|Home | About | Journals | Submit | Contact Us | Français|
Multistep genetic alteration is required for methicillin-resistant Staphylococcus aureus (MRSA) to achieve the level of vancomycin resistance of vancomycin-intermediate S. aureus (VISA). In the progression of vancomycin resistance, strains with heterogeneous vancomycin resistance, designated hetero-VISA, are observed. In studying the whole-genome sequencing of the representative hetero-VISA strain Mu3 and comparing it with that of closely related MRSA strains Mu50 (VISA) and N315 (vancomycin-susceptible S. aureus [VSSA]), we identified a mutation in the response regulator of the graSR two-component regulatory system. Introduction of mutated graR, designated graR*, but not intact graR, designated graRn, could convert the hetero-VISA phenotype of Mu3 into a VISA phenotype which was comparable to that of Mu50. The same procedure did not appreciably increase the vancomycin resistance of VSSA strain N315, indicating that graR* expression was effective only in the physiological milieu of hetero-VISA cell to achieve a VISA phenotype. Interestingly, the overexpression of graR* increased the daptomycin MICs in both Mu3 and N315 and decreased the oxacillin MIC in N315.
Vancomycin-intermediate Staphylococcus aureus (VISA), first described in 1997, has continuously been a worldwide problem in the treatment of methicillin-resistant S. aureus (MRSA) hospital infection (16). VISA has a unique mechanism of resistance; the resistant cell produces a thickened cell wall, whereby many vancomycin molecules are trapped within the cell wall. The trapped molecules clog the peptidoglycan meshwork and finally form a physical barrier towards further incoming vancomycin molecules (5, 6, 13). VISA does not directly emerge from vancomycin-susceptible MRSA. It emerges from hetero-VISA that expresses heterogeneous-type vancomycin resistance (14). Hetero-VISA spontaneously produces VISA cells within its cell population at a frequency of 10−6 or above (14). Using differential hybridization, we have identified the two-component system vraSR, which is constitutively activated in Mu50 (VISA) and Mu3 (hetero-VISA) but strongly depressed in vancomycin-susceptible S. aureus (VSSA) strain Mu50Ω (18). vraSR turned out to be an up-regulator of cell wall peptidoglycan synthesis (7). Its overexpression in VSSA strain N315 decreases VSSA's susceptibility to the level of hetero-VISA but not to the level of VISA (defined as a vancomycin MIC of ≥4 mg/liter). This study was conducted to search for the second genetic alteration that promotes hetero-VISA-to-VISA phenotypic conversion.
VISA strain Mu50 and hetero-VISA strain Mu3 have been described previously (14). N315, isolated in 1982 in Japan, is a prototype of Japanese hospital-associated MRSA strains with genotype IIA (type IIa SCCmec and multilocus sequence type 5), to which Mu3 and Mu50 belong (12). N315 is a pre-MRSA strain; that is, its expression of methicillin resistance is repressed due to the presence of an intact copy of the mecI gene encoding a mecA gene transcription repressor on the chromosome (21). A collection of 13 VISA strains from six countries was used for graSR sequencing. They were MI, NJ, and IL (isolated from the United States); AMC11094 (Korea); 99/3700 (United Kingdom); LIM2, 98141 (France); 28160 (South Africa); and BR1 to BR5 (Brazil). These strains have been described previously (8).
