Homeologous recombination refers to genetic exchanges between DNA partners containing similar but not identical DNA sequences. Heteroduplex intermediates in such exchanges are expected to contain multiple DNA mismatches at positions of sequence divergence and hence are potential targets for mismatch correction. Thus recombination of this type is of particular interest in understanding the role of DNA mismatch correction on recombination fidelity. Previous studies that examined this question in prokaryotic systems suggested that mismatch repair acts as a barrier to recombination between diverged sequences. The central hypothesis of this thesis is that mismatch correction acts as a barrier to homeologous recombination in yeast. The objectives of these studies was to elucidate the role of mismatch correction in homeologous recombination as a means of dissecting its mechanism in eukaryotic organisms. To examine homeologous genetic events in yeast, I developed an in vivo assay system to evaluate recombination between diverged DNA sequences. A homeologous gene pair consisting of Saccharomyces cerevisiae SPT15 and its Schizosaccharomyces pombe homolog were present as a direct repeat on chromosome V, with the exogenous S. pombe sequences inserted either upstream or downstream of the endogenous S. cerevisiae gene. Each gene carried a different inactivating mutation, rendering this starting strain Spt15-. Recombinants that regenerated SPT15function were identified by genetic selection and the rates of recombination in different backgrounds were compared. The homeologous mitotic recombination assay was utilized to test the role of S. cerevisiae mismatch repair genes PMS1, MSH2 and MSH3 on recombination fidelity. In strains wild type for mismatch repair, homeologous recombination was reduced 150-180 fold relative to homologous controls, indicating that multiply mispaired sequences act in cis as part of an inhibitory mechanism. In the upstream orientation of the homeologous gene pair, msh2 or msh3 mutations resulted in 17-fold and 9.6-fold elevations in recombination and the msh2 msh3 double mutant exhibited an 43-fold increase, implying that each MSH gene can function independently in trans to prevent homeologous recombination. Homologous recombination was not significantly affected by the msh mutations. In the other orientation, only msh2 strains were elevated (12-fold) for homeologous recombination. A mutation in MSH3 did not affect the rate of recombination in this orientation. Surprisingly, a pms1deletion mutant did not exhibit elevated homeologous recombination in either case. Next, I investigated whether mismatch correction acts as a specific or general obstacle to homeologous recombination by blocking one or many exchange pathways. To answer this question, I performed structural analysis on numerous recombinant products from each strain to determine the percentage of products that fell into a given class (crossovers or gene conversions). Each percentage was then multiplied by the overall rate to arrive at a rate of recombination for individual events. Typically 90-100% of homologous and homeologous recombinant products could be accounted for, either as crossovers or gene conversions. Recombination for all classes of products was inhibited when divergent sequences were present, indicating that homeology blocks formation of both crossovers and gene conversions. Sequence analysis of a limited number of homeologous recombinants indicated that transfer of DNA occurred in continuous stretches and that endpoints fell within regions of 3-11 base pairs of perfect homology. Mutations in the mismatch repair genes MSH2 or MSH3that elevate the overall rate of homeologous recombination produced similar rate increases in formation of each recombinant class. This suggests that mismatch correction proteins block multiple types of homeologous recombination events. Taken together, these results support the hypothesis that homeologous and homologous recombination occur by the same (or similar) pathways, with mismatch repair superimposed as an extra level of control over the fidelity of the process. I also investigated whether this homeologous recombination system would be useful in a genetic screen to identify novel genes or new alleles of genes known to increase exchanges between diverged DNA sequences. Hyperhomeologous recombination mutants were isolated following ethylmethane sulfonate mutagenesis of yeast that harbored the spt15 homeologous duplication. Preliminary characterization of these mutants demonstrated that some of these isolates yielded phenotypes that were consistent with mutations in mismatch repair genes verifying the utility of this method to identify such mutants. To improve the use of this system for a mutant screen, I developed a second generation homeologous duplication using URA3. These new starter strains are expected to be important for efficient isolation and characterization of hyperhomeologous recombination mutants.

