Brachydactyly type A1 (BDA1) is a rare, autosomal-dominant genetic trait that affects normal limb and bone development, resulting in short or absent middle halan es. Despite being the first human trait described in terms of autosomal dominant Mendelian inheritance, the underlying molecular cause had not been examined until relatively recently. In 2001, mutations of the Indian Hedgehog (IHH) gene were identified in three Chinese families with BDA1. Preliminary studies in our laboratory further demonstrated linkage of BDA1 to chromosome 5p in a Canadian kindred with a mild variant of the trait. The goal of this thesis was to characterize the BDA1 locus on chromosome 5 by reducing the critical region and examining the transcribed portions of candidate genes for mutations. To reduce the chromosome 5p critical region in the Canadian kindred, 23 microsatellite markers were examined and the number of family members analyzed was increased to include 22 affected and 16 unaffected individuals. By doing so, we defined a 7 cM (3.03 Mb) critical region flanked by the recombinant markers D5S1986 and D5S426. Fifteen candidate genes within the region were examined for mutations, however none were identified. Further characterization of the chromosome 5p critical region was made by recruiting twelve additional families into the study. Of the twelve families, five had a G → A transition at codon 100 of the IHH gene. The mutation causes an aspartate to asparagine amino acid substitution and supports a role for IHH in normal bone development. Linkage analysis to chromosome 5 in two other pedigrees indicated that the trait was not linked to this locus in these families, and provided further support for genetic heterogeneity underlying the aetiology of BDA1. IHH mutations were not found in the remaining five families, thus increasing the likelihood that the chromosome 5 locus is involved in these pedigrees. Linkage could not be established in these families, however, since the trait was sporadic, and not familial. Identification of additional BDA1 families in which the trait is linked to chromosome 5, and further mutational analysis will be required to identify the gene from chromosome 5p that causes BDA1.

The soybean (Glycine max L. Merrill) root nodule is an organ specialized for nitrogen fixation resulting from a symbiosis between the bacterium Bradyrhizobium japonicum and the roots. Root nodule senescence, its last stage of development, was the focus of this study. The progression of soybean root nodule senescence was examined using nitrogenase activity. During nodule development, nitrogenase activity was first observed with the appearance of nodules and peaked just prior to flowering, which marks the onset of nodule senescence. Accordingly, nitrogenase activity declined after flowering. Although nitrate and dark treatment both caused significant declines in nitrogenase activity, darkness caused a greater decline in nitrogenase activity than nitrate. The SAN1A gene, originally isolated from senescent root nodules (Chan, 1995), and two other members of the SAN1 gene family, were characterized. These genes represent a novel family of highly conserved, tandemly repeated genes in soybean. SAN1A and SAN1B encode predicted proteins of 352 and 353 amino acids, respectively, whereas SAN1C encodes a truncated protein of 126 amino acids and is likely a pseudogene. Despite their occurrence in tandem, gene conversion does not appear to have occurred within the evolutionary history of these genes. During natural senescence, SAN1A and SAN1B expression patterns suggested that SAN1A may play a role throughout nodule development, whereas SAN1B is likely to be more important during senescence. Both nitrate and dark treatment caused a decrease in SAN1A expression and an increase in SAN1B expression. Although both SAN1A and SAN1B were expressed in all soybean tissues examined, their patterns of expression differed. These results suggest that SAN1A and SAN1B have divergent functions or that SAN1B may replace SAN1A under conditions such as senescence. Preliminary results also suggest that SAN1A may be light regulated. Finally, a test of SAN1C expression in soybean tissues showed that it is expressed, despite the fact that it encodes a truncated protein. Database searches suggested that SAN1 genes are related to plant dioxygenases. A gene tree and BLASTP searches revealed that the SAN1 gene products are most closely related to uncharacterized dioxygenases identified in Arabidopsis, rice, and maritime pine, with the closest, identifiable dioxygenases being gibberellin 20-oxidases. Thus, the SAN1 genes may encode paralogues of known dioxygenases or novel dioxygenases. The evidence collected to date suggests that the SAN1 genes encode dioxygenases with roles in various aspects of plant metabolism. Further characterization of this family may yield more clues about their role(s).

