Apt (Adenine Phosphoribosyltransferase) Mutation in Laboratory-Selected Vancomycin-Intermediate Staphylococcus aureus

Comparative genomic sequencing of laboratory-derived vancomycin-intermediate Staphylococcus aureus (VISA) (MM66-3 and MM66-4) revealed unique mutations in both MM66-3 (in apt and ssaA6), and MM66-4 (in apt and walK), compared to hetero-VISA parent strain MM66. Transcriptional profiling revealed that both MM66 VISA shared 79 upregulated genes and eight downregulated genes. Of these, 30.4% of the upregulated genes were associated with the cell envelope, whereas 75% of the downregulated genes were associated with virulence. In concordance with mutations and transcriptome alterations, both VISA strains demonstrated reduced autolysis, reduced growth in the presence of salt and reduced virulence factor activity. In addition to mutations in genes linked to cell wall metabolism (ssaA6 and walK), the same mutation in apt which encodes adenine phosphoribosyltransferase, was confirmed in both MM66 VISA. Apt plays a role in both adenine metabolism and accumulation and both MM66 VISA grew better than MM66 in the presence of adenine or 2-fluoroadenine indicating a reduction in the accumulation of these growth inhibiting compounds in the VISA strains. MM66 apt mutants isolated via 2-fluoroadenine selection also demonstrated reduced susceptibility to the cell wall lytic dye Congo red and vancomycin. Finding that apt mutations contribute to reduced vancomycin susceptibility once again suggests a role for altered purine metabolism in a VISA mechanism.


