- Open Access
Epistasis analysis uncovers hidden antibiotic resistance-associated fitness costs hampering the evolution of MRSA
- Maho Yokoyama1,
- Emily Stevens2,
- Maisem Laabei3,
- Leann Bacon2,
- Kate Heesom4,
- Sion Bayliss1,
- Nicola Ooi5,
- Alex J. O’Neill5,
- Ewan Murray6,
- Paul Williams6,
- Anneke Lubben7,
- Shaun Reeksting7,
- Guillaume Meric1,
- Ben Pascoe1,
- Samuel K. Sheppard1,
- Mario Recker8,
- Laurence D. Hurst1 and
- Ruth C. Massey2Email authorView ORCID ID profile
© The Author(s). 2018
- Received: 4 May 2018
- Accepted: 25 June 2018
- Published: 18 July 2018
Fitness costs imposed on bacteria by antibiotic resistance mechanisms are believed to hamper their dissemination. The scale of these costs is highly variable. Some, including resistance of Staphylococcus aureus to the clinically important antibiotic mupirocin, have been reported as being cost-free, which suggests that there are few barriers preventing their global spread. However, this is not supported by surveillance data in healthy communities, which indicate that this resistance mechanism is relatively unsuccessful.
Epistasis analysis on two collections of MRSA provides an explanation for this discord, where the mupirocin resistance-conferring mutation of the ileS gene appears to affect the levels of toxins produced by S. aureus when combined with specific polymorphisms at other loci. Proteomic analysis demonstrates that the activity of the secretory apparatus of the PSM family of toxins is affected by mupirocin resistance. As an energetically costly activity, this reduction in toxicity masks the fitness costs associated with this resistance mutation, a cost that becomes apparent when toxin production becomes necessary. This hidden fitness cost provides a likely explanation for why this mupirocin-resistance mechanism is not more prevalent, given the widespread use of this antibiotic.
With dwindling pools of antibiotics available for use, information on the fitness consequences of the acquisition of resistance may need to be considered when designing antibiotic prescribing policies. However, this study suggests there are levels of depth that we do not understand, and that holistic, surveillance and functional genomics approaches are required to gain this crucial information.
- Mupirocin resistance
- Fitness costs
Antibiotic resistance, which can involve either the mutation of existing genes or the acquisition of genes encoding a resistance mechanism , can evolve in many ways and frequently incurs a fitness cost to the organism . As antibiotics are most commonly used for short and defined periods of time, resistant bacteria are under selection to reduce these costs to avoid displacement once treatment has finished. In many cases this is achieved through compensatory mutations that allow many resistance mechanisms to be stably maintained in populations for long periods of time . However, some antibiotic resistance mechanisms have been reported to incur no detectable fitness costs [4–7], which suggests there should be no barriers to their widespread dissemination.
Staphylococcus aureus is an example of a major human pathogen  that has become more challenging to treat due to the emergence of antibiotic resistance, with methicillin-resistant S. aureus (MRSA) being the most notable example . S. aureus resides asymptomatically as part of the normal nasal flora of up to 50% of humans ; however, this carriage is a significant risk factor for infection  to the extent that carriers are often decolonised using antibiotics such as mupirocin prior to invasive procedures, including surgery and dialysis . Mupirocin is a polyketide antibiotic that is applied as an ointment to eradicate nasal carriage of MRSA in patients at risk of infection . Such decolonisation has been reported to reduce S. aureus infections of post-surgical wounds by 58%, of haemodialysis patients by 80% and of peritoneal dialysis patients by 63% .
The molecular target for mupirocin is the bacterial isoleucyl-tRNA synthetase (IleRS), which charges tRNAs with the amino acid isoleucine (Ile) . By binding to this enzyme the antibiotic halts protein synthesis, and in doing so inhibits bacterial growth . As a consequence of the widespread use of mupirocin, resistance has emerged where the bacteria have mutated the gene encoding IleRS, ileS, resulting in an amino acid substitution (e.g. V588F, encoded by a G to T single nucleotide polymorphism (SNP) at position 1762 in the ileS gene (G1762 T)) that alters the structure of the protein’s active site without removing its functionality, rendering mupirocin less effective . This confers a low to intermediate level of resistance on the antibiotic . Alternatively, the bacteria may acquire an alternative IleRS, encoded by a mupA or mupB gene, on a plasmid, which confers a higher level of resistance to mupirocin [18, 19].
The prevalence of mupirocin resistance varies widely, with the highest rates associated with patient groups repeatedly exposed to the antibiotic [20, 21]. It has also been shown that in countries where restrictions were put in place limiting the use of this antibiotic, such as New Zealand and Australia, the prevalence of resistant strains significantly declined . This suggests that antibiotic exposure is required to maintain selection of this resistance mechanism within a population, despite it being reported as incurring no fitness cost.
The ileS gene is highly conserved across the thousands of sequenced S. aureus isolates, and many failed attempts to inactivate it suggest its activity is essential to the bacteria. It is therefore surprising that the mutation that confers mupirocin resistance, by altering the structure of the encoded protein, does not appear to affect fitness . Particularly given that the replacement of valine 588 with phenylalanine (V588F), a much bulkier residue, is likely to fill and distort the Rossman fold of the enzyme, which is responsible for its ATP binding activity [4, 16]. However, in a recent genome-wide association study (GWAS) on the major hospital acquired MRSA clone, ST239, this mutation was significantly associated with differences in the virulence of S. aureus isolates . Its effect on toxin secretion, a major aspect of S. aureus virulence, was believed to result from epistatic interactions between ileS and other polymorphic loci. Upon analysis of a second genetically and geographically distinct collection of clinical isolates of the USA300 lineage, we detected this epistasis signal again. Here we characterise the effect this mutation has on both toxin production and bacterial fitness and in doing so uncover a potential explanation for the lack of success of these resistant strains.
