Comparative proteomic investigation of multiple methicillin-resistant Staphylococcus aureus strains generated through adaptive laboratory evolution

Abstract

Recent discoveries indicate that tolerance and resistance could rapidly evolve in bacterial populations under intermittent antibiotic treatment. In the present study, we applied antibiotic combinations in laboratory experiments to generate novel methicillin-resistant Staphylococcus aureus strains with distinct phenotypes (tolerance, resistance, and suppressed tolerance), and compared their proteome profiles to uncover the adaptation mechanisms. While the tolerant strains have very different proteomes than the susceptible ancestral strain, the resistant strain largely resembles the ancestral in terms of their proteomes. Our proteomics data and other assays support the connection between the detected mutations to the observed phenotypes, confirming the general understanding of tolerance and resistance mechanisms. While resistance directly counteracts the action mechanism of the antibiotic, tolerance involves complex substantial changes in the cells' biological process to achieve survival advantages. Overall, this study provides insights into the existence of diverse evolutionary pathways for tolerance and resistance development under different treatment scenarios.

Keywords: Evolutionary biology; Microbiology; Proteomics.

Graphical abstract

Figure 1

Evolution of tolerance and resistance on MRSA under different treatment schemes (A) Timeline of the evolution experiment on MRSA. In the first scheme, the culture was treated with DAP (10 μg/mL) for 2 h intermittently for two weeks. In the second scheme, the culture was treated with DAP (10 μg/mL) and RIF (1 μg/mL) combination for 2 h intermittently for two weeks. In the third scheme, the culture was treated with DAP (10 μg/mL) for 2 h intermittently for one week, and RIF (1 μg/mL) was added to the treatment regime in the second week. Population names are based on scheme (S#) and day after treatment (D#). (B) Schematic of the evolution experiment protocol. (C) Time-kill curve of ancestral MRSA and evolved populations after 3, 7, 10 and 14 days of treatment under scheme 1 (left), scheme 2 (middle) or scheme 3 (right) with DAP (10 μg/mL) (mean ±s.d., n = 5). (D and E) Relative MDK₉₉ (minimum duration for killing 99% of the population) for DAP (mean ±s.d., n = 5) (D), and MIC for DAP and RIF (mean ±s.d., n = 3) (E) of the population before and after 3, 7, 10 and 14 days of treatment under scheme 1, scheme 2 or scheme 3. The colored bars below the graph indicate the antibiotic treatment regime.

Figure 2

Characterization of the evolved strains (A) Growth profile of ancestral, S1D14, S2D14, and S3D14 populations in the absence and presence of DAP (0.25 μg/mL) (mean ±s.d., n = 3). (B) Schematic of single point mutations identified in the strains evolved from scheme 1, 2 and 3. CDS, coding sequence.c, Survival of the ancestral population, S1D7 and S2D7, and their RIF-resistant derivative (rpoBH481Y) (patterned fill) after 1 h of antibiotic treatments [DAP (10 μg/mL) and RIF (1 μg/mL)] (mean ±s.d., n = 3). p values for the pairwise comparison were estimated with two-tailed Student’s t-test with unequal variances of the log-transformed values.

Figure 3

Proteomic response of ancestral, S1D14, S2D14 and S3D14 strains after DAP treatment (A) Schematic of proteomics data analysis strategy. First, we compared the proteome profile of each strain after DAP treatment to those before treatment (as controls) to obtain the strain-specific antibiotic response. Then, we also compared the proteome profile of the evolved strains with the ancestral as our control, in the presence and absence of DAP, to pinpoint the commonalities and differences in protein expression profiles between the evolved strains and the ancestral strain, under both normal growth condition and during antibiotic exposure. (B) Venn diagrams for proteome comparison of ancestral, S1D14, S2D14, and S3D14 populations before and after 1 h of DAP treatment (0.25 μg/mL). (C) Volcano plots for ancestral, S1D14, S2D14, and S3D14 populations after 1 h of DAP treatment compared with those before treatment. Differentially expressed proteins (DEPs) are defined to be those with pvalues below 0.05, and absolute fold change greater than 1.5, corresponding to the rectangular regions. Left rectangular regions are the down-regulated proteins and right rectangular regions are the up-regulated proteins. (D)Heatmap of the DEPs across the ancestral, S1D14, S2D14, and S3D14 populations under DAP treatment compared to the untreated populations. Hierarchical clustering was performed using Euclidean distance and a ward linkage model. (E) Protein-protein interaction network of the DEPs of ancestral and evolved populations when treated with DAP compared to the untreated populations, as predicted by STRING v11.0. The lines represent protein interaction (thicker lines mean higher confidence), and the dots in different colors represent different protein functions. Nodes with black outlines are up-regulated proteins, and nodes with white outlines are down-regulated proteins. Uncharacterized proteins are not annotated and nodes without function enrichment are colored gray.