Introduction
Staphylococcus aureus is a notorious human pathogen that is associated with both hospital-and community-acquired infections. Since the emergence of multidrug-resistant methicillin-resistant S. aureus (MRSA), treatment of infections caused by these organisms has become challenging [1]. The glycopeptide antibiotic vancomycin remains a clinically proven drug for the treatment of serious MRSA infections [2]. Vancomycin binds to the carboxy terminus of the D-alanyl-D-alanine residues of the lipid II peptidoglycan precursor in the cytoplasmic membrane and prevents peptidoglycan synthesis, disrupting cell wall metabolism and ultimately leading to cell death [2,3].
The increased use of vancomycin, in large part due to the increased incidence of MRSA infections, eventually led to the selection of S. aureus strains that demonstrated reduced susceptibility and resistance to vancomycin. Based on vancomycin MICs, S. aureus isolates are classified as vancomycin-susceptible S. aureus (vancomycin MIC ≤ 2 mg/L), regulator, and another was between two genes encoding hypothetical proteins (SACOL2608 and SACOL2609) ( Table 1). CGS also identified six intragenic nonsynonomous mutations in MM66-3. One mutation was located in a putative adenine phosphoribosyltransferase gene apt, which leads to an A 57 → V 57 (nonpolar A for the larger nonpolar V alteration) within the Apt β3 domain which is part of the five-stranded enzyme core fold that is surrounded by three α-helices [16] (Figure 1). Three additional mutations were located in a single codon of a gene encoding a staphylococcal secretory antigen (ssaA6). Six S. aureus SsaA homologues [17] including ssaA6 share a common cysteine, histidine-dependent amidohydrolases/peptidases-amidase (CHAP) domain [18] and amino-terminal signal sequence, suggesting that these proteins play a role in cell wall metabolism [19]. A number of ssaA homologues are under the control of walKR, which encodes a two-component regulatory system that controls cell wall autolysis in S. aureus [20,21]. In MM66-3 the mutations in ssaA6 lead to a H 297 → G 297 (basic H for uncharged G alteration) in the putative peptidoglycan-degrading amidase domain of SsaA6. The fifth mutation was located in a gene encoding a hypothetical protein and the sixth mutation was located within a gene encoding a sulfite reductase (NADPH) flavoprotein alpha-component coding gene cysJ (Table 1). All of these sequence discrepancies were confirmed in MM66-3 via PCR amplicon sequencing, yet comparison of all these mutations within the same intergenic regions and genes of 9 S. aureus genomes demonstrated that only the ssaA6 and apt were unique to the MM66-3 genome. CGS data analysis of MM66-3 revealed two intergenic mutations. One was between genes encoding a putative IS1181 transposase and a putative Fur family transcriptional regulator, and another was between two genes encoding hypothetical proteins (SACOL2608 and SACOL2609) ( Table 1). CGS also identified six intragenic nonsynonomous mutations in MM66-3. One mutation was located in a putative adenine phosphoribosyltransferase gene apt, which leads to an A 57  V 57 (nonpolar A for the larger nonpolar V alteration) within the Apt β3 domain which is part of the five-stranded enzyme core fold that is surrounded by three α-helices [16] (Figure 1). Three additional mutations were located in a single codon of a gene encoding a staphylococcal secretory antigen (ssaA6). Six S. aureus SsaA homologues [17] including ssaA6 share a common cysteine, histidine-dependent amidohydrolases/peptidases-amidase (CHAP) domain [18] and amino-terminal signal sequence, suggesting that these proteins play a role in cell wall metabolism [19]. A number of ssaA homologues are under the control of walKR, which encodes a two-component regulatory system that controls cell wall autolysis in S. aureus [20,21]. In MM66-3 the mutations in ssaA6 lead to a H 297  G 297 (basic H for uncharged G alteration) in the putative peptidoglycan-degrading amidase domain of SsaA6. The fifth mutation was located in a gene encoding a hypothetical protein and the sixth mutation was located within a gene encoding a sulfite reductase (NADPH) flavoprotein alpha-component coding gene cysJ ( Table 1). All of these sequence discrepancies were confirmed in MM66-3 via PCR amplicon sequencing, yet comparison of all these mutations within the same intergenic regions and genes of 9 S. aureus genomes demonstrated that only the ssaA6 and apt were unique to the MM66-3 genome.    We previously reported that CGS analysis identified eight mutations within MM66-4 which included the exact same apt mutation found in MM66-3 and a mutation in the histidine kinase walK gene [22,23]. The mutation in walK leads to a K 263 → E 263 (dramatic charge inversion in K to E alteration) in the highly conserved central PAS domain of the sensor module of this sensor histidine kinase [24]. While all eight of these sequence discrepancies were confirmed via PCR amplicon sequencing in this study, nucleotide sequence alignments with nine other S. aureus genomes demonstrated that only the walK and apt mutation are unique to the MM66-4 genome. A reduction in walRK activity leads to the reduced transcription of multiple ssaA homologues as well as other autolysin genes [19,20] and altered walRK transcription and mutations have been implicated in the control of the VISA mechanism [25][26][27][28].

SACOL1169
Fibrinogen-binding protein precursor-related opuD2 BCCT family osmoprotectant transporter ND −4.6 a Gene ID correspond to strain COL genome. Abbreviations: ND, transcriptional alteration not detected.
The S. aureus lacEFG encodes the lactose-specific transport components and β-galactosidase [29,30]. The highest upregulated genes in both MM66-4 and MM66-3 were the genes lacA1, lacE and lacF ( Table 2). The gene lacA1 could encode a small (38 amino acid) protein which is transcribed downstream of, and in the same transcriptional direction as lacR, the repressor of the lac operon [31]. In other VISA strains, lac operon genes have been either not transcriptionally altered [4] or are downregulated [32]. The relationship between altered lac gene expression and the VISA mechanism is unclear. Compared to MM66, the highest downregulated genes in both MM66 VISA mutants were virulence genes (e.g., SACOL1164, SACOL1173 and SACOL2418) ( Table 2 and Table S3).