Type 2 diabetes (T2D) is a metabolic disease that results from complex interactions between the environment, the genetic variation and epigenetic regulation of gene expression in individuals. Beta-cell dysfunction and insulin resistance are regarded as the hallmarks of the disease as the common presentation of T2D is the inability of beta-cells to adequately respond to the insulin demands of the body. The prevalence of T2D in Africa, and particularly South Africa, is on the rise. This is very likely the result of the combination of genetic susceptibility with increasing availability and accessibility of relatively cheap, highly palatable, calorie-dense meals with no corresponding lifestyle adjustment. This study aims to utilize available data from GWAS and gene expression arrays to identify potential variants that likely influence T2D susceptibility in African populations. Two public data repositories were mined – the National Center for Biotechnology Information’s (NCBI) Gene Expression Omnibus (GEO) and the National Human Genome Research Institute’s (NHGRI) GWAS Catalog. The criteria for selecting the studies for inclusion were based on ten descriptive T2D-related terms taken from the GWAS catalog’s pre-defined search categories. These terms were also applied to the selection of gene expression studies in GEO. These terms are: “fasting glucose-related traits”, “fasting insulin-related traits”, “fasting plasma glucose”, “insulin resistance/response”, “insulin traits”, “diabetes-related insulin traits”, “pro insulin levels” “Type 2 diabetes”, “type 2 diabetes and 6 quantitative traits” and “type 2 diabetes and other traits”. Ten Affymetrix platform-based studies in human tissues were chosen from GEO using these criteria. A Benjamin-Hochberg adjusted p-value of 0.05 was set as a cut-off for significant differentially expressed genes (7,887 genes) with 497 genes occurring in two or more studies, based on tissue- or array-type, considered candidates for downstream analysis. The GWAS catalogue presented 175 “reported” genes and 218 SNPs from 51 studies matching the set T2D-related criteria. Functional analyses done with the Database for Annotation, Visualization and Integrated Discovery (DAVID) on both the GWAS and expression studies genes lists,

The gene that produces the precursor RNA transcript to the three largest ribosomal RNA molecules (rDNA) is present in multiple copies and organized into gene clusters. They represent 0.5% of the diploid human genome but are critical for cellular viability. The individual genes possess very high levels of sequence identity and are present in high local concentration, making them ideal substrates for genomic rearrangement driven by dysregulated homologous recombination. Our laboratory has developed a sensitive physical assay capable of detecting recombination-mediated genomic restructuring in the rDNA by monitoring changes in lengths of the individual clusters. In order to determine whether dysregulated recombination is a potential driving force of genomic instability in human cancer, adult patients with either lung or colorectal cancer, and pediatric patients with leukemia were prospectively recruited and assayed. Over half of the adult solid tumors show detectable rDNA rearrangements relative to either surrounding non-tumor tissue or normal peripheral blood. In contrast, there is a greatly reduced frequency of alteration in pediatric leukemia. This finding makes rDNA restructuring one of the most common chromosomal alterations in adult solid tumors, illustrates the dynamic plasticity of the human genome, and may have prognostic or predictive value in disease progression.

The superinfection exclusion gene (sieB) of Salmonella phage P22 was mapped using phage deletion mutants. The DNA sequence in the region was re-examined in order to find an open reading frame consistent with the deletion mapping. Several discrepancies from the previously published sequence were discovered. The revised sequence revealed a single open reading frame of 242 codons with six likely translation initiation codons. On the basis of deletion and amber mutant phenotypes the second of these six sites was inferred to be the translation initiation site of the sieB gene. The sieB gene encodes a polypeptide with 192 amino acid residues with a calculated molecular weight of 22,442, which is in reasonable agreement with that estimated from polyacrylamide gels. The transcription start-site of sieB was identified by the use of an RNAase protection assay. The sieB promoter thus identified was inactivated by a two-base substitution in its -10 hexamer. The sieB gene of coliphage λ was also identified. The promoter for λ sieB was identified by homology to that of P22 sieB. sieB aborts the lytic development of some phages. P22 itself is insensitive to the lethal effect of SieB because it harbours a determinant called esc. It was found that the sieB gene encodes two polypeptides-SieB, which is the exclusion protein, and Esc, which is a truncated version of SieB that inhibits its action. Superinfecting P22 synthesizes an antisense RNA, sas, that inhibits synthesis of SieB but allows continued synthesis of Esc, thus allowing P22 to by-pass SieB-mediated exclusion. This translational switch induced by sas RNA is essential to vegetatively developing P22; a mutation that prevents this switch causes P22 to commit SieB-mediated suicide. It was also found that P22's Esc allows it to circumvent the SieB-mediated exclusion system of bacteriophage λ.

