogaster. The results show that DNA sequences which are located upstream of the transcribed region of the amylase gene (Amy) control glucose repression in D. melanogaster larvae. Hybrid constructs, in which the upstream flanking sequences of the amylase gene were fused with the transcribed region of the alcohol dehydrogenase gene (Adh), were expressed in transgenic $Adh\sp{\rm null}$ larvae. The expression of ADH from the hybrid gene was found to be subject to glucose repression. The function of potential regulatory cis-acting elements within the glucose responsive upstream region of the amylase gene was examined by deletion analysis and site-directed mutagenesis, coupled with expression assays in somatically transformed $Amy\sp{\rm null}$ larvae. The upstream deletion analysis showed that essential elements, both for overall activity and glucose repression of the amylase gene, are located close to the transcription start site (within 109 base pairs). In vitro site-directed mutagenesis of upstream sequences revealed that the TATA motif, at position $-$28, and a novel 35 base pair element, at position $-$109 to $-$74 with respect to the transcription start site, were necessary for overall activity of the amylase promoter. None of the introduced mutations, which scanned the upstream regulatory region of the amylase gene, resulted in the loss of glucose responsiveness. These results suggest that glucose repression, in Drosophila, is mediated by transcriptional mechanisms which involve multiple functionally redundant DNA elements. Variation in the degree of glucose repression of the amylase genes was examined at a molecular level in D. melanogaster. Gene-specific elements located within a short upstream DNA region were shown to regulate the difference in response to dietary glucose between duplicated amylase genes. In addition, in order to assess the effect of interspecific variation of regulatory DNA sequences on glucose repression, the function of glucose repressible amylase promoters was examined in distantly related species. The D. virilis amylase gene was shown to be repressed by dietary glucose in somatically transformed D. melanogaster larvae. (Abstract shortened by UMI.)

The role of Ca$\sp{2+}$ in catecholamine actions on hepatocyte metabolism was studied in three fish species: American eel, brown bullhead and rainbow trout. The cellular Ca content, Ca$\sp{2+}$ fluxes across the cell membrane and cytosolic free-Ca$\sp{2+}$ concentration ((Ca$\sp2\rbrack\sb{\rm i}$) changes in Fura-2-loaded single hepatocytes were measured respectively using atomic absorption spectrophotometer, a $\sp{45}$Ca exchange technique and a computer-controlled microspectrofluorimeter technique. The results indicated the following about Ca$\sp{2+}$ and hepatocyte metabolism. First, fish hepatocytes contain higher Ca content than equivalent mammalian cells (8.25 $\pm$ 1.03 (eel) and 10.49 $\pm$ 1.26 (bullhead) $\mu$moles$\cdot$g$\sp{-1}$ wet wt., respectively). Second, Ca$\sp{2+}$ uptake is a passive or energy-independent process whereas Ca$\sp{2+}$ efflux may be an active or energy-dependent process. Third, Ca$\sp{2+}$ uptake was not significantly stimulated by the catecholamines studied. Fourth, Ca$\sp{2+}$ efflux was significantly stimulated by both epinephrine and phenylephrine in eel hepatocytes and these effects were blocked by the $\alpha$-antagonist phentolamine. The $\alpha$-agonist isoproterenol and $\beta$-antagonist propranolol did not affect basal or hormone-stimulated Ca$\sp{2+}$ efflux. Fifth, basal (Ca$\sp{2+}\rbrack\sb{\rm i}$ was similar in eel and trout hepatocytes, but significantly higher in bullhead cells (184 $\pm$ 23 nM). Sixth, (Ca$\sp{2+}\rbrack\sb{\rm i}$ was significantly increased in eel hepatocytes by either epinephrine or phenylephrine at 10$\sp{-7}$M, but not the $\beta$-agonist isoproterenol (4% increase). Epinephrine and phenylephrine were found to induce repeated $\rm Ca\sb{i}\sp{2+}$ transients (i.e., Ca$\sp{2+}$ oscillations) with variable patterns in individual eel hepatocytes. Trout hepatocytes exhibited little sensitivity to epinephrine with less than 20% of the cells tested showing changes in (Ca$\sp{2+}\rbrack\sb{\rm i}$ in response to catecholamines. Seventh, the initial rise in Ca$\sb{\rm i}\sp{2+}$ induced in eel hepatocytes by epinephrine was independent of external Ca$\sp{2+}$, although external Ca$\sp{2+}$ is required for the long-term maintenance of Ca$\sp{2+}$ oscillations. Bullhead cells depend totally on external Ca$\sp{2+}$ for increasing Ca$\sb{\rm i}\sp{2+}$. (Abstract shortened by UMI.)