Validation of the epistatic association between mupirocin resistance and toxicity
The epistatic signal between the mupR mutation and other loci appears to be indirect
Epistatic signals can emerge from population-based GWAS analyses for several reasons. It might signal that direct interactions are occurring between two loci, such that together they affect a phenotype, for example a transcriptional regulator and the region of DNA it binds to. SNPs can arise in the DNA binding region that affect how the regulator binds to it, or the regulator might acquire SNPs that affect its ability to bind the DNA. As these loci evolve and mutate an epistatic signal associated with the polymorphisms and the phenotype they affect could emerge. Alternatively, the epistatic effect may be a less direct lineage-specific effect, where a polymorphism can have an effect on a phenotype in one genetic background but not in another. In this case the polymorphisms that are associated epistatically with the primary SNP of interest may simply reflect the isolate’s genetic background. To examine whether the mupR-conferring SNP has a direct or indirect epistatic effect, we compared the list of loci it was epistatically paired with from the ST239 and the USA300 collections (Additional file 1: Table S1). Other than the mupR-conferring SNP, no other loci were in common between the two collections, suggesting that the epistatic effect may be indirect.
To examine whether the mupR-conferring epistatic effect was dependent upon a strain’s genetic background, we introduced it into three distinct S. aureus strains that all belonged to the ST8 clonal background (as determined by MLST: SH1000, RN6390B and USFL34). The presence of the V588F-conferring mutation and the mupR phenotype were confirmed; however, it was only in the SH1000 background that any effect on toxicity was observed, where the mupirocin-resistant strain, MY40, lysed only 71% of the cells (cultured THP-1 cells) compared to the wild-type mupirocin-sensitive SH1000 strain, which lysed 87% of the cells (n = 6, two-tailed t-test; p = 0.023; the effect of this mutation was further confirmed in four more independently selected mupirocin-resistant isolates of the SH1000 background (Additional file 1: Figure S1). By comparison, this mupR mutation had no effect on toxicity in either the RN6390B or USFL34 backgrounds (n = 6; two-tailed t-test p = 0.59 and 0.12, respectively). Together these data suggest that the mupR-associated epistatic effect observed here is likely to be indirect and its effect on toxicity limited to specific genetic backgrounds.
The effect of mupR on toxicity is independent of the Agr quorum sensing system
Given the effect on toxicity the mupR-conferring SNP has in the SH1000 S. aureus strain, we quantified the levels of production of some of the major cytolytic toxins produced by S. aureus: alpha toxin and the phenol soluble modulins (PSMs). For this, we first confirmed that no other toxicity-affecting mutations had arisen during our selection of the MY40 (SH1000 mupR) strain by comparing the genome sequences of the SH1000 and MY40 strains. The only non-synonymous SNP difference found between these strains was that conferring the V588F change in IleRS, although a small number of synonomous SNP differences were also detected (Additional file 1: Table S2).
Given the dominant effect the Agr quorum regulatory system has on toxin production, we hypothesised that its activity may be affected in the mupR mutant. To test this we quantified RNAIII expression levels, an effector molecule of the Agr system; no difference in expression was detected (n = 6, two-tailed t-test, p = 0.49). We also quantified the response of the mupR and mupS strains to exogenous AIP, the autoinducing pheromone of the Agr system, and we quantified how much AIP the mupR and mupS strains produced. However, we could detect no difference in Agr activity between theses strains by any of these assays (Additional file 1: Figure S2), suggesting that the effect of mupR on toxicity is independent of the Agr system.
Mupirocin resistance exerts a pleiotropic effect on the S. aureus proteome
The protein synthesis-disrupting effect of mupirocin has been used to induce and study the stringent response in bacteria . Exposure to mupirocin has also been shown to affect the expression of enzymes involved in the biosynthesis of other branched chain amino acids (BCAA) and the CodY virulence regulon . As the activity of the Agr system does not appear to be affected by mupR, we hypothesised that this mutation may have a similar effect on S. aureus that exposure to mupirocin has, which might explain its effect on toxicity. To examine this we adopted a proteomic approach using tandem-mass-tagging (TMT) protein mass spectroscopy  on whole cell lysates. As toxin production by S. aureus is growth dependent, we harvested proteins during the stationary phase of growth, and where the effect of exposure to mupirocin was being sought, we used a concentration of mupirocin that was below the minimum inhibitory concentration (MIC) to avoid any confounding effects of growth repression. The proteins were therefore extracted from triplicate 18-h cultures in tryptone soy broth (TSB), where we used a concentration of 10 ng/ml mupirocin, which we found to be the highest concentrations of mupirocin for which we found no inhibition of growth of the mupS SH1000 strain.
Of the 3026 open reading frames predicted for the NCTC8325 chromosome (which is the closest reference genome to SH1000), we detected and quantified 1284 proteins, and we used a cut-off of a minimum of a 50% difference in protein abundance to identify differentially produced proteins. When we compared the proteomes of SH1000 and MY40, 135 proteins were differentially produced (Additional file 2: Table S3), suggesting that this resistance mutation has pleiotropic effects on protein production. Exposure of the wild-type SH1000 strain to mupirocin resulted in differential production of 60 proteins when compared to the untreated SH1000 (Additional file 3: Table S4), but the production of only eight proteins was affected both by mupirocin exposure and in the mupR mutant (indicated in Additional file 2: Table S3 and Additional file 3: Table S4), suggesting that while there are some common downstream effects of interfering with the activity of IleRS through either mutation or antibiotic exposure, there are many differences in the outcome.