Plasmid DNA was purified from Escherichia coli by use of an SV Minipreps DNA purification system (Promega Corporation, Madison, WI) according to the manufacturer's instructions. DNA isolation from S. aureus cells, restriction endonuclease digestion, ligation reactions, and DNA cloning were carried out as described previously (7). The electroporation of S. aureus was performed with a Bio-Rad gene pulser with a pulse controller as described previously (21). Recombinant plasmids pgraRn and pgraR* were constructed by cloning into the pYT3 plasmid vector the PCR-amplified graR gene (SA0614) from N315 and Mu50 chromosomes, respectively, by use of the primers (5′-tttttggatccGGATTAAAGATTTTCAAAGTC-3′) and (5′-tttttggatccGAGATTTCAAAAAATAAGCTAC), in which the underlined parts represent added BamHI sites. The integrity of the cloned graR gene was ascertained by sequencing each recombinant plasmid. The resultant plasmids, pgraRn and pgraR*, were then introduced into N315 and Mu3, resulting in strains N315(pgraRn), N315(pgraR*), Mu3(pgraRn), and Mu3(pgraR*). For the complementation of the graSR-knocked-out strain N315ΔgraSR (see below), plasmid pgraSR* was used, which was constructed by cloning the graSR genes (SA0615 and SA0614) of Mu50 into pYT3. The graSR* genes were PCR amplified with the Mu50 genomic DNA as a template by use of primer sets 5′-tttttggatccTACATCTATACGATTATATC-3′ and 5′-tttttggatccACATATGACTAACATCTATC-3′, in which the underlined parts represent added BamHI sites.
The graSR-null knockout mutants of N315 and Mu3 were constructed by an allelic exchange method as described previously (7), with the exception that the 1.5-kb and 1-kb fragments of the left and right respective junctions of graSR were PCR amplified and cloned into the upstream and downstream regions of the cat gene of chloramphenicol resistance cassette vector pUC118cat (7). Since Mu3 is tetracycline resistant, the graSR knockout mutant of this strain was generated by the following protocol, also described by Bae and Schneewind (1). Briefly, the sequences flanking graSR were PCR amplified with primers attB1-KO614&5F (ggggacaagtttgtacaaaaaagcaggctGTAACTAAAAGGTGGAGTAA) and attB1-KO614&5R_SacII (TTTTTCCGCGGCCATATCACCCAATATCATT) for the upstream region and primers attB2-KO614&5F_SacII (TTTTTCCGCGGACATGCGTTTTGTTACTTAG) and attB2-KO614&5R (ggggaccactttgtacaagaaagctgggtGATAGATGGCATAATGTGAT) for the downstream region. The PCR fragments were then ligated in vitro and recombined with pKOR1 (1). The pKOR1 recombinant was then introduced into S. aureus Mu3 by electroporation. Selection of the knockout strains was performed as described previously (1). The successful deletion of graSR genes was verified by sequence determination.
Mu3 genome sequencing was performed as described previously (20), with shotgun sequencing with 6.4-fold coverage by use of a random small-insert library (1.5 to 2.5 kb). We used the chromosome sequence of Mu50 as a scaffold to assemble and orient Mu3 contigs, followed by gap closure using long-range PCR primer walking. We then compared the sequences of Mu3 and Mu50 chromosomes for single-nucleotide polymorphisms (SNPs) and mismatches. Identified discrepancies were verified by resequencing of the corresponding loci of Mu3 and Mu50 genomic DNAs. Open reading frame (ORF) prediction and gene annotation were performed as described previously (20). The entire Mu3 genome sequence and annotation have been deposited in DDBJ under accession number AP009324.
Preparation and examination of S. aureus cells by transmission electron microscopy were performed as described previously (9). Morphometric evaluation of cell wall thickness was performed using photographic images at a final magnification of ×30,000. One hundred thirty cells of each strain with nearly equatorial cut surfaces were measured for the evaluation of cell wall thickness, and results were expressed as mean values ± standard deviations.
The MIC determination was performed by agar dilution methods according to CLSI criteria (4). To detect small changes in susceptibility, linear sets of antibiotic concentrations with 1-mg/liter increments were adopted for the MIC determinations for vancomycin, teicoplanin, and daptomycin, whereas the orthodox twofold dilution system was used for oxacillin. Taking care not to miss a slow-growing resistant cell subpopulation of heterogeneous resistance expression, MIC was evaluated not only at 24 h but also at 48 h of incubation time (15).