This thesis is divided into three parts, united by the theme of the development of selective inhibitors of mammalian cell DNA polymerase delta (pol δ). The first part consists of an investigation of the cytotoxic mechanism(s) of certain 2-substituted adenine analogs, selected on the basis of their inhibitory properties towards DNA polymerase alpha (pol α) and mammalian cell DNA synthesis. The second is a direct search for inhibitors of isolated pol δ, and an investigation of inhibitory mechanisms. The third consists of measurement of the effects of a selective pol δ inhibitor on cellular DNA synthesis. Mechanism of Cytotoxicity of 2-substituted adenine analoqs. The mechanism of inhibition by 2-(p-n-butylanilino)-2'-deoxyadenosine (BuAdA), and related compounds, of Chinese hamster ovary (CHO) cell ([3H]thymidine [3H]TdR) incorporation, was investigated. The potency of the compound could largely be explained by its potency (IC50 = 23 μM) as an inhibitor of CHO cell [3H]TdR uptake. BuAdA inhibited incorporation by CHO cells of [32p]phosphate into DNA relatively weakly, displaying an IC50value of 80 μM. Differential inhibition of DNA polymerases alpha and delta. Known DNA polymerase inhibitors of a structurally wide range were screened for their ability to inhibit pol δ derived from calf thymus selectively with respect to pol α derived from the same tissue. Pyrophosphate (PPi) and difluoromethanediphosphonate each inhibited pol δ weakly, but with greater potency than pol α. Based on this lead, an expanded series of PPi analogs was screened. Carbonyldiphosphonate (COMDP) inhibited pol δ with a potency (Ki = 1.8 μM) twenty-two times greater than that displayed for pol α. Kinetic studies indicated that COMDP inhibited pol δ competitively with the dNTP specified by the template, but not competitively with the template:primer. Analogous experiments with pol α showed that the compound inhibited that enzyme uncompetitively with the dNTP, and not competitively with the template:primer. COMDP was a weak inhibitor of the 3' → 5' exonuclease activity of pol δ, displaying an IC50value greater than 1 mM. Inhibition of permeabilized cell DNA synthesis bv a selective pol δ inhibitor. The potency of COMDP as an inhibitor of permeabilized CHO cell DNA synthesis (IC50= 200 μM) did not clearly indicate the participation of pol δ in cellular DNA replication. Prospectus. The thesis concludes with a prospectus for the development of pol δ inhibitors with improved properties compared to COMDP.

The error free repair of DNA double strand breaks through the homologous recombinational repair pathway is essential for organisms of all types to sustain life. A detailed structural and mechanistic understanding of this pathway has been the target of intense study since the identification of bacterial recA, the gene whose product is responsible for the catalysis of DNA strand exchange, in 1965. The work presented here began with defining residues that are important for the assembly and stability of the RecA filament, and progressed to the identification of residues critical for the transfer of ATP-mediated allosteric information between subunits in the protein's helical filament structure. My work then evolved to investigate similar mechanistic details concerning the role of ATP in the human RecA homolog, Rad51. Results from non-conservative mutagenesis studies of the N-terminal region of one subunit and the corresponding interacting surface on the neighboring subunit within the RecA protein, led to the identification of residues critical for the formation of the inactive RecA filament but not the active nucleoprotein filament. Through the use of specifically engineered cysteine substitutions we observed an ATP-induced change in the efficiency of cross subunit disulfide bond formation and concluded that the position of residues in this region as defined by the current crystal structure may not accurately reflect the active form of the protein. These ATP induced changes in positioning led to the further investigation of the allosteric mechanism resulting in the identification of residue Phe217 as the key mediator for ATP-induced information transfer from one subunit to the next. In transitioning to investigate homologous mechanisms in the human pathway I designed a system whereby we can now analyze mutant human proteins in human cells. This was accomplished through the use of RNA interference, fluorescent transgenes, confocal microscopy and measurements of DNA repair. In the process of establishing the system, I made the first reported observation of the cellular localization of one of the Rad51 paralogs, Xrcc3, before and after DNA damage. In addition we found that a damage induced reorganization of the protein does not require the presence of Rad51 and the localization to DNA breaks occurs within 10 minutes. In efforts to characterize the role of ATP in human Rad51 mediated homologous repair of double strand breaks we analyzed two mutations in Rad51 specifically affecting ATP hydrolysis, K133A and K133R. Data presented here suggests that, in the case of human cells, ATP hydrolysis and therefore binding, by Rad51 is essential for successful repair of induced damage.