Myoclonus-dystonia (M-D; OMIM 159900) is an autosomal dominant movement disorder characterized by alcohol responsive myoclonic jerks typically affecting the arms and axial muscles. Mutations in the epsilon-sarcoglycan (SGCE) gene on chromosome 7q21 were shown to cause M-D but not in all cases. Currently, in the literature mutations in SGCE are responsible for 20--50% of all cases of M-D. By direct sequencing, we have identified mutations in four families and we were able to find large deletions in two additional families using a semi-quantitative PCR-based DHPLC method. One of these families (family 2) has an 18 kb deletion that includes exons 2 and 3 while another family (family 6) has a 13 kb deletion covering exons 2 through 5. We have confirmed these large deletions by identifying the exact deletion breakpoints. In our cohort of 17 unrelated M-D families, 35% of patients with M-D have mutations in the SGCE gene. Previously, we characterized a large five-generation Canadian family with M-D (family 1) that did not have a mutation in SGCE and mapped a novel locus for this disorder to a 4 Mb region between the markers D18S1132 and D18S843 on 18p11 (OMIM number: 607488). Since additional informative STS markers were not found between the flanking recombinant and non-recombinant markers, we utilized single nucleotide polymorphisms (SNP) for fine-mapping. As a result, we identified three recombinant SNPs (rs385769, rs727951 and rs727952) in the proximal flanking region between markers D18S1163 and D18S843 in family 1. Our previous 4 Mb critical region has been reduced to 3.18 Mb, flanked by markers D18S1132 and rs385769. There are seven known genes and eight predicted genes within or very close to this region. We have sequenced all the exons and exon-intron boundaries of these seven known genes ( MGC17515, ZFP161, EPB41L3 (KIAA0987), L3MBT4, ARHGAP28 (FLJ10312), LAMA1, and PTPRM) and four predicted genes (LOC388459, LOC388460, LOC400643 and LOC388461) without identifying the causative mutations. Future studies should be directed to the investigation of the functional effects of the known SGCE mutations using in vitro and in vivo models. With respect to the identification of a novel gene in family 1, we will focus on the novel transcripts in the critical region based on gene prediction and EST (expressed sequence tag) mapping. Once the second M-D causing gene is identified, we will perform gene expression and cellular localization analyses of the mutated and wild type protein in tissues and generate a mouse model.

Cancer can be considered as a class of genetic diseases which arises as a result of mutations in oncogenes and tumour suppressor genes. These mutations alter the function, morphology, and behaviour of the affected cells, ultimately leading to tumour initiation, progression, and metastasis. These phenotypic changes also impact upon the ability of tumour cells to either respond to chemotherapy with cell death, or evade chemotherapeutic induced cell death through drug-resistant pathways. To identify cancer-related genes which contribute to cell survival during chemotherapy, hybrids between normal human fibroblasts and murine melanoma cells (B78) were generated and screened for the ability to survive short term exposure to the chemotherapeutic agent, PALA (N(phosphonoacetyl)- scL-aspartate). Whereas PALA induced growth arrest in normal fibroblasts, B78 responded to PALA with apoptosis (active cell death). It was hypothesized that the introduction of normal human chromosomes from the fibroblasts into B78 would result in an alteration in some of the hybrids' responses to the drug; in this instance, survival via growth arrest as opposed to apoptosis. Those human chromosomes which were associated with improved PALA survival were theorized to express genes responsible for cellular survival during chemotherapeutic treatment. The hypothesis was first tested using whole cell hybrids constructed between human fibroblasts and B78. Whole cell hybrids showed improved survival during PALA exposure, as opposed to B78, and showed morphological features of growth arrest. To identify individual human chromosomes involved in this response, a panel of dual selectable microcell hybrids was constructed and screened. These hybrids contained one or more normal human chromosomes in a B78 background. A drug selectable marker permitted the selective retention, or removal of the tagged human chromosome, as desired. The property of double selection was required to show that phenotypic changes were always due to the presence or removal of the tagged chromosome, rather than due to clonal variation. A pharmacological screen (using PALA) of the microcell hybrid panel resulted in the identification of human chromosome 3, which was consistently associated with improved survival and growth arrest during drug exposure. To attempt to identify genes which were involved in drug survival, a molecular strategy, known as differential display reverse transcriptase PCR (DDRT-PCR) was employed. This strategy was used to identify those cDNAs derived from differentially expressed mRNAs during PALA exposure in hybrids that showed improved survival during PALA exposure, as compared to B78 and the hybrids that died readily in the presence of PALA. Several candidate cDNAs were identified, cloned and sequenced. Characterization of these cDNA's may ultimately lead to the identity of new survival genes and an improved understanding of cancer cell drug responses.