Bacteria were grown in 10 ml of brain heart infusion (BHI) broth to exponential phase (optical density at 600 nm [OD600] = 0.6) before harvest. Pellets were then suspended with precooled T10E10 buffer (10 mM Tris-HCl [pH 8.0]), followed by the addition of lysostaphin solution (WAKO, Japan) at a 3.0-μg/ml final concentration. The suspensions were then incubated at 37°C for 3 min until complete cell lysis occurred. Immediately, 7 ml of acidic phenol (pH 5.2, equilibrated with 20 mM sodium acetate) was added, followed by the addition of 600 μl of 3 M sodium acetate. The samples were then frozen (−80°C) and thawed (65°C) three times. Phenol-chloroform extraction and ethanol precipitation was then carried out. After that, the resulting RNA pellet was subjected to digestion with RNase-free DNase I (Roche, Mannheim, Germany) at 37°C for 30 min. The RNA samples were then purified again with phenol-chloroform extraction and ethanol precipitation. Pellets were then resuspended in 25 μl of diethyl pyrocarbonate-treated water. The construction and analysis of the DNA microarray have been previously described (7). To confirm reproducibility, RNA extraction and hybridization were performed in duplicate for each experiment. In addition, extractions of up-regulated and down-regulated genes in the Mu50/Mu3 combination were done using three independent experimental data to ensure a high level of credibility for the results. The values for transcription rates given in the Mu50/Mu3 array study are the averages of the three experiments.
Growth curve and autolysis experiments were carried out in succession. An overnight culture of a tested strain was diluted 1/1,000 in 10 ml fresh BHI broth and grown at 37°C with 25-rpm shaking in a photorecording incubator (TN-2612; Advantec, Tokyo, Japan). The OD was monitored automatically every 2 min, and cells were grown to an OD600 of 1.2. Following this, the cells were cooled on ice for 10 min and then pelleted by centrifugation at 7,500 rpm for 5 min. Cells were then washed twice and resuspended with chilled 0.1 M phosphate-buffered saline buffer (pH 7.0). The cell suspension was then incubated at 30°C with continuous shaking at 25 rpm. The decrease in optical density was monitored every 2 min for 24 h with the same photorecording incubator as described above. For growth curve and doubling time determination, values for OD versus time in the exponential growth phases of each strain were plotted. Doubling times were then calculated as follows: doubling time = [(t2 − t1) × log 2]/[log OD600 at t2 − log OD600 at t1].
We determined the whole-genome sequence of hetero-VISA strain Mu3 and compared it with those of Mu50 and N315. (The whole-genome sequence of Mu3 is available under accession no. AP009324 ). Table Table11 summarizes the differences in the nucleotide sequences of Mu3, N315, and Mu50 chromosomes. Among 16 SNPs and three mismatches, we found a Mu50-specific missense mutation in graR, a two-component response regulator gene (Table (Table1).1). The gene was also present in the list of 17 genes whose increased transcription raises vancomycin resistance (7).
The graR mutation replaces the 197th Asn of GraR with Ser. Besides Mu50, comparative sequencing of graSR from 13 clinical VISA strains obtained from six countries identified two other missense mutations in 10 of the 13 strains: Ser79 to Phe (strain NJ) and Asp148 to Gln (strains 99/3700-W, LIM2, 98141, 28160, and BR1 to BR5). The other three strains, MI, IL, and AMC11094, shared the same graR with N315 and Mu3. In contrast, no mutation was found in the graS genes of the 13 strains. This frequent mutation in graR in VISA clinical strains prompted us to test the effect of mutation on vancomycin resistance.
