Massively parallel functional analyses of human PTEN variants

Oregon Health & Science University, Portland OR

Investigators

Linked publications, trials & patents

Abstract

Project Summary We are now able to routinely sequence human genomes at single-base resolution. However, our ability to interpret the functional consequences of detected mutations has lagged behind. Computational approaches scale well but have poor accuracy, whereas retrospective analysis of detected variants has high accuracy but does not scale well. In order to solve this problem, a new experimental paradigm has emerged to empirically characterize the effects of mutations with high accuracy at scale. This approach takes advantage of recent and ongoing improvements in DNA synthesis and sequencing, and has the potential to offer unprecedented insight into protein biochemistry and human disease. We believe these insights will prove to be critical for unlocking the potential of genomic medicine. In this project we seek to comprehensively assess multiple molecular effects of PTEN mutations on protein function, and assess the utility of this data as a predictor for human clinical phenotype. The PTEN protein is a tumor suppressor that is frequently mutated in diverse human cancers and in the germline of some individuals with overgrowth disorders, cancer predisposition syndromes, or autism. Currently, it is impossible to predict the effects of the vast majority of PTEN germline mutations. Since the phenotypic spectrum of PTEN mutation carriers is broad, it would be highly valuable to understand the ways in which phenotypic outcomes arise from PTEN mutation genotypes. In Aim 1, we will first employ a yeast-based screen to assess the effects all PTEN single amino acid mutations on lipid phosphatase activity, the primary biochemical function of PTEN protein. It is known that several pathogenic variants are destabilized. Therefore, in Aim 2, we will perform a second, independent screen to assess the steady state protein stability of all PTEN single amino acid mutations. In Aim 3, we will use the data derived from this study as well as publically available biochemical information to train a classifier model to predict the relationship between the mutation genotypes and clinical phenotypes observed in humans. These data will increase our fundamental understanding of PTEN function and the role of mutations in diverse disorders, and could provide a valuable clinical tool that would increase the quality of life for PTEN mutation carriers.