Epiocentre Bioitechnologies

Accurate, large-scale gene annotation is a major challenge in biology. Traditionally, gene function can be assigned by observing the phenotype of a null mutant under various conditions. This principle has engendered the creation of genome-wide deletion collections in which each gene in a genome is knocked out or disrupted. Phenotype can then be assayed for each gene knockout individually. Parallelization of phenotypic assays via the introduction of molecular barcodes has proved invaluable for interrogating genome-wide collections of mutants in many conditions. This approach has been most visibly successful in the case of the budding yeast Saccharomyces cerevisiae, but it has the potential to bring insight to the genomes of many more microorganisms. The strategy of creating the genome-wide mutant collections that permit parallel phenotypic analysis has not been broadly applied beyond S. cerevisiae due primarily to technical limitations. In this dissertation, we describe a universal approach to rapidly generate comparable tagged, mutant collections. This approach combines DNA tag technology with transposon mutagenesis, and thereby can be generalized to any microorganism amenable to transposon mutagenesis. We created a universal collection of tags whose utility can extend to any application requiring sample tracking or multiplexing. We describe the validation of this tag resource as well as its use in the transposon mutagenesis of two different microorganisms, the bacteria Shewanella oneidensis MR-1 and the yeast Candida albicans. We then describe the expansion of this method to generate a genome-wide tagged transposon mutant collection in C. albicans. We used this collection to identify i) haploinsufficient genes in different nutrient conditions, and ii) mechanisms of drug-induced haploinsufficiency in C. albicans, and illustrated how these data can be used as a resource for genome annotation and hypothesis generation. Moreover, these studies illustrate why direct study of C. albicans is necessary, because relying solely on similarities between its traditional model organism, S. cerevisiae, would result in exclusion of C. albicans-specific genes and processes that are involved in its pathogenesis and may prove to be novel therapeutic targets.