Figure 4

Proteome profile comparison between the evolved strains (S1D14, S2D14, and S3D14) and the ancestral strain (A and B) Volcano plots for S1D14, S2D14, and S3D14 populations compared to the ancestral in the absence of antibiotic (A) and after 1 h of DAP treatment (B). Differentially expressed proteins (DEPs) are defined to be those with pvalues below 0.05, and absolute fold change greater than 1.5, corresponding to the rectangular regions. Left rectangular regions are the down-regulated proteins (expression higher in the ancestral strain) and right rectangular regions are the up-regulated proteins (expression higher in the evolved strain). (C–G) Gene Ontology (GO) analysis and pathway enrichment study (KEGG) by DAVID of the DEPs of S1D14 compared to the ancestral (C), S2D14 compared to the ancestral (D) and S3D14 compared to the ancestral (E) after DAP treatment. Fold enrichment is defined as the ratio of the proportion of the input information to the background information. f, g, Venn diagrams of the DEPs comparison of S1D14, S2D14, and S3D14 populations (compared to ancestral) in the absence of antibiotic (F) and after 1 h of DAP treatment (G). (H) Overlapped differentially expressed proteins among the evolved strains in the absence (top) and presence (bottom) of daptomycin.

Figure 5

Biofilm assay of the ancestral strain and the evolved strains (A) Biofilm cells formation of the ancestral and evolved strains after 24 h of growth, measured by MTT assay through the OD₅₇₀ value (mean ±s.d.,$n$ = 12). (B and C) Antimicrobial activities of DAP and VAN against WT ancestral MRSA, represented by the minimum concentration needed for inhibiting 90% biofilm formation (minimum biofilm inhibitory concentration, MBIC) (B) and minimum concentration needed for eradicating 50% mature biofilms (minimum biofilm eradication concentration, MBEC) (C) ($n$ = 4). The MBIC₉₀ for both DAP and VAN toward ancestral MRSA is 1.25–2.5 μg/mL, while the MBEC₅₀ for both DAP and VAN toward ancestral MRSA is 10–20 μg/mL, corresponding to the highlighted area. d, e, Biofilm inhibiting activity of DAP (2.5 μg/mL) (D) and biofilm eradication activity of DAP (20 μg/mL) (E) on the ancestral and evolved strains (mean ±s.d., n = 4). (F and G) Biofilm inhibiting activity of VAN (2.5 μg/mL) (F) and biofilm eradication activity of VAN (20 μg/mL) (G) on the ancestral and evolved strains (mean ±s.d., n = 4). Significance of difference with the ancestral: ns, not significant, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 (two-tailed $t$-test with unequal variances).

Figure 6

The expression level of the genes that are mutated in the evolved strains (A–C) Relative abundance of the proteins phosphatidylglycerol lysyltransferase (MprF) (A), CDP-diacylglycerol-glycerol-3-phosphate 3-phosphatidyltransferase (PgsA) (B) and phosphate acetyltransferase (Pta) (C) across the ancestral and evolved strains, measured by label-free quantitative proteomics using spectral counting, where the yaxis is the normalized spectral abundance factor (NSAF) values (mean ± s.d., n = 3). The horizontal dashed line shows the mean expression level of the ancestral strain (without DAP treatment). Asterisks indicate zero NSAF value (not detected by the mass spectrometer). d, e, Fold changes of the gene expression of pgsA (D) and pta (E) across the evolved strains as compared to the ancestral, measured by RT-qPCR (mean ± s.d., n = 3).

Figure 7

Evolved strains possessed cell membrane and other cell wall-related modifications (A and B) Induced autolysis assay in the ancestral, S1D14, S2D14, and S3D14 strains. Cells were treated with 0.05% Triton X-100 (a) or 1000 ng/mL lysostaphin (B) and incubated at 37°C. Autolysis was measured by monitoring the decrease in OD₆₀₀ over time (mean ± s.d., n = 3). (C) Killing assay of the WT ancestral population and DAP-tolerant S1D14 evolved strain with VAN (30 μg/mL) (mean ± s.d., n = 3). (D) MIC test toward VAN carried out using disc diffusion antibiotic sensitivity testing. The text on the lower right corner marks the diameter of the zone of inhibitions.