This thesis tested the hypothesis that the synergistic effect of cisplatin in combination with radiation was due to cisplatin inhibition of repair of radiation-induced DNA damage. Two model cell systems were used. DNA repair inhibition caused by cisplatin and the fundamental mechanism of DNA double strand break repair were studied in human fibroblast cells. The effect of cisplatin on stress response and the induction of radiation resistance was studied in yeast. It was shown that normal human fibroblasts cells (AG1522), irradiated in the presence or absence of cisplatin adducts have biphasic DNA repair curves with a fast and slow component. When cells were treated with cisplatin immediately before irradiation, fast repair was not affected but slow repair was inhibited. The rate of the slow repair in the presence of cisplatin adducts was ten fold less than the rate in the absence of cisplatin adducts. Cisplatin treatment 24 hours prior to radiation resulted in inhibition of both fast and slow repair. Yeast experiments tested whether defects in DNA repair mechanisms would amplify the radiosensitizing effect of cisplatin. It was shown that cisplatin sensitized cells with a competent recombinational repair mechanism, but did not sensitize cells defective in recombinational repair. It was also shown that under certain circumstances cisplatin did not radiosensitize normal, repair competent cells but induced stress response and cellular radiation resistance. Repair proficient wild type yeast cells became thermal tolerant and radiation resistant two hours after a sublethal cisplatin treatment. Cisplatin pretreatment also suppressed mutations caused by N-methyl-N$\sp\prime$-nitro-N-nitrosoguanidine (MNNG) treatment, a response previously shown in wild type cells with radiation pretreatment. Like radiation, cisplatin-induced stress response did not confer radiation resistance or suppress MNNG mutations in a recombinational repair deficient mutant (rad52). These results support the idea that cisplatin crosslinks in DNA can induce a stress response which subsequently confers radiation resistance, thermal tolerance, and mutation resistance in yeast. The mechanism of this cisplatin-induced resistance was determined to be dependent on an error free recombinational repair pathway. The results obtained in this thesis indicated that error free recombinational repair plays a major role in the interaction between cisplatin and radiation. To further examine error-free recombinational repair in mammalian cells, confocal microscopy in conjunction with fluorescence in situ hybridization was used to show that the distance between homologous chromosome domains was reduced after radiation. The mechanism for homologous chromosome rearrangement following exposure to radiation is not known, however, it is plausible that a recombinational repair process is responsible. Overall, this work has shown that the radiosensitizing effect of cisplatin was due to inhibition of DNA repair processes involving recombinational repair. However, it was also shown that, under specific conditions, cisplatin treatment induced a stress response that conferred cellular radiation resistance. Therefore, the effect of cisplatin in combination with radiation was variable depending upon the cellular processes influenced by cisplatin.

