In addition to ATP generation, mitochondria are essential in various cellular processes ranging from biosynthetic pathways, apoptosis, cell cycle progression, and calcium buffering. Studies in living cells have now firmly established that mitochondria exist as a dynamic network sculpted by fission and fusion reactions, rather than separated, individual organelles. Not surprisingly, mutations in genes involved in mitochondrial dynamics and quality control lead to human diseases such as Charcot-Marie-Tooth disease type 2A, Optic atrophy, and autosomal recessive Parkinson disease. I have designed a high-throughput protocol to permit genome-wide screening for novel genes that are required for normal mitochondrial morphology. I have executed a genome-wide RNA interference screen and identified several novel genes required for mitochondrial dynamics in addition to known regulators of mitochondrial dynamics. A detailed high-throughput genome-wide screening protocol is presented. I have shown that TID1, a gene identified from the screen, has a dual-role in maintaining the integrity of mitochondrial DNA and preventing the aggregation of complex I subunits. My analysis of the mitochondrial role of TID1 supports the existence of a TID1- mediated stress response to ATP synthase inhibition. The genome screen also identified the novel gene ROMO1 as essential for normal mitochondrial morphology. I have shown that ROMO1 may have an additional role in maintaining mitochondrial spare respiratory capacity, possibly by affecting cellular substrate availability. Finally, in a collaborative effort, we have shown that homozygous mutations in the mitochondrial fusion gene MFN2 lead to multiple symmetric lipomatosis (MSL) associated with neuropathy. Mechanistically, this mutation reduces MFN2 homocomplex formation. Taken together, these results show the utility of genome-wide screening in identifying genes involved in mitochondrial quality control.

Meiosis is a specialized form of cell division in sexually reproducing eukaryotes. Crossovers are physical connections formed between homologous chromosomes during meiosis; these connections help ensure normal segregation of homologous chromosomes at meiosis I. However, the yeast Saccharomyces cerevisiae and other eukaryotes can still segregate homologs properly even in the absence of some crossovers. This is due to a backup mechanism known as distributive segregation, which correctly segregates non-crossover chromosomes at a higher rate than if segregation were completely random. To study distributive segregation, we have generated diploid yeast with one homeologous chromosome pair consisting of a Saccharomyces cerevisiae chromosome V and a Saccharomyces carlsbergensis chromosome V. This pair of chromosomes rarely recombine resulting in crossing over occurring in less than 3% of meiosis. Appropriate segregation of this chromosome pair during meiosis will depend on distributive segregation; we can then assess the possible roles of candidate proteins in distributive segregation through determination of the effect of mutation on segregation of this chromosome pair. Our work has focused on the roles of three proteins, Ndj1, Tid1, and Spo16. These three proteins affect meiosis in many ways, including the efficiency of crossover regulation and the overall timing of meiosis, but their roles during distributive segregation are not fully known. A comparison of spore viability among WT, ndj1, and tid1 strains reveals an elevated incidence of 2-spore-viable tetrads (suggestive of chromosome nondisjunction) in ndj1, but not tid1; these results suggest that the Ndj1 protein, but not the Tid1 protein, plays some role in distributive segregation. spo16 strains seem to also show elevated levels of 2-spore-viable tetrads, but due to a lack of data no deductions can be made about the role of Spo16 in distributive segregation.

Activity of the hypoxia inducible HIF-alpha transcription factors drive the hypoxic response, resulting in enhancement of angiogenesis, tumour growth, invasion and metastasis. Seeking to uncover a role for Tid1 in control of HIF2-alpha, we used lentiviral shRNA to knock-down Tid1 in 786-0 RCC cells with and without pVHL. In 786-0 cells stably expressing pVHL30, Tid1 knock-down resulted in a dramatic reduction in HIF2-alpha levels relative to controls. Adenoviral-mediated overexpression of Tid1S rescued this decline in HIF2-alpha levels, while overexpression of Tid1L enhanced this decline. A protective role of Tid1S for HIF2-alpha was reproduced in a HEK293 cell model. Immunoprecipitations in HEK293 cells revealed a lack of direct binding between HIF2-alpha and Tid1 in vivo, while adenoviral-mediated overexpression of Tid1 in this model failed to alter in vitro binding between HIF2-alpha and pVHL30. We present a model in which Tid1 regulates HIF2-alpha stability through regulation of pVHL30 nuclear import.