Altered Cell Wall Metabolism Genes in MM66 VISA
In addition to the ssaA6 and walK mutations, other cell wall metabolism genes demonstrated altered expression in the MM66 VISA mutants. ssaA1, ssaA2, and ssaA3 (Table 2), the autolysin gene aaa [39], and aaa1 (demonstrated 38% identity along its entire length to aaa) ( Table 2) were all downregulated in MM66-4. In addition, MM66-3 demonstrated reduced expression of the walRK-controlled autolysin gene alt (Table 2). arlS which encodes the sensor kinase of the two-component regulatory system ArlSR which regulates genes involved with autolysis and cell division [38,[40][41][42], was upregulated in both MM66 VISA investigated (Table S3). arlR was also upregulated only in MM66-4 (Table S2). ArlRS was initially characterized as a regulator of autolysis [41] and it was reported that ArlRS repressed the expression of the autolysin gene lytN [42][43][44]. ArlRS also controls the expression of atlR which encodes a gene product that regulates the expression of the autolysin gene atlA [42,43]. The gene murA which encodes a UDP-N-acetylglucosamine enolpyruvyltranferase that catalyzes the first step of peptidoglycan biosynthesis [45], was upregulated in MM66-4. Lastly, the gene sceD which encodes for an autolysin important for cell separation [46] was upregulated in both VISA mutants (Table 2). We hypothesized that akin to other VISA strains [2,3] the MM66 mutants would demonstrate altered whole cell autolysis, and as expected, both MM66 VISA mutants demonstrated reduced whole cell autolysis compared to parent strain MM66 (Figure 2). gene atlA [42,43]. The gene murA which encodes a UDP-N-acetylglucosamine enolpyruvyltranferase that catalyzes the first step of peptidoglycan biosynthesis [45], was upregulated in MM66-4. Lastly, the gene sceD which encodes for an autolysin important for cell separation [46] was upregulated in both VISA mutants (Table 2). We hypothesized that akin to other VISA strains [2,3] the MM66 mutants would demonstrate altered whole cell autolysis, and as expected, both MM66 VISA mutants demonstrated reduced whole cell autolysis compared to parent strain MM66 ( Figure 2).

Physiological Corroboration of apt Mutation in MM66 VISA Mutants and Characterization of MM66 apt Mutants
It has been reported that apt plays a role in the uptake of exogenous adenine [9,47] and that adenine inhibits the growth of S. aureus [48]. Additionally, apt mutants of Bacillus subtilis exhibited increased tolerance to the toxic adenine analogue 2-fluoroadenine [47]. We therefore hypothesized that the apt mutation in the MM66 VISA mutants would reduce adenine and 2-fluoroadenine accumulation and toxicity compared to MM66. As expected, MM66 grew slightly slower than MM66-3 and MM66-4 in the presence of 5 mM adenine ( Figure 3) and the growth of MM66 was inhibited by the addition of 2-fluoroadenine, while MM66-3 and MM66-4 grew in the presence of this toxic adenine analogue ( Figure 3). Note that MM66 VISA mutants and MM66 exhibited almost identical growth curves in drug-free media ( Figure 3). In addition to increased adenine and 2-flouroadenine resistance, we also noted that several genes required for purine (purLFMNHD) and pyrimidine (pyrABFE) biosynthesis were upregulated in both MM66 VISA mutants compared to MM66 (Table 2).