Interestingly, for both the mupR mutation and exposure to mupirocin we detected an upregulation of IleRS production, suggesting that the bacteria may be able to partially compensate for the interference in the activity of this enzyme by upregulating its production. In relation to previously reported effects on BCAA enzyme production, induction of the stringent response and effects on the CodY virulence regulon, the only effect we observed here was on BCAA enzyme production upon exposure to mupirocin (Additional file 3: Table S4). We believe the differences we report here compared to these previous studies [25, 26] are due to our use of sub-inhibitory concentrations of mupirocin and extracting proteins after 18 h, compared to the methodology used in the other studies where 3× and 5× MICs of mupirocin were used and the protein extracted within an hour of this exposure.
Although we detected no effect on the CodY regulon we did observe a difference between the wild type and mupR mutant in the abundance of the cytoplasmic response regulator of the Agr toxicity regulating system, AgrA, where the mupR mutant produced, on average, sevenfold higher levels of this protein (Additional file 2: Table S3). The agrA and agrC genes are co-transcribed, but there was no difference in AgrC protein abundance across the protein preparations, which suggests that the effect on AgrA abundance must be post-transcriptional. The AgrA protein needs to be phosphorylated to become transcriptionally active, and as such we would expect to see increased transcription of the RNAIII effector molecule of the Agr system were it active. As mentioned above, however, we detected no difference in RNAIII expression, which suggests that although AgrA may be more abundant in the mutant, it does not seem to be transcriptionally active.
Mupirocin resistance affects the activity of the PSM secretory apparatus, Pmt
Of the toxins encoded on the SH1000 genome, there was significantly more of both delta toxin and PSMα1 in the lysate of the mupirocin-resistant mutant (35- and 21-fold respectively; Additional file 2: Table S3). These are two of the most abundantly produced members of the PSM family of cytolytic toxins . This result was surprising given that our quantification of the levels of the PSMs in supernatant demonstrated a significantly lower production by the mupR mutant compared to the wild-type strain.
As the PSMs are more abundant in the intracellular environment of the mupR mutant, but less abundant in the extracellular environment, we hypothesised that PSM secretion may be affected in the mutant. This is facilitated by the activity of the PSM export system, Pmt , and although the pmt genes have not been annotated on the NCTC8325 genome, they are present (locus tags SAOUHSC_02155, SAOUHSC_02154, SAOUHSC_02153, SAOUHSC_02152, SAOUHSC_02151). This locus encodes a regulatory protein, two membrane bound ABC transporter proteins and two ATP binding proteins that fuel the export system, where the four structural genes are transcribed from a single promoter. Each ATP binding protein interacts with its paired transporter protein and the maintained selection of the 1:1:1:1 stoichiometry suggests it is critical to the exporters activity. In our mupR mutant we found that there was more than twice as much of one of the two ATP binding proteins (SAOUHSC_02152, gene name pmtC) compared with the wild-type strain. We hypothesised that interference with the stoichiometry of the proteins in this export system may affect its activity, which might explain why we find more PSM inside the cell but less outside. To test this we cloned and expressed the pmtC gene, from an inducible promoter in the SH1000 wild-type background, and found that when this ATPase was overexpressed it secreted significantly less PSMs into the extracellular environment (2.8-fold difference, n = 6, two-tailed t-test, p = 0.003; a representative image of this is presented in Fig. 2c). This suggests that the effect mupirocin resistance has on S. aureus toxicity is at least partially mediated by interfering with the activity of the PSM secretory apparatus.
It is interesting to consider that the complete inactivation of the Pmt system has been shown to be lethal to the bacteria, presumably as a result of the damage the PSMs can cause to internal membrane structures . Here we demonstrate a partial blocking of the Pmt system by mupirocin resistance, and some accumulation of PSMs internally; however, as we see no in vitro growth defects associated with this it must be a sub-toxic level.
Reducing the production of toxins alleviates the fitness cost of mupR
Toxin production by bacteria is an energetically costly activity [23, 30, 31]. Although we see a significant reduction in alpha toxin production by the mupR mutant, for the PSMs we see an increase in intracellular abundance but a decrease in the extracellular abundance. This makes it difficult to speculate on what fitness consequences there might be as a result of these difference. To clarify this, we compared the total amount (both intra- and extracellular) of PSMs being produced by the wild-type and mupR mutant by pooling whole cell lysates with supernatant prior to performing butanol extractions. Despite there being more PSMs intracellularly in the mutant, this represents such a small proportion of the total amount of PSMs produced by the bacteria that it had no effect on the relative abundance of the PSMs (Fig. 2d), demonstrating that the mupR mutant does produce significantly less of two of the major toxin groups produced by S. aureus.
