Transcriptional response of Saccharomyces cerevisiae to lactic acid enantiomers

Abstract

The model yeast, Saccharomyces cerevisiae, is a popular object for both fundamental and applied research, including the development of biosensors and industrial production of pharmaceutical compounds. However, despite multiple studies exploring S. cerevisiae transcriptional response to various substances, this response is unknown for some substances produced in yeast, such as D-lactic acid (DLA). Here, we explore the transcriptional response of the BY4742 strain to a wide range of DLA concentrations (from 0.05 to 45 mM), and compare it to the response to 45 mM L-lactic acid (LLA). We recorded a response to 5 and 45 mM DLA (125 and 113 differentially expressed genes (DEGs), respectively; > 50% shared) and a less pronounced response to 45 mM LLA (63 DEGs; > 30% shared with at least one DLA treatment). Our data did not reveal natural yeast promoters quantitatively sensing DLA but provide the first description of the transcriptome-wide response to DLA and enrich our understanding of the LLA response. Some DLA-activated genes were indeed related to lactate metabolism, as well as iron uptake and cell wall structure. Additional analyses showed that at least some of these genes were activated only by acidic form of DLA but not its salt, revealing the role of pH. The list of LLA-responsive genes was similar to those published previously and also included iron uptake and cell wall genes, as well as genes responding to other weak acids. These data might be instrumental for optimization of lactate production in yeast and yeast co-cultivation with lactic acid bacteria. KEY POINTS: • We present the first dataset on yeast transcriptional response to DLA. • Differential gene expression was correlated with yeast growth inhibition. • The transcriptome response to DLA was richer in comparison to LLA.

Keywords: Budding yeast; D-lactate; L-lactate; Lactic acid; Low pH stress; RNA-seq.

Fig. 1

Overview of gene expression changes in all comparisons. a–f show volcano plots for major comparisons with differentially expressed genes (absolute log₂ fold change > 1 and adjusted p-value < 0.05) are indicated by red dots (each dot corresponds to one gene). g, h show intersections of the lists of genes upregulated and downregulated, respectively, in response to different treatments, represented as Venn diagrams. Full expression data are available in Supplemental Table S3 and from the GEO database (GSE231937)

Fig. 2

Overview of transcriptional response to DLA. Shown are a logarithmic (base 2) normalized expression counts of the genes DE in response to 45 mM DLA and annotated with the GO terms “lactate metabolic process” or “lactate biosynthesis process” (ORF (open reading frame) symbols and gene names shown in the titles of the panels), as well as the correlation between expression changes in response to 5/45 mM DLA of shared DEGs (b) and all genes (c). Pearson’s correlation coefficient = 0.99 for (b) and 0.72 for (c)

Fig. 3

Comparison of transcriptional responses to LLA and DLA shows significant similarities and reveals several candidate qualitative DLA sensor genes. Correlation between expression changes in response to 45 mM DLA or LLA of shared DEGs (a) and all genes (b). Pearson’s correlation coefficient = 0.98 for a and 0.65 for b. c Clustering of expression profiles does not reveal any groups with quantitative response to DLA. d Expression profiles of potential qualitative sensors, genes responding to DLA but not LLA. The vertical axis shows logarithmic (base 2) normalized expression counts of the genes

Fig. 4

Addition of 45 mM sodium D-lactate compensates for the growth defect caused by 45 mM DLA treatment (a) and does not trigger changes in the expression of selective DLA-responsive genes (b). a Relative growth rate (the difference between logarithmic (base 2) final OD600 and initial OD600 divided over incubation time in hours) of cultures incubated for 3 h in the media with 45 mM DLA or DLS. *p < 0.05; ns, not significant (paired pairwise Wilcoxon rank sum test). b Expression levels of the AQR1, DLD3, FIT2, and YPS3 genes relative to the geometric mean of the reference genes ACT1 and CDC19 measured with quantitative PCR. Raw qPCR data are presented in Supplemental Table S5