Physiological Corroboration of apt Mutation in MM66 VISA Mutants and Characterization of MM66 apt Mutants
It has been reported that apt plays a role in the uptake of exogenous adenine [9,47] and that adenine inhibits the growth of S. aureus [48]. Additionally, apt mutants of Bacillus subtilis exhibited increased tolerance to the toxic adenine analogue 2-fluoroadenine [47]. We therefore hypothesized that the apt mutation in the MM66 VISA mutants would reduce adenine and 2-fluoroadenine accumulation and toxicity compared to MM66. As expected, MM66 grew slightly slower than MM66-3 and MM66-4 in the presence of 5 mM adenine ( Figure 3) and the growth of MM66 was inhibited by the addition of 2-fluoroadenine, while MM66-3 and MM66-4 grew in the presence of this toxic adenine analogue ( Figure 3). Note that MM66 VISA mutants and MM66 exhibited almost identical growth curves in drug-free media ( Figure 3). In addition to increased adenine and 2-flouroadenine resistance, we also noted that several genes required for purine (purLFMNHD) and pyrimidine (pyrABFE) biosynthesis were upregulated in both MM66 VISA mutants compared to MM66 ( Table 2).
Selection of B. subtilis in 2-flouroadenine resulted in 2-flouroadenine reduced susceptibility (FARS) mutants harboring mutations in apt [47]. To ascertain a role for apt mutations on vancomycin susceptibility, FARS mutants of MM66 were selected and their apt regions were sequenced. Suspected MM66 FARS mutants arose on media containing 5 mM 2flouroadenine at a mutation frequency of 4.7 × 10 −7 and all randomly selected MM66 colonies demonstrated higher 2-flouroadenine MICs than MM66 (Table 3). In addition, all FARS MM66 mutants harbored one of a variety of mutations (e.g., deletion, insertion and nonsense mutations) within the apt gene ( Table 4). The penultimate 2-fluoroadenine MICs could not be determined for all FARS MM66 mutants (Table 3), since 2-flouroadenine precipitated out of the susceptibility testing media at concentrations above 7 mM. Antibiotics 2021, 10, x FOR PEER REVIEW 7 of 17  (Table 3). In addition, all FARS MM66 mutants harbored one of a variety of mutations (e.g., deletion, insertion and nonsense mutations) within the apt gene ( Table 4). The penultimate 2-fluoroadenine MICs could not be determined for all FARS MM66 mutants (Table 3), since 2-flouroadenine precipitated out of the susceptibility testing media at concentrations above 7 mM.   FARS mutants MM66-FARS-1 and MM66-FARS-6 demonstrated identical growth curves and slightly decreased tolerance to 5 mM adenine growth inhibition compared to the parent strain MM66 (Figure 4). In addition, both MM66-FARS-1 and MM66-FARS-6 demon-Antibiotics 2021, 10, 583 8 of 17 strated reduced Triton X-100-stimulated autolysis compared to MM66 ( Figure 5). Furthermore, compared to MM66, all MM66-FARS mutants investigated demonstrated increased resistance to Congo red, similar to apt mutants previously reported on ( Figure 6) [12]. In contradiction to this finding, MM66-3 and MM66-4 demonstrated reduced resistance to Congo red compared to parent strain MM66 (data not shown). FARS mutants MM66-FARS-1 and MM66-FARS-6 demonstrated identical growth curves and slightly decreased tolerance to 5 mM adenine growth inhibition compared to the parent strain MM66 (Figure 4). In addition, both MM66-FARS-1 and MM66-FARS-6 demonstrated reduced Triton X-100-stimulated autolysis compared to MM66 ( Figure 5). Furthermore, compared to MM66, all MM66-FARS mutants investigated demonstrated increased resistance to Congo red, similar to apt mutants previously reported on ( Figure  6) [12]. In contradiction to this finding, MM66-3 and MM66-4 demonstrated reduced resistance to Congo red compared to parent strain MM66 (data not shown).   Vancomycin resistance population analysis demonstrated that MM66-FARS-6 produced CFUs on 3 mg/L vancomycin while strains MM66 and MM66-FARS-1 stopped producing Antibiotics 2021, 10, 583 9 of 17 viable colonies on 2.5 mg/L (Figure 7). However, all FARS MM66 mutants investigated demonstrated increased distances grown on the vancomycin gradients investigated (Table 5).