As the ability to produce toxins is selected for in environments such as the skin and nose [23, 33], we hypothesised that, in such an environment, the costs associated with this resistance mechanism should become apparent. To test this we developed a growth medium with very low levels of nutrients (0.1× TSB) but added 5% horse blood such that strains that can produce toxins can release and utilise further nutrients from these blood cells for growth. To avoid the less toxic cells benefiting from nutrients released by the more toxic bacteria we grew the wild-type mupS SH1000 strain, an Agr mutant in the SH1000 background as a control, and MY40, the SH1000 mupR mutant, separately and quantified their relative fitness (Malthusians parameters) at 3, 6 and 24 h in this medium. After 3 h of incubation in the medium there was no detectable growth of any of the strains; however, by 6 h all three strains had grown, although the Agr mutant and the mupR strain grew faster than the SH1000 strain (Fig. 3b; n = 12, Mann-Whitney U test, p = 0.003 for the Agr mutant and 0.002 for the mupR mutant). This suggests that the bacteria enter a long lag phase in this medium, but once adapted (after 3 h) sufficient nutrients are available to sustain some growth. At this time point the SH1000 strain is at an apparent disadvantage, possibly by expending energy on producing more toxins than the Agr and mupR mutants, although we would not typically expect S. aureus to produce toxins this early in its growth phase.
After the 6-h time point there was a distinct shift in the relative fitness of the strains, with the SH1000 strain growing faster than the Agr mutant and the mupR strain (Fig. 3b; n = 12, Mann-Whitney U test, p = 0.018 and 0.01, respectively). The increased relative growth rate of the wild-type SH1000 strain could be explained by its increased capability to lyse cells and release nutrients necessary for its growth. To verify this, at the end of the growth assay we removed the bacteria and intact blood cells by centrifugation, and quantified by spectroscopy the levels of haem released from the cells. Significantly more haem was present in the supernatant of the SH1000 strain (Fig. 3c; n = 12, Mann-Whitney U test, p < 0.001 in both cases), demonstrating that SH1000 has released more cellular material into the broth.
While the relative fitness of any organism is dependent upon its environment, here we demonstrate how readily this can fluctuate within an environment. That the fitness consequences of mupirocin resistance became apparent when cell lysis became necessary may provide an explanation for why this resistance-conferring mutation, despite the ease with which it can emerge, is not more prevalent and is readily lost within healthy communities .
Understanding the evolution of antibiotic resistance, in terms of both how it emerges and how it is stably maintained, is critical if ways to curtail its further emergence are to be developed. With some antibiotic resistance mechanisms being reported as incurring no fitness costs [4–6], and others where physiological and regulatory compensation of costs has been demonstrated [23, 31, 33–35], it is perhaps surprising that they have not become more prevalent. Here, by adopting a population-based functional genomics approach, we have uncovered hidden fitness costs associated with an apparently cost-free, antibiotic resistance mechanism. We have demonstrated that the mupirocin resistance-conferring mutation of the gene encoding the IleRS enzyme has pleiotropic effects on bacteria, which in hindsight is perhaps unsurprising given the highly conserved and essential nature of this gene. In this instance, while reducing the energetic expense associated with toxin production decreases the resistance-associated fitness costs, this appears to be unsustainable when the ability to produce toxins is required. Given that the nose of healthy carriers has been shown to be such a toxicity-favouring environment for S. aureus, this toxicity-related fitness off-setting may explain why this resistance mechanism is not more prevalent, and is readily displaced, once this antibiotic has been removed from their environment. An effect we would have been unable to explain had we not adopted a population-based approach to studying this major bacterial pathogen.
Strains and growth conditions
All strains used in this study are listed in Additional file 1: Table S5. S. aureus strains were grown at 37 °C, in either tryptic soy agar or broth (TSA/TSB) with the appropriate antibiotic where necessary. The Escherichia coli TOP10 strain containing the pAgrC(his)A plasmid was grown in Luria-Bertani (LB) media with 100 μg/ml ampicillin. THP-1 cells were grown in RPMI 1640 supplemented with foetal bovine serum (10%), L-glutamine (2 mM), penicillin (100 units/ml) and streptomycin (100 μg/ml) and incubated at 37 °C with 5% CO2.
Analysis of SNP–SNP epistatic interactions was performed using the “epistasis” option in the PLINK software package (http://zzz.bwh.harvard.edu/plink/) , which is based on linear regression analysis and tests the inclusion of an interaction term (into the regression equation) for statistical significance. The total number of non-synonomous SNPs across our collection of 130 USA300 isolates was just under 4000; however, only those with a minor allele frequency of > 3% were considered, and as interactions between two SNPs can only be tested where both coincide, the number of tests per locus was on average just over 100.
Selection of the mupirocin resistant strain
a control set to confirm the presence of the ileS gene: IleS-F-control and IleS-R
a set to determine the presence of a G at position 1762: IleS-F-mupS and IleS-R
a set to determine the presence of a T at position 1762: IleS-F-mupR and IleS-R
Forward control primer common to both resistant and sensitive bacteria
Forward primer matched to the wild-type mupSsequence
Forward primer with two mismatches which amplifies the mupR sequence
Reverse primer common to both resistant and sensitive bacteria
Genome sequencing of the mupR strain
S. aureus strain MY40 was sequenced in this study; DNA was extracted using the QIAamp NA Mini Kit (QIAGEN, Crawley, UK) using the manufacturer’s instructions with 1.5 μg/μL lysostaphin (Ambi Products LLC, NY, USA) to facilitate cell lysis. DNA was quantified using a Nanodrop spectrophotometer, as well as the Quant-iT DNA Assay Kit (Life Technologies, Paisley, UK) before sequencing. High-throughput genome sequencing was performed using a MiSeq machine (Illumina, San Diego, CA, USA) and the short read, paired-end data were assembled using the de novo assembly algorithm SPAdes . Sequence data are archived in the NCBI repositories: GenBank accession NIPL00000000.1, Short Read Archive (SRA) SRR5651527, associated with BioProject PRJNA384009. Assembled genomes are also available on FigShare (doi.org/10.6084/m9.figshare.5089939.v1).