Cytoplasmic dynein is a multisubunit complex involved in retrograde transport of cellular components along microtubules. The heavy chains (HC) are very large catalytic subunits which possess microtubule binding ability. The intermediate chains (IC) are responsible for targeting dynein to its appropriate cargo by interacting with the dynactin complex. The light intermediate chains (LIC) are previously unexplored subunits that have been proposed to modulate dynein activity by regulating the motor or the IC-dynactin interaction. The light chains (LC) are a newly identified class of subunit which are also thought to have regulatory functions. In the first part of this work, I analyzed the relationship between the four SDS-PAGE gel bands that comprise the light intermediate chains. 1- and 2-D electrophoresis before and after alkaline phosphatase treatment revealed that the four bands are derived from two different polypeptides, each of which is phosphorylated. Peptide microsequencing of these subunits yielded sequences that indicated similarity between them. cDNA cloning of the rat LICs revealed the presence of a conserved P-loop sequence and a very high degree of homology between the two different rat LICs and among LICs from different species. The second series of experiments was designed to analyze the association of pericentrin with cytoplasmic dynein. First, various dynein and dynactin subunits were co-associate with pericentrin in these experiments. Co-precipitation from 35S labeled cell extracts revealed a direct interaction between LIC and pericentrin. Comparison of pericentrin binding by LICl and LIC2 showed that only LICl was able to bind. Further investigation of the relationship between LICl and LIC2 demonstrated that each LIC will self-associate, but they will not form heterooligomers. Additionally, using co-overexpression and immunoprecipitation of LICl, LIC2, and HC, I have shown that binding of the two LICs to HC is mutually exclusive. Finally, I investigated the relationships between dynein HC, IC, and LIC by examining the interactions among the subunits. IC and LIC were both found to bind to the HC, but not to each other. Despite the lack of interaction between IC and LIC, they are, in fact, present in the same dynein complexes and they have partially overlapping binding sites within the N-terminal sequence of the HC. The HC dimerization site was determined to extend through a large portion of the N-terminus, and it includes both the IC and LIC binding sites, although these subunits are not required for dimerization. Together these studies implicate the light intermediate chains in dynein targeting. Targeting of dynein to its cargo has been thought to be performed by the dynactin complex, and for one particular cargo, the kinetochore, there is considerable evidence to support this model. The results presented here suggest that the light intermediate chains appear to function in a separate, non-dynactin-based targeting mechanism.

The lcrF gene of Yersinia pestis, the etiological agent of plague, encodes a transcription activator responsible for inducing expression of several virulence-related proteins (Yops) in response to temperature. The mechanism of this thermoregulation was investigated. Using a yopE::lacZ reporter fusion, lcrF-mediated thermal regulation was observed in Y. pestis and Escherichia coli. The lcrF gene was sequenced, the 30.8 kDa. LcrF protein identified and purified, and LcrF-dependent yopE-specific DNA binding activity was detected. A sequence similarity search revealed that LcrF exhibits 98% homology to VirF of Yersinia enterocolitica and significant homology to the carboxy termini of other members of the AraC family of transcription activators. During localization studies, a significant proportion of LcrF was found associated with the membrane fraction in E. coli. However, pulse-chase experiments indicated that this result is an artifact of fractionation. lcrF-mediated thermal induction of the yopE::lacZ reporter fusion remains intact in a Shigella flexneri virR mutant. The virR mutation is known to affect thermal induction of Shigellavirulence genes, which are also controlled by an activator in the AraC family. As a first step toward identifying the temperature-sensitive step in the regulation of yop expression, lcrF::lacZ transcriptional fusions were constructed and analyzed in Y. pestis and E. coli. The activity of the fusions was not affected by the native pCD1 virulence plasmid, an intact lcrF gene, or temperature. Thus, induction of lcrF transcription is not essential for temperature-dependent activation of yopE transcription. To confirm these results, attempts were made to identify both the native lcrF message in Y. pestis, and a lcrF-lacZ hybrid message in Y. pestis and E. coli. These attempts were unsuccessful. Examination of LcrF protein production revealed temperature-dependent expression in Y. pestis. Surprisingly, high-level T7 polymerase-directed transcription of the lcrF gene in Escherichia coli also resulted in temperature-dependent production of the LcrF protein. Pulse-chase experiments showed that the LcrF protein was stable at both 26 and 37°C, suggesting that translation rate or message degradation is thermally controlled. Comparison of the amount of LcrF protein produced per unit of message at 26 and 37°C in E. coli indicated that the efficiency of translation of lcrF message increased with temperature. mRNA secondary structure predictions suggest that the lcrF Shine-Dalgarno sequence is sequestered in a stem-loop. A model in which decreased stability of this stem-loop with increasing temperature leads to increased efficiency of translation initiation of lcrF message is presented.

