Unstable chromosome rearrangements in Staphylococcus aureus cause phenotype switching associated with persistent infections

Significance Staphylococcus aureus is a major human pathogen known to exhibit subpopulations of small-colony variants (SCVs) that cause persistent or recurrent infections. The underlying mechanisms promoting the SCV phenotypic switching and adaptation to persistent infection are poorly understood. Moreover, the instability of this frequently reverting phenotype hampers diagnosis and study. Here we show that SCVs with reduced virulence but increased immune evasion and persistence properties can arise from reversible chromosomal instability. This mechanism of SCV generation implies an asymmetric chromosome inversion and the activation of prophage-encoding genes used for immune evasion. Assessment of major S. aureus lineages indicates this genomic plasticity is a common but previously unrecognized mechanism used by S. aureus to cause persistent and relapsing infections.

and "non-invérsion" standard curvés génératéd from 10-fold serial dilution of gel purified amplicons corresponding to the inverted or non-inverted chromosomal loci. In order to control for potential PCR hybrid formed during PCR, a limit of detection was determined by using as DNA template a 1 to 1 ratio of a non-recombined hsdM1 and hsdM2 purified amplicons obtained from the WT strain. All qPCR experiments were performed in triplicate.

RNA-seq transcriptomic analysis
RPMI 1640 (Gibco) media were inoculated with overnight RPMI 1640 cultures of WT, NCV and SCV to a starting adjusted to OD600 of 0.02. Cultures were grown to mid-exponential phase (OD600 = 0.6) before 5 mL of each culture was mixed with 5 mL of RNAprotect bacteria reagent (Qiagen) and incubated for 10 min at room temperature. Bacterial cells were pelleted by centrifugation at 3300 xg for 20 min at room temp. Total RNA were extracted from the bacterial pellets using FastRNA pro blue kit (MP Biomedicals) using manufacturer's protocol with modifications. Cells were resuspended in 1 mL of RNA pro solution and lysed using Bertin Precellys 24 homogenizer set at 6000 rpm for 40 sec followed by phenol/ chloroform extraction. Samples were ethanol precipitated overnight at -20C with 0.3M sodium acetate. Samples were pelleted and dried at room temp before resuspending in DEPC treated water which were then treated with TURBO DNase (Ambion) followed by clean up with RNeasy kit (Qiagen). The absence of DNA contamination was checked by PCR and RNA integrity and purity was checked with a bioanalyzer RNA kit (Agilent).  (12). Differential gene expression analysis was performed with Degust (http://degust.erc.monash.edu/) using the four RNA-seq replicates of each strains.

Conservation of ΦSa3 and inverted repeats among S. aureus genomes.
A global dataset of 7099 S. aureus genomes previously compiled was used to estimate ΦSa3 conservation (2). The DNA sequence of the ΦSa3 integrase and the three IEC genes SCIN, CHIPS and staphylokinase genes were detected among the 7099 genome assemblies using abricate v0.8.11 (https://github.com/tseemann/abricate). Genomes where the integrase plus at least two genes of the IEC were detected were considered as having ΦSa3. The 29 fully assembled S. aureus genomes used to detect potential loci potentially promoting chromosomal inversion are listed in S Table S4. Inverted repeats at least 1000 bp and with more than 95% identity have been detected with a custom script using nucleotide blast (https://github.com/rguerillot/scripts/blob/master/inversion_scan.py).

Statistical analysis
Two-tailed Mann Whitney tests were performed with GraphPad Prism 7 (GraphPad Software).

Long read whole genome sequence data analysis
De novo assembly were performed using HGAP2 algorithm. To remove remaining SNP/Indel errors, the resulting fully assembled genomes were polished with the corresponding short read sequences using snippy and then annotated with prokka v 1.13.3 (7).

Whole human blood killing assay and human neutrophil chemotaxis assay
Wholé blood was colléctéd from héalthy voluntéérs into EDTA tubés. Wholé blood killing assays wéré conductéd as déscribéd in (10)