Altered Growth of MM66 VISA Mutants in the Presence of Salt
Osmoprotectants (e.g. choline and glycine betaine) enhance the growth of S. aureus

Altered Growth of MM66 VISA Mutants in the Presence of Salt
Osmoprotectants (e.g. choline and glycine betaine) enhance the growth of S. aureus at high osmolarity [52] and a variety of gene products are required for the uptake and biosynthesis of these osmoprotectants [53]. The glycine betaine transporter-encoding opuD1 was upregulated in both MM66-3 and MM66-4 ( Table 2). The genes betA and betB, which encode proteins that catalyze the oxidation of choline to glycine betaine, and opuD2 which encodes a glycine betaine transporter were downregulated in MM66-4 ( Table 2). The qRT-PCR results demonstrated that betA was also downregulated in MM66-3 (Table 2). A third gene encoding a glycine betaine transporter component (opuAC) that demonstrated 38% identity along its entire length with the OpuAC of B. subtilis [54] was upregulated in MM66-4 ( Table 2). The data above indicated that the ability of the MM66 VISA to grow in high osmolarity might be altered. In support of this hypothesis, both VISA strains demonstrated reduced growth in the presence of 2 M NaCl and 2 M KCl ( Figure 8) compared to parent strain MM66.

Conclusions
Based on experimental evidence and bioinformatics, we propose that the ssaA6, walK and apt mutations are loss of function mutations. These mutations (apt and ssaA6 in MM66-3, and apt and walK in MM66-4) in conjunction with the altered regulation of autolysin and cell wall metabolism/envelope genes, contribute to the MM66 VISA mechanism and reduced whole cell autolysis observed. With the exception of the apt mutation, this proposal is in line with a number of previous studies that demonstrate the importance of loss of function mutations in VISA mechanisms that directly affect cell wall physiology [3,27,28,55]. Mutations in walK lead to the acquisition of the VISA phenotype [25,27,28] and decreased walRK expression has been reported to reduce Triton X-100 stimulated whole cell autolysis [19]. While ssaA6 is a new mutation reported for VISA, it is not difficult to envision a role for a mutated autolysin gene in the MM66-3 mechanism since autolytic activity and autolysin gene expression are commonly reduced in VISA strains [3,27,28,55,56].
Since Apt plays a role in the uptake of adenine [8,9], we propose that the apt mutation leads to a reduction in the accumulation of adenine and 2-flouroadenine in the MM66 VISA. This in turn, allowed for the faster growth of MM66 VISA in media amended with adenine or 2-fluoroadenine, compared to MM66. In conjunction with the apt mutation, both MM66 VISA also demonstrated increased expression of purine and pyrimidine biosynthetic genes. It is possible that the upregulation of purine biosynthesis genes observed in MM66 VISA is due to defective purine salvage that occurred as a result of the apt mutation. Genes involved with purine and pyrimidine biosynthesis have also been reported to be upregulated in VISA with a walK or walR mutations [25]. It is of interest to note that Apt catalyzes the production and consumption of phosphoribosyl-pyrophosphate, an intermediate metabolite that lies at the intersection between purine and pyrimidine metabolism and carbohydrate metabolism [57,58].
All MM66-FARS mutants that harbored one of a variety of apt mutations demonstrated increased 2-flouroadenine MICs. Compared to MM66, MM66-FARS-1 and MM66-FARS-6 also demonstrated identical growth curves, slightly decreased susceptibility to adenine growth inhibition, slightly reduced autolysis and increased resistance to Congo red. Only one FARS mutant demonstrated survival on a higher concentration of vancomycin compared to MM66 in the vancomycin resistance population analysis. Gradient plates utilize a gradual gradient of vancomycin, while population analysis employs abrupt changes in vancomycin concentrations and therefore, gradient plates can better detect smaller differences in vancomycin resistance levels. All MM66 FARS mutants demonstrated increased distances grown on the vancomycin gradients examined compared to MM66, which indicated that the apt mutations altered vancomycin susceptibility levels. We conclude that 2-fluoroadenine-selected apt mutations can contribute to reduced vancomycin susceptibility in MM66 FARS mutants.
Compared to MM66, the decrease in autolysis, increased adenine susceptibility and increase in vancomycin resistance exhibited by the MM66-FARS mutants, were not as great as that observed when we compared the MM66 VISA and MM66 (this study and [15]). This is likely due to the lack of the walK and/or ssaA6 mutation in the MM66-FARS mutants. In addition, both VISA mutants which also harbored an apt mutation, demonstrated increased susceptibility to Congo red. This is in contrast to what was observed in the MM66 FARS mutants which demonstrated increased resistance to this dye. We propose that the additional mutations and altered cell wall physiology in the VISA mutants masked the effects of the apt mutation and supported the enhanced Congo red susceptibility observed.
Our results demonstrated that sigB and/or sigB operon genes are upregulated in the MM66 VISA strains. Since sigB is required for the full expression of vancomycin intermediate-levels in VISA strains [38,59,60], this finding was not unexpected. Having a primed general stress response might allow the MM66 VISA cell greater control over genes required for survival in the presence of vancomycin.
A large percentage of downregulated genes shared by both MM66 VISA investigated were represented by genes encoding virulence factors. In addition, our results demon-strate an increased in coagulation time and reduced hemolytic activity in the MM666 VISA mutants compared to the parent strain. These findings are in line with reports of downregulated virulence genes and reduced virulence factor production by other VISA strains [55].
Our transcriptional profiling data demonstrated the altered expression of a number of genes involved with the acquisition and biosynthesis of osmoprotectants in the MM66 VISA mutants. This finding in turn led us to investigate the growth of the MM66 VISA mutants in the presence of salt. While MM66 and the MM66 VISA mutants demonstrated similar growth curves in unamended media, the MM66 VISA mutants did not grow as well as parent strain MM66 in the presence of salt. Salt can cause the contraction of Gram-positive cell walls [61] and causes S. aureus to produce larger cells [62]. It is possible that salt affects the unique peptidoglycan metabolism/structural alterations in the MM66 VISA [3] in such a way that inhibits their growth greater than MM66. Salt addition has been shown to affect susceptibility to cell wall active antibiotics, for instance it is well know that salt addition positively affects the expression of resistance to cell wall active β-lactams by MRSA [63,64].