THP-1 cells  were grown as described above and harvested by centrifugation and washed in phosphate buffered saline (PBS) and diluted to a final density (determined by haemocytometer) of 2 × 106 cells/ml of PBS. Bacterial supernatant was harvested after 18 h of growth in TSB at 37 °C. The bacterial supernatant (20 μl) was mixed with 20 μl of THP-1 cells and incubated for 12 min at 37 °C. Guava ViaCount (260 μl; Milipore) was added to the sample and incubated at room temperature for 5 min before analysing the viability on the Guava flow cytometer (Milipore). Each strain was assayed six times (three biological replicates and two technical replicates).
The Malthusian parameters of the mupS and mupR strains were compared using a Mann-Whitney U test as the data were not normally distributed. For the individual fitness assays the strains were cultured overnight, diluted and added to 0.1× TSB (made with PBS rather than water to maintain the integrity of the horse blood cells) to which 5% horse blood (Oxoid) was added. Bacterial growth was determined by plating the cultures on TSA plates after 3, 6 and 24 h of incubation at 37 °C in air. Each competition was performed between 12 and 20, either three biological replicates and three technical or four biological replicates and five technical.
Deletion of the agr locus from SH1000 and MY40
Phage transduction was used to construct the mupR and mupS Agr mutants. In the S. aureus strain ROJ48, the entire Agr locus has been replaced with an erythromycin resistance cassette . This was moved from ROJ48 into SH1000 by phage transduction as follows: ROJ48 ϕ11 lysates were prepared from 200 μl of overnight ROJ48 culture in LK (1% tryptone, 0.5% yeast extract, 1.6% KCl) which was added to 3 ml of fresh LK and 3 ml of phage buffer (10 mM MgSO4, 4 mM CaCl2, 50 mM Tris-HCl pH 7.8, 100 mM NaCl and 0.1% gelatine powder in molecular/MiliQ water), and to this 500 μl ϕ11-RN6390B lysate was added. This was incubated at 30 °C shaking until the media became clear, which indicated bacterial lysis. The lysates were then filter sterilised and a second round of lysis was carried out on ROJ48 with this first round lysate. After these two lysis steps, transduction into SH1000 and MY40 strains was performed by adding 200 μl of overnight culture to 1.8 ml LK with 10 μl 1 M CaCl2 and 500 μl the ϕ11-ROJ48 lysate. This was incubated at 37 °C with shaking for 45 min; then 1 ml ice-cold 20 mM trisodium citrate was added and the transducing mixture was placed on ice for 5 min. The bacteria were harvested by centrifugation and re-suspended with 1 ml ice-cold 20 mM trisodium citrate. This was incubated on ice for 2.5 h and plated onto TSA plates with 20 mM trisodium citrate, erythromycin and lincomycin (25 μg/ml), which was incubated overnight at 37 °C.
Overnight cultures were diluted 1:500 into 3 ml fresh TSB-chloramphenicol. After 18 h of growth 2 ml of this culture was mixed with 4 ml RNA Protect Bacteria (Qiagen), and the RNeasy Mini Kit (Qiagen) was used to extract RNA following the manufacturer’s protocol. Lysostaphin (200 μg/ml) was added to Tris-EDTA buffer (Ambion), and this was added to the sample after the RNA Protect step before continuing with the protocol. When the RNA was extracted, a Turbo DNA-free kit (Thermo) was used to remove genomic DNA from the RNA samples; 3 μl Turbo DNase was added to the sample and incubated for 1.5 h at 37 °C, then a further 4 μl Turbo DNase was added and incubated for 1.5 h. DNase inactivation reagent (35 μl) was added to the samples to inactivate the DNase according to the protocol. The concentration of RNA in the samples was measured using a Qubit RNA Broad Range Kit (Thermo) and normalised before using a QuantiTect Reverse Transcription Kit (Qiagen) to convert the RNA samples into cDNA according to the manufacturer’s protocol. After adding the reverse transcriptase, the samples were incubated at 42 °C for 20 min before raising the temperature to 95 °C for 3 min to inactivate the reverse transcriptase. Primers for gyrB, a housekeeping gene, were used alongside those for RNAIII to standardise transcript levels (gyrB forward, CCAGGTAAATTAGCCGATTGC; gyrB reverse, AAATCGCCTGCGTTCTAGAG; RNAIII forward, AGCATGTAAGCTATCGTAAACAAC; RNAIII reverse, TTCAATCTATTTTTGGGGATG). ssoAdvanced SYBR Green Supermix (Bio-Rad) was used using a standard curve of known genomic DNA concentrations for each primer set. Samples, standards and water (5 μl) were pipetted into the wells of a 96-well PCR plate. The supermix was added to water and primers according to the manufacturer’s protocol, and 15 μl of this mix was pipetted over the DNA samples. This was then placed into a qPCR machine and run using the manufacturer’s recommendation. The quantity of RNAIII cDNA was divided by the quantity of gyrB cDNA to get a ratio of RNAIII transcription levels. This was performed six times, three biological replicates and two technical.
