Functional Synonymous SNVs: Novel Methodology and Framework for Identification

Date of Award

Winter 12-15-2012

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Human & Statistical Genetics)

Degree Name

Doctor of Philosophy (PhD)

Degree Type

Dissertation

Abstract

Despite their demonstrated effect on select proteins through the alteration of splicing, protein expression, folding, and function, the functional importance of synonymous single nucleotide variations (sSNVs), because they don't change the sequence of the encoded protein, continues to be discounted by the majority of the scientific community. The lack of any large-scale studies focused on sSNVs has hindered the development of a community-wide mechanistic understanding and framework for the identification of potentially functional synonymous SNVs.

We motivate functional interest in sSNVs in the second chapter of this dissertation by showing that sSNV mutational preferences are highly polarized and non-compensatory for this class of variations. For the first time, we demonstrate that variations that result in codons with the same base(2,3) dinucleotide exhibit more consistent and less variable cancer tissue distributions than sSNVs. We show that transmembrane signal transduction proteins and proteins involved in cell movement are significantly enriched in the set of proteins containing sSNVs thought to be involved in cancer. Such enriched ontologies point to proteins for which small variations in expression can lead to significant downstream consequences.

In the third chapter, we, for the first time, provide a methodological framework for identifying potentially functional synonymous mutations. We develop a novel metric that can take codon usage frequency information and relate it to tRNA abundance to 98% correspondence. This allows for the approximation of the consequence of a sSNV on the tRNA selection rate and resulting protein expression level. We combine this with information relating to whether an sSNV introduces or eliminates splicing element binding sites to predict the functionality of sSNVs in dbSNP.

Having examined the effect of variations that introduce or eliminate codons that can alter the translation rate (and thus expression level) of proteins in the preceding chapters, in the fourth chapter, we look at potential translation-rate-altering codons in wildtype proteins. We address the hotly-debated issue of whether rare (translation-slowing) codons cluster around human protein domain boundaries. Utilization of our novel metric that addresses tRNA selection rate variations indicates that rare codons do not cluster around domain boundaries, and that such clusters are actually selected against shortly downstream of protein boundaries.

Together, the findings presented in this dissertation support the notion that synonymous codons are not functionally equivalent, and that they have the capacity to affect everything from protein folding to cancer progression. As such, the functional relevance of synonymous changes to such codons should not continue to be ignored, as doing so actively hinders a thorough mechanistic understanding of disease susceptibility, progression, and regulation.

Language

English (en)

Chair and Committee

Michael Province

Committee Members

Anne Bowcock, Justin Fay, Michael Lovett, John Rice, Nancy Saccone, Gary Stormo

Comments

Permanent URL: https://doi.org/10.7936/K7S46PVH

This document is currently not available here.

Share

COinS