Mapping small effect mutations in Saccharomyces cerevisiae: impacts of experimental design and mutational properties

Abstract

Genetic variants identified by mapping are biased toward large phenotypic effects because of methodologic challenges for detecting genetic variants with small phenotypic effects. Recently, bulk segregant analysis combined with next-generation sequencing (BSA-seq) was shown to be a powerful and cost-effective way to map small effect variants in natural populations. Here, we examine the power of BSA-seq for efficiently mapping small effect mutations isolated from a mutagenesis screen. Specifically, we determined the impact of segregant population size, intensity of phenotypic selection to collect segregants, number of mitotic generations between meiosis and sequencing, and average sequencing depth on power for mapping mutations with a range of effects on the phenotypic mean and standard deviation as well as relative fitness. We then used BSA-seq to map the mutations responsible for three ethyl methanesulfonate-induced mutant phenotypes in Saccharomyces cerevisiae. These mutants display small quantitative variation in the mean expression of a fluorescent reporter gene (-3%, +7%, and +10%). Using a genetic background with increased meiosis rate, a reliable mating type marker, and fluorescence-activated cell sorting to efficiently score large segregating populations and isolate cells with extreme phenotypes, we successfully mapped and functionally confirmed a single point mutation responsible for the mutant phenotype in all three cases. Our simulations and experimental data show that the effects of a causative site not only on the mean phenotype, but also on its standard deviation and relative fitness should be considered when mapping genetic variants in microorganisms such as yeast that require population growth steps for BSA-seq.

Keywords: Bulk Segregant Analysis; FASTER MT; Mutagenesis screen; Next Generation Sequencing; TDH3.

Figure 1

Experimentally controllable parameters affect statistical power for detecting a significant difference in the frequency of a causal mutation between bulks. (A) An overview of the modeled BSA-seq experiment is shown, with the four experimental parameters we allowed to vary (population size, generations of growth, cutoff for bulk selection, and average coverage of sequencing) indicated. Power is shown for various population sizes (B−D), generations of growth (E−G), bulk selection cutoffs (H−J), and average sequencing coverages (K−M), for a range of effects of the causal mutation on mean expression (B, E, H, K), standard deviation of expression (C, F, I, L), and fitness (measured in terms of the selection coefficient) (D, G, J, M). In all plots, the dashed line indicates 90% power. Gray shaded regions represent 90% confidence intervals of the mean effect and standard deviation of the fluorescence phenotypes observed in a recent set of trans-regulatory mutants (Gruber et al. 2012, see Figure S2, A and B). The 90% confidence interval for selection coefficients was inferred from fitness assays performed on 8 mutants (see Materials and Methods and Figure S2C). In all analyses, only the indicated parameters were allowed to vary; all other experimentally controllable parameters were fixed at values ultimately used in our mapping experiment (sequencing depth = 100, population size = 10⁷, cutoff percent = 5%, generations = 20), and mutational parameters were fixed at values representative of the mutants used for mapping (mean effect = 5%, standard deviation = 100%, selection coefficient = 0.03).

Figure 2

Inherent properties of mutations affect statistical power to detect a difference in the frequency of a causal mutation between bulks. Power is shown for various mutation effects on mean (B, C), standard deviation (D, F), and relative fitness (G, H). Comparisons of hypothetical wild-type (red) and mutant (blue) populations with effects of a mutation on mean expression (A), standard deviation of expression (E), and relative fitness (I) are also shown. In all plots, the dashed line indicates 90% power. Gray shaded regions represent values of the mean effect, standard deviation, or selection coefficient of causal mutations observed in a recent set of expression mutants (see Figure S2). In all analyses, only the indicated parameters were allowed to vary; all others were fixed. These fixed values were: mean effect = 5%, standard deviation = 100%, selection coefficient = 0.03, sequencing depth = 100, population size = 107, cutoff percent = 5%, generations = 20.

Figure 3

Overview of experimental design for mapping small effect mutations affecting the expression of a fluorescent reporter in EMS-induced mutants. This approach is based on the isolation of a large number of random F₁ segregant haploid cells, followed by high-throughput phenotypic selection using FACS, and estimation of allele frequencies genome-wide using next generation sequencing. Note the selection of haploid MATα cells using expression of the RFP reporter linked to MATα locus that is indicated with a red dot. Quantitative differences in the level of YFP expression are indicated by differences in the intensity of yellow background.

Figure 4

Analysis of Illumina sequencing data. A set of high confidence variants was called using the somatic command in VarScan (dark gray), with reads from the mapping strain treated as “normal” data and reads from merged bulks treated as “tumor” data. Allele frequencies were then estimated for these sites in the low fluorescence and high fluorescence bulks with Popoolation2 (light gray). Differences between these two bulks were assessed using G-tests.

Figure 5

BSA-seq clearly identified a single causative site in three trans-regulatory mutants affecting fluorescence of a reporter gene. Significance of the difference in allele frequency between low fluorescence and high fluorescence bulks is shown as the negative of logarithm of P-value from G-test for mutants YPW89 (A), YPW94 (B), and YPW102 (C). Each bar shows significance for an individual EMS-induced mutation with its genomic position represented on x-axis. Roman numerals indicate each of the 16 S. cerevisiae chromosomes. Insets in (A) and (B) are magnifications of chromosomes harboring causative sites and show linked mutations with significant effects. Horizontal dotted lines represent a significance threshold of α = 0.001.

Figure 6

Single mutations identified by BSA-seq completely explain the mutant phenotypes. Mean expression level (measured as YFP fluorescence) for 8 replicates each of the wild-type genotype (WT); the YPW89, YPW94, and YPW102 mutants; and the three allele-replacement strains (“Single site”) for each mapped mutation (SSN2(Q971Stop), TUP1(G696D), and ROX1(R12K)) are shown. For each replicate, the median level of fluorescence of at least 5000 cells was quantified and expressed relative to mean fluorescence in the WT reference strain. Mutant genotypes and allele-replacement strains were compared to the WT strain using t-tests (***P < 0.001). In all three cases, the single site mutant was found to phenocopy the EMS mutant strain (P = 0.58 for SSN2, P = 0.23 for TUP1, and P = 0.44 for ROX1).