Proteomic insights into survival strategies of Escherichia coli in perchlorate-rich Martian brines

Abstract

Brines, potentially formed by the deliquescence and freezing point depression of highly hygroscopic salts, such as perchlorates (ClO₄^-), may allow for the spatial and temporal stability of liquid water on present-day Mars. It is therefore of great interest to explore the microbial habitability of Martian brines, for which our current understanding is, however, still limited. Putative microbes growing in the perchlorate-rich Martian regolith may be harmed due to the induction of various stressors including osmotic, chaotropic, and oxidative stress. We adapted the model organism Escherichia coli to increasing sodium perchlorate concentrations and used a proteomic approach to characterize the adaptive phenotype. Separately, the microbe was adapted to elevated concentrations of sodium chloride and glycerol, which enabled us to distinguish perchlorate-specific adaptation mechanisms from those in response to osmotic, ion and water activity stress. We found that the perchlorate-specific stress response focused on pathways alleviating damage to nucleic acids, presumably caused by increased chaotropic and/or oxidative stress. The significant enrichments that have been found include DNA repair, RNA methylation and de novo inosine monophosphate (IMP) biosynthesis. Our study provides insights into the adaptive mechanisms necessary for microorganisms to survive under perchlorate stress, with implications for understanding the habitability of Martian brines.

Keywords: Adaptation; Brines; Habitability; Mars; Perchlorate; Proteomics.

Fig. 1

Schematic illustration of the experimental workflow. (a) E. coli was incrementally adapted to increasing concentrations of NaClO₄, NaCl and glycerol. (b) Example growth profiles of wild-type (WT) E. coli (grey) and after its adaptation and subsequent propagation at concentrations of 0.2 mol/kg NaClO₄ (blue), 1.2 mol/kg NaCl (green) and 4 mol/kg glycerol (pink). Data represent means ± SD (n = 3). (c) The adaptive phenotypes were analyzed using proteomics. First, triplicates of each treatment and the WT were prepared, harvested during late exponential growth and analyzed by mass spectrometry. Next, we aimed to identify those adaptation mechanisms, which were specific for perchlorate-induced stress in E. coli. For this, a post hoc test (FDR ≤ 0.01) was performed, giving information on which treatment combinations significantly differ from each other for each protein. Sets of proteins for solute-specific and shared adaptation mechanisms were extracted using a strict filtering process and graphically depicted as Venn diagrams (see Materials and Methods). Lastly, the protein sets were used to identify significant adaptive pathways (FDR ≤ 0.05) using protein-protein interaction networks with the help of the STRING database.

Fig. 2

Overview of proteomic phenotypes after adaptation of E. coli to 0.2 mol/kg NaClO₄, 1.2 mol/kg NaCl and 4.0 mol/kg glycerol. (a) Principal component analysis (PCA) depicting 72.8% of the total variance in the data. (b) Fractions of significantly regulated proteins of all proteins that were detected by LC-MS in each treatment. Significant regulation relative to the control was determined by ANOVA (FDR ≤ 0.01) and post hoc testing (FDR ≤ 0.01) (c) Hierarchical clustering of proteins passing both statistical tests. The protein abundances were log2-transformed and z-normalized, with downregulated proteins colored blue and upregulated ones colored red. The heatmap and PCA were generated by the Perseus software.

Fig. 3

Proteomic adaptation mechanisms of E. coli to NaClO₄ stress. (a) Sets of proteins that are part of solute-specific and shared proteomic adaptive responses were extracted using a strict filtering process (see Materials and Methods) and visualized using Venn diagrams for up- and downregulated proteins. Each protein set was then fed into the STRING database (Version 11.5), which was used to identify significant adaptive pathways (FDR ≤ 0.05) of protein clusters. The corresponding cluster annotations are listed for each protein set, with a plus indicating overlapping cluster annotations. For many clusters, several enrichments were found, which, however, were very general and can be found in Supplementary Table S2-S4. (b) Significantly enriched pathways of the NaClO₄-specific upregulated (left) and downregulated (right) stress response as identified using the STRING database (Version 11.5). The protein clusters are outlined, and the significant pathways are annotated, with unique colors representing each enrichment.