Fitness analyses of all possible point mutations for regions of genes in yeast

Abstract

Deep sequencing can accurately measure the relative abundance of hundreds of mutations in a single bulk competition experiment, which can give a direct readout of the fitness of each mutant. Here we describe a protocol that we previously developed and optimized to measure the fitness effects of all possible individual codon substitutions for 10-aa regions of essential genes in yeast. Starting with a conditional strain (i.e., a temperature-sensitive strain), we describe how to efficiently generate plasmid libraries of point mutants that can then be transformed to generate libraries of yeast. The yeast libraries are competed under conditions that select for mutant function. Deep-sequencing analyses are used to determine the relative fitness of all mutants. This approach is faster and cheaper per mutant compared with analyzing individually isolated mutants. The protocol can be performed in ∼4 weeks and many 10-aa regions can be analyzed in parallel.

Figure 1

Bulk competition of libraries of point mutants in yeast. (a) Plasmid libraries are transformed into yeast. (b) Yeast that have taken up a plasmid are selected for and amplified. (c) Selection pressure is applied to the library copy of the mutated gene and samples are collected over time in bulk-competition.

Figure 2

Steps to generate plasmid libraries of point mutants. (a) Whole-plasmid PCR to generate inverted BsaI vector. (b) Digestion of this vector to generate directional sticky-ends. (c) Cassette ligation to introduce point mutants.

Figure 3

Steps to prepare DNA for deep sequencing. (a) PCR amplify mutant library using primers specific to plasmid library. (b) Perform second PCR step to add MmeI site to 5’ end and Illumina universal primer sequence to 3’ end. (c) Perform MmeI digestion to create sticky end adjacent to randomized region of the mutant library. (d) Ligate an adapter to the 5’ end containing a barcode. (e) PCR with universal deep sequencing primers. (f) Parallel analyses of a wild-type plasmid provide information on mis-reads.

Figure 4

Analysis pipeline for measuring fitness effects of mutations from deep sequencing data. (a) Sequences that pass quality filtering at all positions in the read are stored in a sequence only file. (b) The occurrence of each unique sequence read is summed resulting in a dramatic compression of file size. (c) At each time-point, calculate the relative abundance of each point-mutant in the library. For the control samples, calculate the mis-read rate per base (d) Calculate the fitness of each point mutant based on its change in relative abundance over time.