Variation in Filamentous Growth and Response to Quorum-Sensing Compounds in Environmental Isolates of Saccharomyces cerevisiae

Abstract

In fungi, filamentous growth is a major developmental transition that occurs in response to environmental cues. In diploid Saccharomyces cerevisiae, it is known as pseudohyphal growth and presumed to be a foraging mechanism. Rather than unicellular growth, multicellular filaments composed of elongated, attached cells spread over and into surfaces. This morphogenetic switch can be induced through quorum sensing with the aromatic alcohols phenylethanol and tryptophol. Most research investigating pseudohyphal growth has been conducted in a single lab background, Σ1278b. To investigate the natural variation in this phenotype and its induction, we assayed the diverse 100-genomes collection of environmental isolates. Using computational image analysis, we quantified the production of pseudohyphae and observed a large amount of variation. Population origin was significantly associated with pseudohyphal growth, with the West African population having the most. Surprisingly, most strains showed little or no response to exogenous phenylethanol or tryptophol. We also investigated the amount of natural genetic variation in pseudohyphal growth using a mapping population derived from a highly-heterozygous clinical isolate that contained as much phenotypic variation as the environmental panel. A bulk-segregant analysis uncovered five major peaks with candidate loci that have been implicated in the Σ1278b background. Our results indicate that the filamentous growth response is a generalized, highly variable phenotype in natural populations, while response to quorum sensing molecules is surprisingly rare. These findings highlight the importance of coupling studies in tractable lab strains with natural isolates in order to understand the relevance and distribution of well-studied traits.

Keywords: filamentous growth; invasive growth; phenylethanol; pseudohyphal growth; tryptophol.

Figure 1

Pseudohyphal growth. Images depict: a small colony with pseudohyphae surrounding it (A), a close-up of a colony perimeter (B), and an image of the same perimeter in a different focal plane showing the pseudohyphae growing into the agar (C). Scale bar represents 200 μm. To obtain images, strain YJM1439 was streaked on a SLAD plate and grown for 6 days.

Figure 2

Population structure of the 100-genomes panel supplemented with strain YJM311, as inferred by the program structure. Following the analysis of Strope et al. (2015), each vertical line represents an individual strain with its fractional ancestry of K = 6 subpopulations represented by colors: green (North American), orange (Malaysian), red (West African), purple (Sake), blue (European/wine), and gray (human associated). Strains were assigned membership based on a threshold of >60% ancestry in a subpopulation, except for mosaic strains which had less than 60% ancestry in any other subpopulation.

Figure 3

Image processing pipeline. First column: three sample colonies from an Omnitray, which was blotted with 1.58 mm pins; in order, strains YJM984, YJM1336, and YJM1341, derived from 96-112, a clinical strain, M28s2, a European wine strain, and NRRL Y-12637, a South African wine strain, respectively. Second column: original images processed to differentiate white ring, filamentous growth, and background. Third column: inner part of the colony separated and pseudohyphal pixels counted to generate the filamentous index listed on the right.

Figure 4

Psh for the 100-genomes panel and the YJM311 mapping population. (A) and (C) plot average filamentous index for individual strains or segregants (+/− 2 s.e.m.), respectively, and were ordered based on their filamentous index. Panel A also contains the means for the subpopulations; points not connected by the same letter are significantly different. Panels (B) and (D) represent population distributions. In (A-B), black is control, red is PheOH treatment, and blue is tryptophol treatment. In (C-D), green is the high pool and orange is the low pool.

Figure 5

Genome-wide comparison of the allele frequencies in the high and low psh pools of YJM311 F5 segregants. The G-prime statistic was calculated with a sliding window size of 60,000 bp (A) and 20,000 bp (B). Red line represents the cut-off for significance at a false discovery rate of 0.01. Candidate loci are listed above major, significant peaks.