AIP EC50 and AIP quantification assays
These assays were performed as described previously . Strains were grown overnight in Brain Heart Infusion (BHI), and 1 ml of the overnight was washed three times with PBS. The pellet was re-suspended in fresh BHI and grown for a further 2 h. A dilution series of AIP-1 was created, with concentrations ranging from 250 to 0.125 nM in 1:2 dilutions, plus 1250 and 2500 nM. The culture (190 μl) was pipetted into a black 96-well plate, and 10 μl of this dilution series was also pipetted into the wells. The OD600 and luminescence readings were carried out every 15 min for 80 cycles in a Tecan plate reader. For the AIP quantification strains were grown in BHI with chloramphenicol (10 μg/ml). The reporter strain was diluted 1/20 into fresh BHI and grown for a futher 2 h, while the supernatants of the mupS and mupR strains were filtered through a 0.22 μm filter. The reporter strain was diluted 1/50 and pipetted into a black 96-well plate. The supernatant was added to the wells at 5% final concentration, and the OD600 and luminescence readings were carried out every 15 min for 80 cycles in a Tecan plate reader. This was performed six times, three biological replicates and two technical.
PSM quantification in supernatants
An overnight culture of SH1000 and MY40 was diluted 1:1000 into 50 ml fresh TSB and grown for 18 h. The cultures were centrifuged at 18,000 rpm for 10 min and 35 ml of the supernatant was mixed with 10 ml butanol. The samples were incubated at 37 °C shaking for 3 h and then centrifuged at 3000 rpm for 3 min; 1 ml of the upper organic layer was taken off. The samples were then freeze-dried overnight and then re-suspended in 160 μl 8 M urea, separated on an SDS-PAGE gel and PSMs quantified by densitometry analysis using the ImageJ software. This was performed six times, three biological replicates and two technical.
TMT labelling and high pH reversed-phase chromatography
Aliquots of 100 μg of up to ten samples per experiment were digested with trypsin (2.5 μg trypsin per 100 μg protein; 37 °C, overnight), labelled with Tandem Mass Tag (TMT) ten plex reagents according to the manufacturer’s protocol (Thermo Fisher Scientific) and the labelled samples pooled. An aliquot of the pooled sample was evaporated to dryness and resuspended in buffer A (20 mM ammonium hydroxide, pH 10) prior to fractionation by high pH reversed-phase chromatography using an Ultimate 3000 liquid chromatography system (Thermo Fisher Scientific). In brief, the sample was loaded onto an XBridge BEH C18 column (130 Å, 3.5 μm, 2.1 mm × 150 mm, Waters, UK) in buffer A and peptides eluted with an increasing gradient of buffer B (20 mM ammonium hydroxide in acetonitrile, pH 10) from 0 to 95% over 60 min. The resulting fractions were evaporated to dryness and resuspended in 1% formic acid prior to analysis by nano-LC MSMS using an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific).
Nano-LC mass spectrometry
High pH RP fractions were further fractionated using an Ultimate 3000 nanoHPLC system in line with an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific). In brief, peptides in 1% (vol/vol) formic acid were injected onto an Acclaim PepMap C18 nano-trap column (Thermo Scientific). After washing with 0.5% (vol/vol) acetonitrile 0.1% (vol/vol), formic acid peptides were resolved on a 250 mm × 75 μm Acclaim PepMap C18 reverse phase analytical column (Thermo Scientific) over a 150 min organic gradient, using seven gradient segments (1–6% solvent B over 1 min, 6–15% B over 58 min, 15–32% B over 58 min, 32–40% B over 5 min, 40–90% B over 1 min, held at 90% B for 6 min and then reduced to 1% B over 1 min) with a flow rate of 300 nl min− 1. Solvent A was 0.1% formic acid and solvent B was aqueous 80% acetonitrile in 0.1% formic acid. Peptides were ionised by nano-electrospray ionisation at 2.0 kV using a stainless steel emitter with an internal diameter of 30 μm (Thermo Scientific) and a capillary temperature of 275 °C.
All spectra were acquired using an Orbitrap Fusion Tribrid mass spectrometer controlled by Xcalibur 2.0 software (Thermo Scientific) and operated in data-dependent acquisition mode using an SPS-MS3 workflow. FTMS1 spectra were collected at a resolution of 120,000 with an automatic gain control (AGC) target of 200,000 and a max injection time of 50 ms. Precursors were filtered with an intensity threshold of 5000 according to charge state (to include charge states 2–7) and with monoisotopic precursor selection. Previously interrogated precursors were excluded using a dynamic window (60s ± 10 ppm). The MS2 precursors were isolated with a quadrupole mass filter set to a width of 1.2 m/z. ITMS2 spectra were collected with an AGC target of 10,000, max injection time of 70 ms and CID collision energy of 35%.
For FTMS3 analysis, the Orbitrap was operated at 50,000 resolution with an AGC target of 50,000 and a max injection time of 105 ms. Precursors were fragmented by high energy collision dissociation (HCD) at a normalised collision energy of 60% to ensure maximal TMT reporter ion yield. Synchronous precursor selection (SPS) was enabled to include up to five MS2 fragment ions in the FTMS3 scan.
Proteomic data analysis
The raw data files were processed and quantified using Proteome Discoverer software v2.1 (Thermo Scientific) and searched against the UniProt Staphylococcus aureus strain NCTC 8325 database using the SEQUEST algorithm . Peptide precursor mass tolerance was set at 10 ppm and MS/MS tolerance was set at 0.6 Da. Search criteria included oxidation of methionine (+ 15.9949) as a variable modification and carbamidomethylation of cysteine (+ 57.0214) and the addition of the TMT mass tag (+ 229.163) to peptide N-termini and lysine as fixed modifications. Searches were performed with full tryptic digestion and a maximum of two missed cleavages were allowed. The reverse database search option was enabled and all peptide data were filtered to satisfy a false discovery rate of 5%.
MY is a self-funded PhD student and ES is on a BBSRC DTP PhD training program.
MY, ES, ML, LB, KH, SB, EM, AL, SR and BP performed the experiments. MY, ML, NO, AJoN, PW, GM, MR, LDH and RCM designed the experiments. MY, SKS and RCM provided resources and reagents. MY, KH, SB, BP, MR, LDH and RCM analysed the data. MY, ML, MR and RCM wrote the manuscript. ES and ML contributed equally to this work. All authors read and approved the final manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Tenover FC. Mechanisms of antimicrobial resistance in Bacteria. Am J Med. 2006;119:Suppl 3–10.View ArticleGoogle Scholar
- Moura de Sousa J, Balbontín R, Durão P, Gordo I. Multidrug-resistant bacteria compensate for the epistasis between resistances. PLoS Biol. 2017;15:e2001741.View ArticlePubMedPubMed CentralGoogle Scholar
- Levin BR, Perrot V, Walker N. Compensatory mutations, antibiotic resistance and the population genetics of adaptive evolution in bacteria. Genetics. 2000;154:985–97.PubMedPubMed CentralGoogle Scholar
- Hurdle JG, O’Neil AJ, Chopra I. The isoleucyl-tRNA synthetase mutation V588F conferring mupirocin resistance in glycopeptide-intermediate Staphylococcus aureus is not associated with a significant fitness burden. J Antimicrob Chemother. 2003;53:102–4.View ArticlePubMedGoogle Scholar
- Stanczak-Mrozek KI, Manne A, Knight GM, Gould K, Witney AA, Lindsay JA. Within-host diversity of MRSA antimicrobial resistances. J Antimicrob Chemother. 2015;70:2191–8.View ArticlePubMedPubMed CentralGoogle Scholar
- Knight GM, Budd EL, Lindsay JA. Large mobile genetic elements carrying resistance genes that do not confer a fitness burden in healthcare-associated meticillin-resistant Staphylococcus aureus. Microbiology. 2013;159:1661–72.View ArticlePubMedGoogle Scholar
- Melnyk AH, Wong A, Kassen R. The fitness costs of antibiotic resistance mutations. Evol Appl. 2015;8:273–83.View ArticlePubMedGoogle Scholar
- Rasigade JP, Vandenesch F. Staphylococcus aureus: a pathogen with still unresolved issues. Infect Genet Evol. 2014;21:510–4.View ArticlePubMedGoogle Scholar
- Lowy F. Antimicrobial resistance: the example of Staphylococcus aureus. J Clin Invest. 2003;111:1265–73.View ArticlePubMedPubMed CentralGoogle Scholar
- Wertheim HF, Melles DC, Vos MC, van Leeuwen W, van Belkum A, Verbrugh HA, et al. The role of nasal carriage in Staphylococcus aureus infections. Lancet Infect Dis. 2005;5:751–62.View ArticlePubMedGoogle Scholar
- Gordon RJ, Lowy FD. Pathogenesis of methicillin-resistant Staphylococcus aureus infection. Clin Infect Dis. 2008;46:S350–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Abad CL, Pulia MS, Safdar N. Does the nose know? An update on MRSA decolonization strategies. Curr Infect Dis Rep. 2013;15:455–64.View ArticlePubMedPubMed CentralGoogle Scholar
- Thomas CM, Hothersall J, Willis CL, Simpson TJ. Resistance to and synthesis of the antibiotic mupirocin. Nat Rev Microbiol. 2010;8:281–9.View ArticlePubMedGoogle Scholar
- Simor AE. Staphylococcal decolonisation: an effective strategy for prevention of infection? Lancet Infect Dis. 2011;11:952–62.View ArticlePubMedGoogle Scholar
- Bertino JS Jr. Intranasal mupirocin for outbreaks of methicillin- resistant Staphylococcus aureus. Am J Heal Pharm. 1997;54:2185–91.Google Scholar
- Antonio M, McFerran N, Pallen MJ. Mutations affecting the Rossman fold of isoleucyl-tRNA synthetase are correlated with low-level mupirocin resistance in Staphylococcus aureus. Antimicrob Agents Chemother. 2002;46:438–42.View ArticlePubMedPubMed CentralGoogle Scholar
- Patel JB, Gorwitz RJ, Jernigan JA. Mupirocin resistance. Clin Infect Dis. 2009;49:935–41.View ArticlePubMedGoogle Scholar
- Udo EE, Sarkhoo E. Genetic analysis of high-level mupirocin resistance in the ST80 clone of community-associated meticillin-resistant Staphylococcus aureus. J Med Microbiol. 2010;59:193–9.View ArticlePubMedGoogle Scholar
- Seah C, Alexander DC, Louie L, Simor A, Low DE, Longtin J, et al. MupB, a new high-level mupirocin resistance mechanism in Staphylococcus aureus. Antimicrob Agents Chemother. 2012;56:1916–20.View ArticlePubMedPubMed CentralGoogle Scholar
- Hetem DJ, Bonten MJM. Clinical relevance of mupirocin resistance in Staphylococcus aureus. J Hosp Infect. 2013;85:249–56.View ArticlePubMedGoogle Scholar
- Williamson DA, Carter GP, Howden BP. Current and emerging topical Antibacterials and antiseptics: agents, action, and resistance patterns. Clin Microbiol Rev. 2017;30:827–60.View ArticlePubMedPubMed CentralGoogle Scholar