The Dam-dependent mismatch repair system of Escherichia coli is part of a large network of DNA surveillance and error avoidance systems that identify and repair DNA damage. In this thesis, I have investigated the Dam-dependent mismatch repair system of E. coliand its role in the recognition and repair of DNA substrate molecules containing small insertion/deletion heterologies. This investigation was divided into two parts: the first part utilized genetic techniques to evaluate the specificity of repair and the second part utilized biochemical approaches into the recognition of insertion/deletion heterologies. I have developed a sensitive in vivo transformation system to rapidly evaluate the repair of small insertion/deletion heterologies by Dam-dependent mismatch repair. Heteroduplexes were constructed, for each state of methylation of d(GATC) sequences, by annealing single strand DNA to the linearized complementary strand of duplex DNA. The unmethylated single strand DNA was isolated from f1 phage (R408) propagated on a strain of E. coli containing the dam-16 allele (Chapter 2) to eliminate the possibility of residual Dam-methylation of d(GATC) sequences. Tranformation of E. coli indicator strains with heteroduplexes containing 1, 2, 3, 4 and 5 base insertion/deletion heterologies were scored for repair based on colony color. The results of these experiments show that the Dam-dependent mismatch repair system can recognize and repair 1, 2 and 3 base heterologies as well as repairing G/T mispairs (Chapters 3 and 4). The repair of 4 base heterologies was marginal, while no repair was observed with 5 base heterologies (Chapters 3 and 4). Repair of the 1, 2, 3 and 4 base heterologies proceeded in a Dam-dependent process that required the gene products of mutL, mutS, and I have demonstrated that MutS protein from both Salmonella typhimurium and E. coli can recognize and bind in vitro to the same 1, 2, 3 and 4 base heterologies used for the genetic studies above (Chapters 4 and 5). In fact, MutS protein binds to 1, 2 and 3 base heterologies with greater affinity than it binds to a G/T mismatch. The in vitro observation that MutS does not bind to 5 base heterologies is consistent with the in vivoobservation that 5 base heterologies are not subject to repair. I have also shown that MutS protein specifically binds to 1, 2 and 3 base heterologies since MutS protects about 25 base pairs of DNA flanking the site of the heterology from DNaseI digestion. The results of the genetic and biochemical experiments described in this thesis (and summarized above) serve to re-emphasize the importance of the role that methyl-directed mismatch repair plays in mutation avoidance, and hence in the preservation of genetic integrity.

The cell division of Saccharomyces cerevisiae is controlled by the action of pheromones at the G1 phase of the cell cycle. A general method was developed for the isolation of constitutive mutants in the pheromone response pathway. Recessive alleles of the SCG1 gene (encoding the α subunit of a G protein) were isolated as well as a dominant mutation in the STE4 gene (encoding the β subunit of a G protein). Analysis of double mutants suggested that the STE4 gene product functions after the SCG1 product but before the STE5 gene product. Double mutants carrying either scg1 or STE4Hp1 constitutive alleles together with the temperature-sensitive unresponsive mutation, ste5-3ts, showed arrest and recovery when shifted from 34° C to 22° C. Recovery from the constitutive signal was independent of the receptor. The STE4Hp1 sst2 ste5ts triple mutant was not able to recover from arrest, suggesting that an SST2-dependent mechanism is involved in recovery of the STE4Hp1 mutant from constitutive arrest. In contrast, the scg1-7 sst2 ste5ts triple mutant recovered only partially suggesting that even though SST2 gene product is probably involved in recovery of the scg1-7 mutant, this mutant can recover by an SST2-independent mechanism. This implies existence of another, SST2-independent postreceptor recovery mechanism. The scg1-null mutant do not recover from constitutive arrest (J. Hirschman, personal communication). Both recovery mechanisms probably operate at the G protein step. Isolation of a constitutive allele of STE5 allowed the definition of its site of action as being after the STE4-controlled step. In addition, constitutive activation of the pheromone pathway by STE5Hp1 mutation was found to be partially dependent on the STE4 and STE18 gene products, the β and γ subunits of a G protein. A comprehensive genetic model is presented to explain the mechanisms of signal transduction and recovery.