Sequence-function relationships in intrinsically disordered regions through the lens of evolution

PhD Exit Seminar

Taraneh Zarin (Moses lab)

Intrinsically disordered regions (IDRs) are regions of proteins that do not autonomously fold into stable secondary or tertiary structures. Though they defy the classical view of proteins as rigidly structured macromolecules, IDRs are widespread in living organisms, and are involved in a diverse array of functions. The majority of IDRs appear to be rapidly evolving at the level of the primary amino acid sequence, which makes it difficult to quantify evolutionary conservation and associate these regions with biological function using standard sequence analysis. The aim of my thesis research has thus been to understand evolutionary constraint and sequence-function relationships in IDRs. Using a functionally characterized IDR in the yeast protein Ste50, I first found that highly diverged amino acid sequences can encode conserved phenotypes in IDRs, showing that sequence divergence does not necessarily imply functional divergence in these regions. Using a phylogenetic comparative framework, I found that the net charge of the Ste50 IDR, rather than the precise amino acids, is a functional molecular feature that is preserved over evolution. I next expanded my evolutionary analysis of IDRs to the yeast proteome, and found that most highly diverged IDRs contain many molecular features that are preserved over evolution. I summarized the evolution of these molecular features with an “evolutionary signature” for each IDR, and found that groups of IDRs with similar evolutionary signatures are enriched for specific biological functions. I also found that IDRs with similar evolutionary signatures can rescue function in vivo despite negligible sequence similarity. Finally, I used these evolutionary signatures to train a statistical model, and found that they can be used to classify IDRs for a diverse set of biological functions. I identified the molecular features contributing to these functional predictions, and attributed distinct functions to specific IDRs in proteins with multiple IDRs. Overall, this work shows that there is rich functional information in IDR sequences, and that this information can be revealed through evolutionary analysis.