To test the role of the mutation, we PCR amplified and cloned the mutated graR gene, designated graR*, and intact graR, designated graRn (where “n” stands for N315), into shuttle vector pYT3, obtaining pgraR* and pgraRn. These plasmids were then introduced into Mu3 and N315 by electroporation, and MICs were evaluated (Table (Table2).2). MICs were determined by an agar dilution method according to the CLSI guidelines (4), with an additional condition, i.e., MIC reading after 48 h of incubation (Table (Table2).2). The introduction of graR* into Mu3 significantly increased the vancomycin MIC from 2 to 5 mg/liter (in BHI agar), which was comparable to that of VISA strain Mu50 (Table (Table2).2). The introduction of graR* also increased the daptomycin MIC from 2 to 5 mg/liter, which was also comparable to that (6 mg/liter) of Mu50 (Table (Table2).2). In contrast, the introduction of the intact graR (graRn) into Mu3 did not raise vancomycin or daptomycin MICs appreciably. No significant increase was observed in vancomycin and teicoplanin MICs when graR* was introduced into N315 (Table (Table2).2). However, there was a slight decrease in the oxacillin MIC and a slight increase in the daptomycin MIC. This curious phenomenon was not apparent with the introduction of graRn (Table (Table22).
Since N315 and Mu3 have intact graSR genes on the chromosome, deletion mutant strains N315ΔgraSR and Mu3ΔgraSR were constructed for the confirmation of the role of the two-component system in antibiotic resistance. The graSR deletion of Mu3 caused small decreases in the vancomycin and daptomycin MICs and rather significant decreases in the teicoplanin MICs (Mu3ΔgraSR [Table [Table2]).2]). However, no evident change was observed in the oxacillin MIC. On the other hand, the graSR knockout of N315 (N315ΔgraSR) caused a small increase in the oxacillin MIC, while it did not affect the glycopeptide or daptomycin MICs appreciably (Table (Table2).2). Introduction of the plasmid pgraR* into Mu3ΔgraSR raised its glycopeptide and daptomycin MICs to a level comparable to that seen for Mu3(pgraR*) (Table (Table2).2). On the other hand, the oxacillin and daptomycin MICs of N315ΔgraSR were significantly affected downwards and upwards, respectively, with the introduction of pgraR*. A similar phenotypic change was observed for N315ΔgraSR(pgraSR*), in which graR* was introduced together with graS (Table (Table2).2). These data clearly indicated that the effect of introduced graR* was not dependent on the presence of GraS in the cell.
Transmission electron microscopy revealed that Mu3(pgraR*) showed remarkable cell wall thickening which was comparable to that of Mu50 (Fig. (Fig.1).1). In contrast, only a small degree of cell wall thickening was observed with Mu3(pgraRn).
Mu3(pgraR*) also resembled Mu50 in terms of reduced autolytic activity (Fig. (Fig.2)2) and prolonged doubling time in drug-free medium (Fig. (Fig.3).3). Mu3(pgraRn) also showed a slight reduction in autolytic activity and a prolonged doubling time. These changes, however, were not as significant as those seen for Mu3(pgraR*) (Fig. (Fig.22 and and3).3). Interestingly, the graSR knockout of Mu3 led to an enhanced autolytic activity and an accelerated growth rate (Fig. (Fig.22 and and33).
Figure Figure44 shows the results of three representative microarray analyses on graR-relevant derivative strains, including the one using strains Mu3(pgraR*) and Mu3(pgraRn). Figure Figure4A4A shows the fluorescence intensities of all the array spots (left) and a scatter diagram of normalized spot intensities in the study of Mu3(pgraR*)/Mu3(pgraRn) (right). The diagram shows seven spots whose signals were significantly high in strain Mu3(pgraR*) compared to the level for Mu3(pgraRn). They correspond to the seven genes vraFG, vraDE, isaB, mprF(fmtC), and SAS091 (Fig. (Fig.4B,4B, right). On the left of Fig. Fig.4B4B is shown a magnified view of the signal intensities of these genes in three array experiments.
Table Table33 summarizes the results of nine sets of array experiments performed in this study. The listed genes are the top 14 highly expressed genes in Mu50 compared to their expression in Mu3 (up-regulated genes), and the four most repressed genes in Mu50 relative to their expression in Mu3 (down-regulated genes). Criteria for the entry of up- and down-regulated genes were signal ratios of >1.7 and <0.5, respectively. The listed 14 up-regulated genes were those reproducibly found in the top 20 entries of three separate experiments.
As shown in Fig. