Aureusimines in Staphylococcus aureus Are Not Involved in Virulence

Background Recently, dipeptide aureusimines were reported to activate expression of staphylococcal virulence genes, such as alpha-hemolysin, and increase S. aureus virulence. Surprisingly, most of the virulence genes affected by aureusimines form part of the regulon of the SaeRS two component system (TCS), raising the possibility that SaeRS might be directly or indirectly involved in the aureusimine-dependent signaling process. Methodology/Principal Findings Using HPLC analyses, we confirmed that a transposon mutant of ausA, the gene encoding the aureusimine dipeptide synthesis enzyme, does not produce dipeptides. However, the transposon mutant showed normal hemolysis activity and alpha-hemolysin/SaeP production. Furthermore, the P1 promoter of the sae operon, one of the targets of the SaeRS TCS, showed normal transcription activity. Moreover, in contrast to the original report, the ausA transposon mutant did not exhibit attenuated virulence in an animal infection model. DNA sequencing revealed that the ausA deletion mutant used in the original study has an 83 nt-duplication in saeS. Hemolysis activity of the original mutant was restored by a plasmid carrying the sae operon. A mutant of the sae operon showed elevated resistance to chloramphenicol and erythromycin, two antibiotics widely used during staphylococcal mutagenesis. At 43°C in the presence of erythromycin and aeration, the conditions typically employed for staphylococcal mutagenesis, an saeR transposon mutant grew much faster than a control mutant and the saeR mutant was highly enriched in a mixed culture experiment. Conclusions/Significance Our results show that the previously reported roles of aureusimines in staphylococcal gene regulation and virulence were due to an unintended mutation in saeS, which was likely selected due to elevated resistance of the mutant to environmental stresses. Thus, there is no evidence indicating that the dipeptide aureusimines play a role in sae-mediated virulence factor production or contribute to staphylococcal virulence.


Introduction
Staphylococcus aureus is an important human pathogen that causes a wide variety of diseases, ranging from minor skin infections to life-threatening blood infections, endocarditis, pneumonia, and toxic shock syndrome [1,2]. The bacterial pathogen can infect almost every part of the human body and this versatility of the bacterium is, at least partly, due to numerous virulence factors such as hemolysins, leukocidins, and immune modulators. The expression of these virulence factors is coordinately regulated, for example by multiple transcription factors or two component systems [3]. The SaeRS two component system (TCS) in particular plays a key role in the production of important virulence factors such as alpha-hemolysin, coagulase, and fibronectin binding proteins [4,5,6,7,8] in vitro and it has been shown that SaeRS is essential for in-vivo production of alpha-hemolysin [9].
The SaeRS TCS is encoded by the saePQRS operon, in which saeR and saeS encode the response regulator and sensor kinase, respectively (Fig. 1A). The sae operon has two promoters, P1 and P3 [10,11,12]. Since the P1 promoter is positively autoregulated by the SaeRS system [11,13,14], along with alpha-hemolysin production, P1 promoter activity and P1-driven production of SaeP are good indicators for the function of SaeRS. In strain Newman, a clinical isolate widely used in staphylococcal research, SaeS has an L18P mutation in the first transmembrane domain, which renders this sensor kinase constitutively active [10].
Recently, nonribosomally synthesized dipeptides, named aureusimine A and B, were reported by Wyatt et al. to increase staphylococcal virulence by activating the expression of virulence genes, including those encoding alpha-hemolysin, coagulase, and fibronectin binding proteins [15]. The dipeptides are synthesized by the products of two genes, ausA (NWMN_0123) and ausB (NWMN_0124) (Fig. 2A). In the report by Wyatt et al., the ausA deletion mutant of strain Newman showed reduced hemolysis on blood agar and attenuated virulence in a murine abscess formation model [15]. Intriguingly, almost all (39 out of 40) virulence genes reported to be affected by the ausA mutation in that study are known to be regulated by the SaeRS TCS [15]. In particular, activities of the hla (alpha-hemolysin) and sae P1 promoters were severely reduced, suggesting that the SaeRS TCS is not functional in the ausA mutant. This apparent convergence of regulons implies that aureusimines might be the activation ligands for SaeS. However, since the SaeS sensor kinase is constitutively active in strain Newman [10], exactly how the ausA mutation (i.e., the absence of the dipeptides) could abolish SaeRS-dependent signaling is puzzling. The aim of the present study was to examine whether the aureusimine dipeptides are the activation signals for SaeRS. In the process of addressing this question, we found that the original ausA mutant contains an inadvertent mutation in saeS and, therefore, the observed involvement of aureusimines in staphylococcal virulence was due to that mutation in saeS.

Ethics Statement
The animal study presented in this report (Approval ID: NW-20) was approved by the Institutional Animal Care and Use Committee at Indiana University School of Medicine-Northwest, which complies with the guidelines of the National Institutes of Health.
Bacterial strains, plasmids and culture conditions. The bacterial strains used in this study are listed in Table 1. Staphylococcus aureus cells were grown in tryptic soy broth (TSB). For routine culture of transposon mutants, erythromycin (10 mg/ ml) was added, while chloramphenicol (5 mg/ml) was added to cells containing the lacZ reporter plasmid pCL-P1-lacZ [16].
Generation of sae-deletion mutant (NMDsae). The sae operon was deleted from the genome of S. aureus strain Newman by the allelic replacement plasmid pKOR1 as reported previously [17]. For the deletion, flanking DNA fragments (1 kb each) were PCR-amplified from the genome with the following primers: P667

Chemical synthesis of aureusimine A
The synthesis of aureusimine A was adapted from the previously reported synthesis of Phevalin 2 [18]. Details of the synthesis can be found in Supporting Information (Fig. S1).

HPLC analysis of culture supernatants
Preparation and HPLC analysis of organic cell extracts followed the procedures described by Wyatt et al. [15] with a few modifications. Briefly, overnight cultures of S. aureus Newman and the transposon mutant of ausA (WNJ-12202) were diluted 1:100 into 1 L of fresh TSB containing nalidixic acid (10 mg/ml) and grown for 3 days on a rotary shaker (37uC, 175 rpm). The cells and broth were extracted with 1 L of ethyl acetate twice, then evaporated and dissolved in 1 ml of methanol. Aliquots of 0.25 ml of the concentrated extract were analyzed by HPLC with a C18 column (Higgins Analytical Proto 200). The flow rate of the mobile phase was 3 ml/min. Compounds were eluted using a linear gradient from 20% to 80% acetonitrile in 0.1% TFA over 40 min. Eluted compounds were detected by UV absorbance at 280 nm and 322 nm. In the analysis, aureusimine A and B eluted at 19.66 min and 26.82 min, respectively.
PCR characterization of the transposon mutant. The wild-type Newman and the transposon mutant WNJ-12202 strains were grown in 3 ml of TSB at 37uC overnight, and chromosomal DNAs were purified with a Zippy DNA purification kit (Zymo Research). To detect the transposon insertion, the primers PM1 (59-TTTTAACAATTGGTGCTAGCATGC-39) and PM2 (59-TTTATGGTA CCATTTCATTTTCCTGCTTTTTC-39) were used. To amplify the 59 part of ausA, we used the primers PM1 and PM3 (59-TTAGC GAGTTTAACC CAATACGC-39). The primers were annealed to chromosomal DNA at 55uC for 15 seconds, and the DNA was amplified by the enzyme Phusion (New England Biolabs) at 72uC for 1 min with 30 cycles.
Construction of plasmid pCL55-sae. The entire saePQRS operon was PCR-amplified from the genome of S. aureus strain Newman using Phusion polymerase (New England Biolabs) with the primers: P671 (59-AAC GAATTC TTGGTACTTG-TATTTAATCGTCTATC-39)/P783 (59-AAACTT CCGCGG TTATGACGTAATGTCTAATTTGTG-39). The PCR product was cut with EcoRI and ligated into the single-copy integration plasmid pCL55 [19] cut with EcoRI and SmaI. The ligated plasmid was transformed into Escherichia coli DH5a. To confirm that no mutation was introduced during PCR amplification, the resulting plasmid pCL55-sae was sequenced; no mutation was found in the cloned saePQRS operon.
Cell fractionation and Western blot hybridization. Cells from an overnight culture were diluted in TSB containing appropriate antibiotics, and grown with shaking at 37uC to the stationary growth phase (OD 600 = 6). All cultures were normalized by measuring optical density (OD 600 = 0.6). Supernatant and cells were separated by centrifugation. Proteins in the culture supernatants were precipitated by trichloroacetic acid (10% final concentration) on ice for 20 min, washed with 1 ml acetone, and then air-dried. The dried proteins were suspended in 20 ml 16 SDS-PAGE sample buffer (40 mM Tris HCl, pH 6.8, 2% SDS, 2 mM beta-mercaptoethanol, 4% glycerol, 0.01% bromophenol blue). Cells collected by centrifugation were suspended in 50 ml of Tris HCl (pH 8.0); then 2 ml lysostaphin (2 mg/ml) was added. After incubation at 37uC for 30 min, 50 ml of 26 SDS-PAGE sample buffer was added. The supernatant sample was used to detect alpha-hemolysin, while the cell pellet sample was used to detect SaeP. The samples were separated by 15% SDS-PAGE and the proteins were transferred onto a nitrocellulose membrane (0.45 mm, Whatman). The membrane was blocked with 10% skim milk and incubated with primary antibody (either alpha-hemolysin antibody from rabbit or SaeP antibody from chicken) for 1 h at room temperature; then the blot was incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (either anti-rabbit IgG or anti-chicken IgG). Signals were detected by a luminal enhancer solution detection kit (Thermo). The alphahemolysin antibody was purchased from Sigma-Aldrich. The SaeP antibody was generated by Genscript. LacZ assay. Cells carrying the plasmid pCL-P1-lacZ [16] were grown overnight and diluted in fresh media containing chloramphenicol (5 mg/ml). The cells were further grown with shaking at 37uC to mid-log phase (OD 600 = 0.8). Then cells were harvested, washed with AB buffer (60 mM K 2 HPO 4 , 40 mM KHPO 4 , 100 mM NaCl, pH 7.0), and suspended in 100 ml of the same buffer. After 5 ml lysostaphin (2 mg/ml) was added, the cell suspension was incubated for 15 min at 37uC and then mixed with 900 ml of AB buffer containing 0.1% (v/v) Triton X-100. The bgalactosidase assay was performed at room temperature. As a substrate, 4-methyl umbelliferyl b-D galactopyranoside (MUG, Sigma) was used in the hydrolysis reaction, which was read at 366 nm excitation and 445 nm emission wavelengths.
Murine model of abscess formation. The effect of the ausA mutation on staphylococcal virulence was examined as described previously [20]. Briefly, after the weight of the mice was measured, 1610 7 cfu (colony forming unit) of Newman wild type and the ausA transposon mutant WNJ-12202 were administered to 10 Balb/c mice (Harlan, USA) via retro-orbital injection. Four days after the injection, the mice were sacrificed and their organs (heart, kidney, liver, and spleen) were harvested. The harvested organs were homogenized; then the cfu of bacteria in the organs was measured using serial dilutions on TSA plates.
Antibiotic resistance test. Antibiotic resistances of the sae deletion mutant and the wild-type Newman strains were compared by measuring growth with Bioscreen C and EZExperiment software (Growth Curves USA). In brief, approximately 2610 5 cfu of bacteria was inoculated into 200 ml of TSB containing a test antibiotic and incubated for 24 h without shaking at either 37uC or 43uC. The highest concentration of the test antibiotic was 10 mg/ml, which was serially diluted using 1:2 dilution steps. For data stability, the bacterial growth was measured by the optical density (OD) at 420 nm-580 nm. The presented data are the mean of quadruplicate measurements.
Environmental effects on growth. The transposon insertion mutants of geh (glycerol ester hydrolase) and saeR were grown in 3 ml TSB containing 10 mg/ml of erythromycin (TSB erm10 ) at 37uC overnight. The next day, cells were collected by centrifugation and suspended in phosphate-buffered saline (PBS) to OD 600 = 1. The cell suspension (500 ml) was inoculated into 500 ml TSB and incubated under various conditions (37uC vs. 43uC, +/2 erythromycin, +/2 aeration at 250 rpm). Cell growth was determined over 24 h by measuring OD 600 .
Enrichment test. The transposon insertion mutants of geh and saeR were prepared as described above. The cell suspensions (1.5 ml each) were mixed and 500 ml of the mixed culture was inoculated into 500 ml TSB either with or without erythromycin (10 mg/ml). The cultures were grown at 43uC for 16 h with shaking (250 rpm). Chromosomal DNA was purified from 1 ml of the initial mixed culture (input) and the final cultures either with or without erythromycin. The changes of each strain in the final cultures were analyzed by quantitative real time PCR (qPCR) using the following primers: gyrB-1 (59-TTGTCAGATGT-AAACAACACGG-39), gyrB-2 (59-GTCCGTTATCCGTTAC-TTTAATCC-39), Tn-F(59-TTTATGGTA CCATTTCATTT-TCCTGCTTTTTC), geh-1 (59-GCAACCA TATTGTTATG ACCACC-39), and saeR-1(59-CAACCAGTTGAACAACT-GTCG-39). The primer pair gyrB-1/gyrB-2 was used to normalize DNA sample loading, while the primer pairs Tn-F/ geh-1 and Tn-F/saeR-1 were used to assess changes of the geh and saeR transposon mutants in the final cultures. The DNA amplifications were performed with a 7900HT Fast Real Time PCR system (Applied Biosystems) using the following conditions: 94uC 30 sec, 55uC 30 sec, 72uC 30 sec, 40 cycles. The change of each strain was calculated with the following formula: where Ct SF = Ct of test gene (i.e. either geh or saeR) in the final cultures; Ct SI = Ct of test gene in the initial mixed culture; Ct GF = Ct of gyrB in the final cultures; Ct GI = Ct of gyrB in the initial mixed culture. The enrichment index was calculated with the following formula: Enrichment Index = (Fold Change of saeR mutant)/(Fold Change of geh mutant).

Results
The ausA transposon mutant WNJ-12202 does not produce aureusimines We used the ausA transposon mutant WNJ-12202 from the phoenix library collection [21], which was kindly provided by Drs. Olaf Schneewind and Dominique Missiakas (University of Chicago), to test whether aureusimines are required for the production of virulence factors. The WNJ-12202 mutant strain has a transposon insertion 36 bp downstream of the putative ATG start codon of ausA ( Fig. 2A). To eliminate any fortuitous mutations acquired during the mutagenesis process, the transposon insertion was transduced into strain Newman with the staphylococcal phage w85. To confirm the transposon insertion mutation of ausA, we subjected the mutant chromosomal DNA to PCR-amplification with the primer pairs PM1/PM2 or PM1/ PM3 ( Fig. 2A). Since primer PM2 binds to the bursa aurealis transposon sequence used for construction of the Phoenix library, a PCR product with the primer pair PM1/PM2 can be amplified only from genomic DNA of the transposon mutant. On the other hand, since both PM1 and PM3 primers recognize ausA sequences, PM1/PM3 PCR products are expected to be amplified from genomic DNA of both the wild type and the transposon mutant strains, but with different sizes. The size of the product amplified from the mutant DNA is expected to be 3.2 kb larger than that from wild-type DNA. Accordingly, using primer pairs PM1 and PM2, and PM1 and PM3, PCR products with the expected sizes of approximately 800 bp, or 3.8 kb, respectively, were amplified using DNA from the transposon mutant (Fig. 2B). The latter PCR product with a molecular weight of 3.8 kb corresponds to the sum of the ausA (621 bp) and the transposon (3.2 kb) sequences. The PM1/PM2 PCR product obtained from the mutant strain DNA was sequenced, confirming the transposon insertion site (data not shown). To verify that the ausA mutant does not produce aureusimines, we compared the culture supernatants from the wild-type and ausA mutant strains by organic extraction and high performance liquid chromatography (HPLC) [15]. We chemically synthesized aureusimine A as standard (Fig. S1). While two peaks corresponding to aureusimines A and B were detected in the supernatant from the wild-type strain Newman, those peaks were not detected in the culture supernatant from the transposon mutant (Fig. 2C), thus confirming the result of Wyatt et al. that ausA encodes the dipeptide synthesis enzyme [15]. As expected, the elution time of the chemically synthesized aureusimine A corresponded to that of the first peak, interpreted as aureusimine A. These results demonstrate that the transposon mutant WNJ-12202 does not produce aureusimines and therefore represents an appropriate tool for the study of the role of aureusimine dipeptides.
SaeRS is functional in the ausA transposon mutant WNJ-12202 Next, we tested whether the aureusimine dipeptides are required for SaeRS signaling using the mutant strain WNJ-12202. First, because SaeRS is critical for alpha-hemolysin production, we examined the hemolysis pattern of the mutant on blood agar. The tested strains were streaked against the strain RN4220, which produces only beta-hemolysin [22]. To our surprise, the hemolysis pattern of the ausA transposon mutant was identical to that of wild type (see the white arrow heads in Fig. 3A), whereas the sae deletion mutant (Dsae in Fig. 3A) showed a clear defect in hemolysis. Furthermore, wild-type level hemolysis was restored by insertion of the complementation plasmid pCL55-sae [Dsae(pCL55-sae) in Fig. 3A] in the sae deletion mutant. These results strongly indicate that ausA is not involved in controlling expression of alpha-hemolysin, which is at variance with the results obtained by Wyatt et al. [15].
To further test whether the ausA mutant produces alphahemolysin, we used Western blot analysis. In addition, we examined expression of SaeP, another well-known gene product of the sae operon. Consistent with the hemolysis patterns on blood agar, the ausA transposon mutant WNJ-12202 produced both alpha-hemolysin and SaeP at the wild-type level (Fig. 3B), suggesting that SaeRS signaling is intact in the ausA mutant. When transcription from the P1 promoter of the sae operon was examined using a P1-lacZ fusion, no significant difference was observed between the wild type and the mutant strain (Fig. 3C), further confirming the normal functionality of SaeRS in the ausA transposon mutant. These data demonstrate that ausA and the biosynthetic products of the ausA locus, the aureusimines, are not required for SaeRS signaling.

Aureusimines do not contribute to staphylococcal virulence
Given that our findings demonstrate that aureusimines are not required for SaeRS signaling and SaeRS-dependent alphahemolysin production, we further tested the contribution of aureusimines to bacterial virulence. Equal numbers of cells of the wild-type Newman and transposon mutant WNJ-12202 strains were injected into mice; then effects on disease development were analyzed in the same way as in the Wyatt et al. study [15]. Bacterial numbers in all tested organs were not significantly different between mice infected with the wild-type or mutant strains (Fig. 4). In addition, no significant difference was observed in weight loss, showing that aureusimines are not required for staphylococcal virulence in this murine infection model.

The original ausA deletion mutant contains an additional mutation in saeS
Since we failed to observe a role of aureusimines in hemolysin production or in-vivo virulence, we suspected that the original ausA deletion mutant might have contained additional mutation(s), especially in the saeRS genes. Therefore, we examined the saeRS region of the original mutant used by Wyatt et al. by DNA sequencing, which revealed that the original mutant has an 83-nt duplication in saeS (Fig. 1). Furthermore, the hemolysis defect of the original ausA mutant on blood agar was fully complemented by the plasmid carrying the sae operon [DausA(pCL-sae) in Fig. 3A]. In addition, when transduced into strain Newman, the resulting ausA deletion mutant showed the wild-type hemolysis pattern ('DausA, transduced' in Fig. 3A). These results demonstrate that the virulence defects of the ausA deletion mutant in the Wyatt et al. study were caused by the inadvertent mutation in saeS that we discovered.
The sae deletion mutant is more resistant to certain antibiotics Next, we examined whether acquisition of the additional mutation in saeS may have resulted from conditions during mutagenesis that selected for the mutant. Since mutagenesis often involves growing S. aureus cells in the presence of antibiotics such as chloramphenicol and erythromycin, if mutations in the saeRS system render cells more resistant to these antibiotics, saeRS mutants would be enriched during the mutagenesis process. We tested this hypothesis by comparing the growth of wild-type Newman and sae deletion mutant strains in the presence of various antibiotics. At both 37uC and 43uC, the sae deletion mutant showed higher resistance to chloramphenicol and erythromycin (Fig. 5A). Although to a lesser extent, the mutant was also more resistant to spectinomycin at 43uC and kanamycin at 37uC (data not shown). Furthermore, because all of the antibiotics tested are protein synthesis inhibitors, we investigated whether the sae deletion mutant is also more resistant to other classes of antibiotics. At 37uC, the sae-deletion mutant showed a low level of resistance to ciprofloxacin, a DNA gyrase inhibitor (Fig. 5B). However, the mutant was more sensitive to cell lysis caused by oxacillin, a cellwall synthesis inhibitor. Both wild-type and the sae deletion mutant strains were equally sensitive to rifampicin, an RNA synthesis inhibitor, and to tetracycline, another protein synthesis inhibitor (data not shown). When complemented by the plasmid pCL55-sae, the resistance to erythromycin was abolished (Fig. 5C), confirming that increased resistance was due to the deletion of the sae operon.
A sae mutant has a growth advantage at high temperature, aeration, and presence of erythromycin The moderate resistance to erythromycin and chloramphenicol strongly suggests that sae mutants can be enriched during mutagenesis when employing antibiotics. However, since in a typical mutagenesis process (e.g., deletion or transposon insertion) the target strains already have plasmid-conferred antibiotic resistance, the comparison between the wild-type and sae deletion mutant strains shown in Fig. 5 might not be directly relevant to the mutagenesis process. To examine whether sae mutants have a growth advantage and can be enriched even with a strain that is already resistant to antibiotics, we used two transposon mutants, NM::geh and NM::saeR. Due to the presence of an erythromycin resistance gene in the transposon, both strains are fully resistant to erythromycin (i.e., minimal inhibitory concentration .100 mg/ml). NM::geh has a transposon insertion in geh (glycerol ester hydrolase) while NM::saeR has a transposon insertion in saeR. Since geh is already disrupted by the lysogenic phage wNM4 [23] in strain Newman, further disruption by the transposon insertion in geh is not expected to affect bacterial gene expression, physiology, or virulence.
The mutants were incubated in TSB containing 10 mg/ml of erythromycin (TSB erm10 ) overnight at 37uC; then equal numbers of cells were inoculated into TSB erm10 and incubated with aeration (250 rpm) at either 37uC or 43uC. The growth advantage was pronounced at 43uC, at which temperature NM::geh showed a diphasic growth pattern (Fig. 6). The doubling time was 37 min for NM::saeR while it was 113 min (4 hr-8 hr) or 44 min (9 hr-12 hr) for NM::geh. On the other hand, at 37uC, although NM::geh showed a little longer lag phase, the growth rate at mid-log phase was almost identical (doubling time for NM::saeR, 41.3 min; NM::geh, 40.8 min). Interestingly, when we repeated the growth test at 43uC in the absence of either erythromycin or aeration, the pronounced growth advantage of NM::saeR over NM::geh mutant was greatly diminished (Fig. 6, bottom panel). Without erythromycin, the doubling time for NM::saeR was 30.6 min, while it was 38.9 min for NM::geh. These results suggest that high temperature, presence of erythromycin, and aeration contribute to the growth advantage of the sae mutant.

Sae mutants can be enriched under typical staphylococcal mutagenesis conditions
In a typical staphylococcal mutagenesis procedure, the target cells are incubated at 43Cu in the presence of an antibiotic and aeration, which according to our results favors growth of sae mutants (Fig. 6). To examine the extent of the enrichment of the sae mutant under those growth conditions, an equal number of NM::geh and NM::saeR cells were combined and incubated for 16 h at 43uC with aeration in the presence of erythromycin (10 mg/ml). For comparison, we also grew the mixed cells in the absence of erythromycin. To assess the relative ratio of the mutants in the resulting cultures, we purified chromosomal DNA and performed quantitative real-time PCR (qPCR), amplifying parts of the transposon sequence, and geh or saeR. While, in the absence of erythromycin, the NM::saeR strain was enriched approximately 20 fold, the enrichment was increased to approximately 700 fold in the presence of erythromycin (Fig. 7). These results demonstrate that sae mutants can be enriched to a great extent under conditions typically employed for staphylococcal mutagenesis.

Discussion
The SaeRS TCS is an important signaling system controlling the production of multiple virulence factors and in vivo survival of S. aureus [4,5,8]. However, how the signaling system is activated remains unknown. Although results from the Wyatt et al. study raised the possibility that the dipeptide aureusimines could be the activation signal for SaeRS [15], we here demonstrate that the results obtained by Wyatt et al., indicating an involvement of the dipeptides in SaeRS signaling, were due to an inadvertent mutation in saeS. We suspect that the acquisition of the saeS mutation was facilitated by the increased resistance of sae mutants to environmental stresses such as high temperature, aeration, and the presence of erythromycin, the antibiotic used for the generation of a single cross-over during the deletion of ausA [15].
The sae deletion mutant showed a moderate level of resistance not only to erythromycin but also to chloramphenicol, another antibiotic widely used in staphylococcal research (Fig. 5A). Both antibiotics inhibit protein synthesis by binding to the 50S subunit of bacterial ribosome [24,25]. Although the exact steps inhibited by these two antibiotics are different, the binding sites of the antibiotics overlap [25]. Therefore, it is possible that the same change caused by the saeRS TCS mutation is responsible for the elevated resistance to the antibiotics. However, the nature of that change is not yet clear. One can hypothesize that some genes in the sae regulon might be directly involved in the protein translation Figure 6. An sae mutant has a growth advantage at high temperature, aeration, and presence of erythromycin. The transposon mutants of saeR and geh (glycerol ester hydrolase) were grown under various conditions and the effects of three environmental factors (temperature, the presence of erythromycin, and aeration) on the growth pattern of the mutants were examined. For clarity, only the altered environmental factor is indicated above the graph being compared. In the test, erythromycin was added at 10 mg/ml and aeration was achieved by shaking at 250 rpm. Erm; erythromycin. doi:10.1371/journal.pone.0015703.g006 Figure 7. Sae mutants can be enriched under typical staphylococcal mutagenesis conditions. Transposon insertion mutants of geh (glycerol ester hydrolase) and saeR were used in this test. Equal numbers of the two mutants were mixed and grown either in the absence or in the presence of erythromycin (10 mg/ml) with shaking for 16 h at 43uC. Chromosomal DNA was purified from each culture and the fold change of each strain was calculated by quantitative real time PCR (qPCR) as described in Materials and Methods. Bar graphs indicate the fold change of each strain. Data are presented as the mean 6 standard deviation from three independent experiments. Input, the mixed culture used for inoculation; -Erm, culture grown in the absence of erythromycin; + Erm, culture grown in the presence of erythromycin. doi:10.1371/journal.pone.0015703.g007 process and can affect ribosomal sensitivity to the antibiotics. Among the genes that contain SaeR-binding sequences in the upstream region, six are involved in protein translation: peptide chain release factor-2 (prf), 30S ribosomal proteins S20 and S21, alanyl-tRNA synthetase, glutamyl-tRNA synthetase, and histidyl-tRNA synthetase [16]. It remains to be determined if these or other factors contribute to the resistance to certain antibiotics in the saeRS mutants.
Given that the synthesis of many exoproteins is decreased in sae mutants, it is not surprising to observe a measurable growth advantage of the saeR mutant over the geh mutant (Fig. 6). Intriguingly, the growth advantage of sae mutants was greatly enhanced by the combination of three environmental stresses: high temperature, the presence of erythromycin, and aeration. A recent study also showed that sae mutants are 10 times more resistant to the oxidative stresses imposed by H 2 O 2 [26], which is known to activate the SaeRS TCS [11]. These results raise the question of why sae mutants are more resistant to those stresses. It is possible that certain gene products of the sae regulon might exacerbate or amplify the growth inhibitory effects of those stresses. However, because not all genes belonging to the sae regulon are yet verified experimentally, it is difficult at this point to identify candidates for the gene products responsible for the exacerbated growth inhibition.
Interestingly, unlike sae mutants, mutants of the agr quorumsensing system, another TCS important for staphylococcal virulence, seem to be enriched in the absence of environmental stresses [27]. In fact, due to the genetic instability of agr, global regulatory roles have been bestowed incorrectly to several genes such as xpr, svrA, traP, and lukSF-PV. The genetic locus xpr (more correctly the mutation D 1058::Tn551) was identified, in a Tn551 mutagenesis, to encode a global regulator required for the expression of lipase, alpha-toxin, delta-toxin, protease, and nuclease [28]. However, later, all of the expression defects of the xpr mutant were found to be caused by an additional mutation in agrC, the gene encoding agr sensor kinase [29]. The gene svrA was identified in a signature-tagged mutagenesis (STM) screen as a positive regulator for agr expression because the svrA mutation abolished expression of agr [30,31]. But the absence of agr expression in the svrA mutant was later attributed to a frame-shift mutation in agrC [32]. Although the traP gene was initially reported to activate the agr system via a pathway distinct from the autoinducing peptide (AIP) pathway [33], Adhikari et al showed that the traP mutant contains a stop codon in agrA, the gene encoding the response regulator, which, in turn, explains the lack of agr activity in the traP mutant [22]. Finally, a PVL phage strain of S. aureus producing Panton-Valentine Leukocidin (PVL), a poreforming toxin encoded by lukSF-PV, was reported to exhibit an agrnegative phenotype (i.e., reduced expression of multiple exoproteins and enhanced expression of surface proteins such as SdrD and Spa) [34]. However, Villaruz et al. demonstrated that the PVL phage strain used in the original study carried a point mutation at the AgrA binding site in agr P2 promoter; therefore, the agrnegative phenotype of the PVL phage strain was caused by the point mutation in the agr P2 promoter, not by expression of PVL [35]. Together with our report on the saeS mutation in the ausA mutant, these examples clearly show that great care should be taken during genetic manipulation of S. aureus in order to prevent acquisition of secondary mutations, especially in saeRS and agr, and appropriate genetic complementation experiments used.
Currently, the use of antibiotics during mutagenesis of staphylococcal genes is inevitable. Since sae mutants show elevated resistance to chloramphenicol, erythromycin, and, to a lesser extent, kanamycin, spectinomycin, and ciprofloxacin, when one of these antibiotics is used, it would be prudent to confirm that the resulting mutant does not contain an inadvertent mutation in the sae operon. For confirmation, a convenient method is using a hemolysis assay on blood agar. In particular, if the test strain is streaked on sheep blood agar against RN4220 as shown in Fig. 3A [22], the enhancement of b-hemolysis by delta-hemolysin at the strain junction can also reveal the functionality of the agr quorumsensing system. This is of particular importance because agr mutants can arise spontaneously in the laboratory as well as in clinical settings [22,27,29,32,35,36,37,38]. Since the sae mutants do not seem to be more resistant to tetracycline, if a choice is given, usage of tetracycline can minimize the enrichment of sae mutants.
Since our results showed that aureusimines do not play a role in sae-mediated virulence factor production or virulence in animal models, it remains to be determined what roles these dipeptides play in S. aureus. Intriguingly, the upstream sequence of ausA contains an imperfect SaeR binding sequence (ATTAA-GAATTTGTTAA), indicating that the expression of ausAB genes might be regulated by saeRS TCS [16]. Certainly, more work is required to identify the role(s) of these dipeptides in S. aureus.