Bacterial Strains, Culture Conditions, and Chemicals
hVISA parent strain MM66 and MM66 VISA mutant strains MM66-3 and MM66-4 were utilized for this study [15]. All bacteria were cultured in Luria Bertani broth (LB) (Difco, Detroit, MI, USA) with shaking (200 rpm, 37 • C) or on LB agar (LBA) as required. Working stock LBA cultures were kept at 4 • C and all strains were stored in LB containing 20% glycerol at −80 • C. All overnight cultures were initiated with single colonies and then allowed to grow at 37 • C overnight. Growth experiments were performed in LB and in LB containing 2 M NaCl or 2 M KCl or 5 mM adenine or 5 mM 2-fluoroadenine. To make the LB plus salt media, NaCl or KCl crystals were added directly to the liquid media before autoclaving. Adenine and 2-fluoroadenine stocks were made up in DMSO and then filter sterilized before adding to liquid media. For growth experiments, triplicates of broth cultures were initiated with an overnight culture to reach an initial OD 580nm of 0.04 and the OD 580nm of each culture was recorded over the incubation time. Unless otherwise noted, chemicals were obtained from Sigma Chemical Co. (St. Louis, MO, USA).

DNA and RNA Purification and cDNA Synthesis
S. aureus chromosomal DNA was extracted by the DNA spooling method as described previously [65]. RNA for quantitative real-time PCR (qRT-PCR) and microarray analysis was isolated from mid-exponential phase cultures (OD 580nm = 0.7) using a bead mill homogenization procedure as previously described [66] following pretreatment of cell pellets with RNA protect (Qiagen Inc., Venlo, The Netherlands). cDNAs were synthesized from DNAfree-treated (Ambion, Austin, TX, USA) RNA using Moloney murine leukemia virus SuperScript III reverse transcriptase (Invitrogen) as previously described [66].