The haemagglutinin-neuraminidase protein (HN) of human parainfluenza virus type 3 is a bifunctional protein. To map the functions to the protein, three truncations of the gene were constructed by digesting the HN gene with the restriction endonucleases Hind III, Bgl II or Xbo I and inserting a stop codon. An internal deletion using Rsa I was also constructed. The four mutants were expressed in the recombinant vaccinia virus system. The expression of the mutant proteins was analysed by both Western Blot and immunoprecipitation. The products of the three truncations migrated at the molecular weights predicted for the truncated proteins. The product of the internal deletion migrated more quickly than predicted and upon sequencing revealed a frame shift mutation and, therefore, a fourth truncation. The migration of the gene product was consistent with the molecular weight predicted for this truncation. The full length HN was functional in both haemagglutination and neuraminidase assays. The four mutants were active only in the neuraminidase assay, none of them haemagglutinated Guinea pig erythrocytes. This allows us to predict that the haemagglutination region is near the carboxy-terminus of the protein while the neuraminidase region is amino-terminal of amino acid 212.

At present the cloning of disease genes relies on the application of forward and reverse genetics which requires prior knowledge of either the protein or the location of the gene locus underlying the disorder. In the present study a third alternative, a lateral approach, which focuses in at the level of expressed sequences (mRNA) is examined. This approach is used in the present study to explore its potential in the isolation of X-linked chorio-retinopathy candidate genes. The strategy developed for this approach uses a probe derived by the PCR amplification of a chorio-retinal cDNA library to screen a hamster/human X hybrid genomic library made in $\lambda$ phage vectors. The immediate achievement of this cross screen is the direct association of tissue expression with chromosome specificity. The initial application and evaluation of this approach is presented in the following thesis.

Low density lipoproteins (LDL), the major cholesterol carriers in human plasma contain a single copy of apolipoprotein B-100 (apoB-100), a 4536 residue, water-insoluble polypeptide as their exclusive protein component. ApoB-100-containing lipoproteins (LpB) are secreted from the liver as large triglyceride-rich lipoproteins and are converted in the circulation to smaller cholesteryl ester-rich LDL. As a ligand for the cell surface LDL receptor, apoB-100 has an important role in cholesterol metabolism and elevated apoB-100 levels increase the risk of atherosclerosis. Genetic polymorphism, oxidative modification, non-enzymatic glycation and the lipid environment of apoB-100 can affect its ability to mediate binding to the LDL receptor. In my Ph.D. program, I have produced and employed a panel of anti-apoB-100 monoclonal antibodies (mAbs) to study apoB structure and function. To this end, I developed and evaluated 3 novel immunization strategies to generate apoB-100 isoform-specific mAbs. Immunization of human apoB-100 transgenic mice proved to be an effective method for production of anti-apoB-100 mAbs having unusual specificities. This strategy could also be applied to other antigens. Several mAbs produced by this protocol are shown to be useful probes of apoB-100 conformation. I have characterized the binding specificities of a large panel of anti-apoB-100 mAbs and have analyzed the expression of their corresponding epitopes on LpB as a function of LpB intravascular metabolism. I show that the apoB-100 carboxy-terminus undergoes a major conformational change as large LDL are converted to smaller particles, a conformational change that could allow the particle to become competent to mediate binding to the LDL receptor. Finally, I demonstrate that, when LDL undergoes non-enzymatic glycation, at least 6 sites in apoB-100 are modified, including 2 that flank the apoB LDL receptor-binding site. As the LDL receptor-binding site itself is not modified, the loss of binding activity may be secondary to a change in apoB conformation.