Activity of the hypoxia inducible HIF-alpha transcription factors drive the hypoxic response, resulting in enhancement of angiogenesis, tumour growth, invasion and metastasis. Seeking to uncover a role for Tid1 in control of HIF2-alpha, we used lentiviral shRNA to knock-down Tid1 in 786-0 RCC cells with and without pVHL. In 786-0 cells stably expressing pVHL30, Tid1 knock-down resulted in a dramatic reduction in HIF2-alpha levels relative to controls. Adenoviral-mediated overexpression of Tid1S rescued this decline in HIF2-alpha levels, while overexpression of Tid1L enhanced this decline. A protective role of Tid1S for HIF2-alpha was reproduced in a HEK293 cell model. Immunoprecipitations in HEK293 cells revealed a lack of direct binding between HIF2-alpha and Tid1 in vivo, while adenoviral-mediated overexpression of Tid1 in this model failed to alter in vitro binding between HIF2-alpha and pVHL30. We present a model in which Tid1 regulates HIF2-alpha stability through regulation of pVHL30 nuclear import.

Meiosis is the specialized process of cell division utilized during gametogenesis in all sexually reproducing eukaryotes, which consists of one round of DNA replication followed by two rounds of chromosome segregation and results in four haploid cells. Crossovers between homologous chromosomes promote proper alignment and segregation of chromosomes during meiosis. Crossover interference is a genetic phenomenon in which crossovers are non-randomly placed along chromosomes. Crossover assurance ensures that every homologous chromosome pair obtains at least one crossover during Prophase I. Crossovers physically connect homologous pairs, allowing spindle fibers to attach and separate homologs properly. However, some organisms have shown an ability to segregate chromosomes that fail to receive at least one crossover, a phenomenon termed distributive disjunction. In Saccharomyces cerevisiae, mutation of either Tid1 or Ndj1 results in a similar defect in crossover interference. The overall number of crossovers is not substantially different from the wild type, however they are distributed more randomly with respect to each other. In this thesis, the roles of Tid1 and Ndj1 on crossover assurance and distributive disjunction have been further elucidated through use of knock-out mutants and tetrad dissection. To analyze meiotic chromosome segregation in isogenic tid1 and ndj1 strains, the spore viability of dissected tetrads was utilized as an indirect measure of nondisjunction events. An elevated number of 2- and 0- spore viable tetrads were seen in ndj1, but not tid1 yeast, confirming previous results. Elevated 2- and 0- spore viable tetrads are an indication of meiosis I (MI) nondisjunction, commonly resulting from failure of crossover formation. These results suggest crossover assurance is disrupted in njd1, but not tid1 mutants. However, MI chromosome segregation is an indirect readout of crossover formation; distributive disjunction, for example, can lead to proper segregation of achiasmate chromosomes. To determine if distributive disjunction is functional in yeast, wild type, tid1 and ndj1 versions of diploid yeast carrying a single homeologous pair of chromosomes were constructed. These strains have one chromosome (chr. III or V) replaced with one from a closely related species of yeast. The homeologous chromosome functionally replaces the homolog, however crossovers are significantly reduced between homeologs. A spore viability pattern typical of MI nondisjunction was detected in ndj1 mutants, but not in tid1 mutants. In the context of these homeologs, this pattern is suggestive of a role for Ndj1, but not Tid1, in distributive disjunction. Further, these results suggest that tid1 and ndj1 mutant yeast may not be different in their competence for crossover assurance. To directly assay competence for crossover assurance in native mutants, the incidence of E0 chromosome pairs (those lacking crossovers) was determined. To do this we assayed crossover formation along the length of chromosome III of isogenic wild type, ndj1 and tid1 mutant strains. The incidence of E0 chromosomes was comparably elevated in both tid1 and ndj1 mutant yeast, suggesting that crossover assurance is nonfunctional in both strains. We find evidence that supports the idea that interference and assurance are genetically linked. Our data also suggests that distributive disjunction may be genetically separable from some meiotic genes.