Figures and data
Experimental evolution design.
A) Growth curves and B) total growth of ancestral MRSA and MSSA populations in vitro with and without sub-MIC oxacillin. Different letters indicate significant differences (
Evolution of virulence is facilitated by exposure to both host and sub-MIC antibiotic.
A) Virulence in terms of C. elegans mortality. Dashed and dotted lines indicate respective ancestral virulence. Shaded areas indicate standard errors of technical replicates of ancestral virulence. Different letters indicate significant differences (
Host and sub-MIC antibiotic selection facilitated pathogen growth in antibiotics.
Pathogen in vitro growth A) without oxacillin (
Regulatory genes likely played an important role in pathogen adaptation.
A) Mutations swept to fixation, excluding intergenic and synonymous mutations, grouped by general function. The size of each point indicates how many populations had acquired at least one mutation in the gene. Colored shapes next to genes indicate whether these genes are regulatory or have been implicated in virulence or antibiotic resistance in the literature (see Table S1). B) In vitro growth of populations with SCCmec and ACME deletions with or without sub-MIC oxacillin (one-sample t-test t = 2.73, df = 8, P = 0.026). C) Change from ancestral pathogen-induced mortality vs. mutations (codY one-sample t-test = 19.56, df = 7, P < 0.001; gdpP: 17.02, df = 3, P = 0.002; pbpA: 7.54, df = 3, P = 0.012). D) Hemolysis status vs. mutations (Fisher’s exact test P < 0.001; agr vs. alr: P = 0.036; argR: P = 0.006; codY: P = 0.005; gdpP: P = 0.006; purR: P = 0.006). E) Oxacillin MIC vs. mutations (Fisher’s exact test P = 0.0015; codY vs. brnQ1: P = 0.043). F) Biofilm production of evolved populations. Different letters indicate significant differences (MRSA: 
Mutations that arose during experiment are enriched in isolates associated with systemic infection in humans.
A) Number of genomes in our public S. aureus genome dataset (see Methods) grouped by isolation source. General host-association refers to descriptions not specific enough to assign to other categories. We excluded samples that were ambiguously specified or missing information. B) to H) Number of genomes in the database containing mutations in the respective gene. I) Number of genomes in the dataset containing the mutations arisen in agr. This gene is separate from the others to facilitate ease of visualization due to the magnitude of the y-axis. Asterisks indicate significant difference in the proportion of blood/systemic-associated genomes compared to the expected distribution in the dataset. *P < 0.05, **P < 0.01, ***P < 0.001
MRSA and MSSA underwent distinct evolutionary trajectories.
Principal component analysis of traits evolved from A) MRSA ancestors (PERMANOVA F(3,23) = 8.38, r2 = 0.557, P = 0.001) and B) MSSA ancestors (PERMANOVA F(3,23) = 2.20, r2 = 0.248, P = 0.064). Phylogenetic tree constructed from frequencies of mutations in populations evolving from C) MRSA ancestors (distance to ancestor: F(3,20) = 2.39, P: 0.099; distance between populations: Kruskal-Wallis 
Most colonies sampled matched their respective population in terms of the ability to hemolyze sheep’s blood (i.e., over 50% of colonies having the same hemolysis status as the population they were sampled from).
Zone of inhibition from Kirby-Bauer disk diffusion susceptibility test with oxacillin for colonies sampled from each of the four populations with the most number of mutations (two from MRSA and two from MSSA), and two additional randomly selected populations.
Count of all mutations (from 10-100% frequency) arisen in evolved populations.
Mutations at frequencies between 0.1 and < 1, excluding synonymous or intergenic mutations, in each evolved population.
Points with darker shades indicate more than one mutation present. All mutations in MRSA +host +ox populations between 0.1 and < 1 are all either synonymous or intergenic. A table of all mutations and frequencies is available on figshare (10.6084/m9.figshare.28745558)
A) Count and B) Mean of mutations swept to fixation in evolved populations.
Genetic distance between the ancestor and evolved populations for A) MRSA and B) MSSA genotypes. Genetic distance between replicate populations within each treatment for C) MRSA and D) MSSA genotypes. Different letters indicate significant differences. Error bars indicate standard errors.
Nonsynonymous mutations occurred in genes with known roles in virulence and antibiotic resistance.
Fixed mutations in genes and intergenic regions appearing in more than two populations.
SYN = synonymous, NONSYN = nonsynonymous
Chi-square goodness-of-fit test results for comparison between the proportion of blood/systemic infection-associated mutations vs. skin/nose/throat-associated mutations against the expected proportion (i.e., all BioSamples in the dataset).
All degrees of freedom equalled 1.