S. aureus DNA Microarray and Quantitative Real-Time PCR Analyses
For microarray analysis, cDNA samples prepared from MM66, MM66-3 and MM66-4 were labeled with Cy3 or Cy5 post-labeling reactive dye following the manufacturer's suggestions (Amersham Biosciences, Piscataway, NJ, USA). Microarray experiments were performed in duplicate, and fluorophore dyes were swapped to produce dual cDNA samples to minimize dye bias for each strain cDNA preparation analyzed. S. aureus DNA microarrays version 4 were used for hybridization and array analysis. Hybridized arrays were scanned with a GenePix 4000B Microarray Scanner (Axon Instruments, Union City, CA, USA) and array TIFF images were analyzed using TIGR Spotfinder followed by data normalization with the LOWESS algorithm using TIGR-MIDAS. Genes that demonstrated ≥ 2.0-fold upregulation or downregulation were considered significant. The iCycler iQ Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA) and iQ SYBR Green Supermix (Bio-Rad) were utilized for qRT-PCR of control and test cDNAs. The expression level of each sample was normalized using 16S rDNA as an internal control and expression ratios were determined using the 2 −∆∆Ct method as previously described [66]. All primers utilized for qRT-PCR are described in Table S4.

Isolation and Characterization of MM66 FARS Mutants
Aliquots (100 µL) of overnight cultures were spread on to LB agar containing 5 mM 2-fluoroadenine and incubated at 37 • C for 24 h. Colonies appearing on 2-fluoroadenine containing plates were then selected as suspected FARS-susceptibility mutants. MICs for 2-fluoroadenine were determined with LB overnight cultures, which were initially diluted to an OD 580nm of 0.01. A ten-microliter aliquot of each diluted overnight culture was then spotted onto LBA plates containing increasing concentrations of 2-fluoroadenine (0 and 4 to 12.0 mM at 0.5 mM increments). The spots were then allowed to dry, and plates were incubated at 37 • C for 24 h. The MIC was the lowest 2-fluoroadenine concentration at which there was no visible growth. A problem we encountered was that after 7 mM 2-fluoroadenine, it became clear that the drug was precipitating in the LBA causing the agar to take on a cloudy appearance, and therefore the final MICs could not be determined for the MM66 FARS mutants. In order to sequence and detect mutations in the apt genes of the MM66 FARS mutants a primer set (apt100U-F and apt100D-R) was designed that would amplify 100 bp upstream and downstream of the apt gene (Table S4). Sequencing of the apt amplicons and apt mutation detection was carried out as described above (Section 4.3). The determination of relative Congo red susceptibilities was carried out according to previous published protocols [12]. Vancomycin-resistance population and vancomycin gradient plate analyses were carried out as described previously [68,69].

Triton-X-100-Induced Whole Cell Autolysis, Coagulase Time and Hemolysis
Triton-X-100-stimulated whole cell autolysis assays were performed following the protocol of Gustafson et al. [70]. For coagulase assays, strains were grown in LB and the OD 580nm was adjusted to 1.0. Then, 50 µL of the OD-adjusted cultures were added to 500 µL of rehydrated coagulase plasma (BD BBL TM ; Becton, Dickinson and Company, Sparks, MD, USA) and the time (min) to complete coagulation for all strains was then recorded. For hemolysis assays, overnight cultures were initially adjusted to an OD 580 of 0.1. The cultures were then serially diluted in sterile LB and 100 µL aliquots of the 10 −6 and 10 −7 dilutions were spread onto the surface of 5% defibrinated sheep blood agar plates (Quad Five, Ryegate, MT, USA). The plates were then incubated at 37 • C for 24 h and then at 4 • C for 2 h. The sizes (mm) of the hemolysis zones around three isolated colonies of each strain investigated were then measured.