- Laabei M, Recker M, Rudkin JK, Aldeljawi M, Gulay Z, Sloan TJ, et al. Predicting the virulence of MRSA from its genome sequence. Genome Res. 2014;24:839–49.View ArticlePubMedPubMed CentralGoogle Scholar
- Laabei M, Uhlemann AC, Lowy FD, Austin ED, Yokoyama M, Ouadi K, et al. Evolutionary trade-offs underlie the multi-faceted virulence of Staphylococcus aureus. PLoS Biol. 2015;13(9):e1002229.View ArticlePubMedPubMed CentralGoogle Scholar
- Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–75.View ArticlePubMedPubMed CentralGoogle Scholar
- Abranches J, Martinez AR, Kajfasz JK, Chávez V, Garsin DA, Lemos JA. The molecular Alarmone (p)ppGpp mediates stress responses, vancomycin tolerance, and virulence in enterococcus faecalis. J Bacteriol. 2009;191:2248–56.View ArticlePubMedPubMed CentralGoogle Scholar
- Reiß S, Pané-Farré J, Fuchs S, François P, Liebeke M, Schrenzel J, et al. Global analysis of the Staphylococcus aureus response to mupirocin. Antimicrob Agents Chemother. 2012;56:787–804.View ArticlePubMedPubMed CentralGoogle Scholar
- Dayon L, Sanchez JC. Relative protein quantification by MS/MS using the tandem mass tag technology. Methods Mol Biol. 2012;893:115–27.View ArticlePubMedGoogle Scholar
- Qi R, Joo HS, Sharma-Kuinkel B, Berlon NR, Park L, Fu CL, et al. Increased in vitro phenol-soluble Modulin production is associated with soft tissue infection source in clinical isolates of methicillin-susceptible Staphylococcus aureus. J Inf Secur. 2017;72:302–8.Google Scholar
- Chatterjee S, Joo HS, Duong AC, Dieringer TD, Tan VY, Song Y, et al. Essential Staphylococcus aureus toxin export system. Nat Med. 2013;19:364–7.View ArticlePubMedPubMed CentralGoogle Scholar
- Somerville GA, Beres SB, Fitzgerald JR, DeLeo FR, Cole RL, Hoff JS, et al. In vitro serial passage of Staphylococcus aureus: changes in physiology, virulence factor production, and agr nucleotide sequence. J Bacteriol. 2002;184:1430–7.View ArticlePubMedPubMed CentralGoogle Scholar
- Collins J, Rudkin J, Recker M, Pozzi C, O'Gara JP, Massey RC. Offsetting virulence and antibiotic resistance costs by MRSA. ISME J. 2010;4:577–84.View ArticlePubMedGoogle Scholar
- Lenski RE, Rose MR, Simpson SC, Tadler SC. Long-term experimental evolution in Escherichia coli. I. Adaptation and divergence during 2,000 generations. Am Nat. 1991;138:1315–41.View ArticleGoogle Scholar
- Recker M, Laabei M, Toleman MS, Reuter S, Saunderson RB, Blane B, et al. Clonal differences in Staphylococcus aureus bacteraemia-associated mortality. Nat Microbiol. 2017;2:1381–8.View ArticlePubMedGoogle Scholar
- Pacheco JO, Alvarez-Ortega C, Rico MA, Martínez JL. Metabolic compensation of fitness costs is a general outcome for antibiotic-resistant Pseudomonas aeruginosa mutants overexpressing efflux pumps. MBio. 2017;8:e00500–17.Google Scholar
- Freihofer P, Akbergenov R, Teo Y, Juskeviciene R, Andersson DI, Böttger EC. Nonmutational compensation of the fitness cost of antibiotic resistance in mycobacteria by overexpression of tlyA rRNA methylase. RNA. 2016;22:1836–43.PubMedPubMed CentralView ArticleGoogle Scholar
- Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol. 2012;19:455–77.View ArticlePubMedPubMed CentralGoogle Scholar
- Tsuchiya S, Yamabe M, Yamaguchi Y, Kobayashi Y, Konno T, Tada K. Establishment and characterization of a human acute monocytic leukemia cell line (THP-1). Int J Cancer. 1980;26:171–6.View ArticlePubMedGoogle Scholar
- Jensen RO, Winzer K, Clarke SR, Chan WC, Williams P. Differential recognition of Staphylococcus aureus quorum-sensing signals depends on both extracellular loops 1 and 2 of the transmembrane sensor AgrC. J Mol Biol. 2008;381:300–9.View ArticlePubMedGoogle Scholar
- MacCoss MJ, Wu CC, Yates JR 3rd. Probability-based validation of protein identifications using a modified SEQUEST algorithm. Anal Chem. 2002;74:5593–9.View ArticlePubMedGoogle Scholar
- Yokoyama M, Stevens E, Laabei M, Bacon L, Heesom K, Bayliss S, et al, Epistasis analysis uncovers hidden antibiotic resistance-associated fitness costs hampering the evolution of MRSA. https://www.ncbi.nlm.nih.gov/bioproject/PRJNA384009 (2018).Google Scholar
- Yokoyama M, Stevens E, Laabei M, Bacon L, Heesom K, Bayliss S, et al, Epistasis analysis uncovers hidden antibiotic resistance-associated fitness costs hampering the evolution of MRSA. doi.org/10.6084/m9.figshare.5089939.v1 (2018).Google Scholar