A mouse model system was developed to sensitively detect mutagenic events arising spontaneously in mouse fibrosarcoma cells grown under in vitro or in vivo conditions. These cells were genetically manipulated to enhance their sensitivity as detectors of mutational events at the X-linked hprt (hypoxantine phosphoribosyltransferase) gene, a frequently used marker of genomic instability. A fibrosarcoma cell with three X-chromosomes and randomly distributed X-specific fragments was isolated from a male mouse. Initially, this cell line was resistant to mutation induction, presumably due to the presence of multiple hprt genes. Inactivation of all but one hprt gene resulted in a cell line with a heterozygous marker which was at least 1000 times more sensitive to detecting induced mutational events. Infection of this derived cell line with a selectable marker for neomycin resistance allowed recovery of cells which subsequently demonstrated that the tumour environment was capable of enhancing spontaneous in vitro mutation induction. A human lymphoblastoid cell line was also genetically manipulated to establish nine stable transfectants as models for the study of genomic instability. The endogenous X-linked hprt gene was mutated, presumably by a large deletion, and subsequently transfected with a human hprt minigene encoding for its own promoter and poly-adenosine signal. Transfectants were assessed for the stability of the transgene. Transfected clones demonstrated a 2 to 600 fold increase in spontaneous mutation rate at the transgene loci compared to the endogenous hprt locus. There was no evidence of mutator phenotype activation in any of the clones as judged by testing at another locus, the endogenous tk (thymidine kinase) gene. An inverse correlation was also noted between the number of the inserted sequences and the stability of gene expression, with single gene insertions being most stable and tandem insertions being the least stable.

The childhood spinal muscular atrophies (SMA) are neurodegenerative disorders characterised by progressive spinal cord motor neuron depletion and are among the most common autosomal recessive disorders. Type I SMA is the most frequent monogenic cause of death in infancy. The loss of motor neurons in SMA, has led to suggestions that an inappropriate continuation or reactivation of normally occurring motor neuron apoptosis may underlie the disorder. The gene resonsible for the SMA's was mapped to a region of 5q13 Banked proximally by the marker CMS-1 and distally by the marker D5S557. In keeping with the hypothesis that a mechanism of apoptosis may underlie the disorder, we have isolated a gene encoding the neuronal apoptosis inhibitor protein (NAIP), which is homologous to baculoviral inhibitor of apoptosis proteins (IAP) and is partially deleted in individuals with Type I SMA. Concurrently, a second candidate gene encoding survival motor neuron (SMN), which is contiguous with the NAIP locus on 5q13.1 was also reported (Levebre et al., 1995. Cell 80, 155). SMN is deleted in a significant majority of SMA individuals, leaving unclear the precise role of the two genes in SMA causation. In an effort to delineate the role of NAIP in SMA pathogenesis, we have studied the effect of NAIP on cell death induced by different apoptotic triggers and determined the cellular distribution of the protein in human spinal cord. We report that overexpression of NAIP in Rat-1, HeLa and CHO cells suppresses apoptosis induced by menadione, tumor necrosis factor alpha (TNF-$\alpha$) and serum withdrawal. Immunocytochemistry employing polyclonal antiserum raised against human NAIP demonstrates immunoreactivity in motor neurons. NAIP mediated inhibition of cell death and the immunolocalization of the protein to motor neurons are consistent with a role for NAIP both in the naturally occurring programmed motor neuron death. and when defective, in the pathogenesis of SMA. Moreover, NAIP appears to be the first member of a novel family of human genes with anti-apoptotic activity. Three novel human IAPs have subsequently been identified and shown to suppress apoptosis (Liston et al. 1996. Nature 379, 349). Two of these were also identified based on their ability to associate with the tumor necrosis factor receptor II (TNFRII) via a TNF receptor associated factor 1 and 2 (TRAF1-TRAF2) heterocomplex (Rothe et al., 1995. Cell 83, 1243) These data provide evidence for a central role of NAIP and IAPs not only in apoptosis but as regulators in these signalling pathways. (Abstract shortened by UMI.)

No description available.

No description available.

No description available.