PgmNr Y3178: Expanding the yeast genetic toolkit: developing a pooled assay for genetic interactions.

Authors:
M. Jaffe 1 ; J. D. Smith 1 ; R. P. St. Onge 1 ; G. Sherlock 1 ; S. F. Levy 2


Institutes
1) Stanford Univ., Stanford, CA; 2) Stony Brook Univ., Stony Brook, NY.


Keyword: Networks

Abstract:

Two genes are said to interact when the fitness of a double mutant deviates from the multiplicative fitness of the two corresponding single mutants. Saccharomyces cerevisiae is an ideal system to measure these interactions quantitatively and in high-throughput because of its compact genome and its wealth of genetic tools. In order to improve current screening methods, we developed, and present here, two iterations of a pooled assay of genetic interactions named iSeq. These pool-based methods increase the scalability of measurements across growth conditions, because, in contrast to prior methods, a pool of mutants must only be generated once before assaying multiple conditions. In the first iteration of iSeq, we generated ~400 mutant strains representing 45 possible single or double gene deletions, with multiple strains per genotype. Starting with deletions from the yeast deletion collection, we introduced a novel DNA double-barcoding technology to each double deletion strain to use as its molecular identifier during pooled growth. iSeq fitness and interaction score measurements for each strain in the pool were reproducible across three independent cultures for three growth conditions tested (ρ = 0.91-0.99). These measurements also correlated with an independent OD-based growth assay (ρ = 0.68-0.69). However, we observed high variability in estimates across strains carrying the same gene deletion(s), but unique barcodes (N = 4—16 per genotype). Whole-genome sequencing of 73 double mutant and parental strains revealed segregating and de novo mutations were common and accumulated over the two rounds of mating and selection used to generate strains. This genetic variability likely led to the observed fitness differences, and observations of aneuploidy were often associated with lower fitness estimates. To reduce this genetic variability for future screens, we are developing a second iteration of iSeq that relies on an inducible CRISPR/Cas9 gene repression system to generate single and double mutants without any mating or selection steps. As validation, we are using this system to perform a preliminary screen for genetic interactions of 9x459 genes, also in a pooled format. Next, we aim to apply this technology to discover novel gene functions and further understand the topology of the genetic interaction network and its dynamics across growth conditions.