UCSF

Precise control of gene expression underlies cellular identity and function in eukaryotes. This control is achieved by gene activation pathways, which select particular sequences for expression, and by gene silencing pathways, which prevent the expression of unnecessary gene products. The latter class of pathways not only prevents the metabolic cost associated with expression of unnecessary genes, but also protects against the expression of deleterious gene products, like transposons. Understanding the mechanisms by which gene silencing pathways achieve specificity for their diverse targets is a central goal in the study of eukaryotic gene expression. In this thesis, I use the human fungal pathogen Cryptococcus neoformans as a model system for the study of target specificity in two gene silencing pathways: post-transcriptional silencing by RNA interference (RNAi) and transcriptional silencing by Polycomb group proteins.

RNAi pathways, in which small RNAs guide the silencing of complementary target transcripts, act to suppress transposons. But the signals that distinguish their targets from host genes remain unclear. Here, I identify SCANR, a spliceosome- associated protein complex that mediates small RNA production specifically from unspliced mRNA precursors whose suboptimal introns cause them to stall on spliceosomes. These findings identify a new role for the spliceosome in genome defense, in which the distinct splicing signals of transposon transcripts provide an opportunity for cells to distinguish transposons from host genes.

Polycomb proteins deposit H3 lysine 27 methylation (H3K27me) to establish heterochromatin domains, but the mechanisms that specify domain extent and location are unknown. Here, I characterize a Polycomb complex in C. neoformans, demonstrate its role in subtelomeric gene silencing, and find that it physically interacts with the methyl mark it deposits: H3K27me. Disruption of this binding activity causes aberrant commingling of Polycomb heterochromatin with other types of repressive heterochromatin, thereby revealing a mechanism that prevents inappropriate crosstalk between discrete chromatin domains.

Finally, I characterize the signal recognition particle (SRP)--a ribonucleoprotein complex required for protein sorting--in C. neoformans, its first description in basidiomycetous yeast. I reveal unprecedented features in its RNA and protein components, thereby contributing to our understanding of SRP diversity, evolution, and function in lower eukaryotes.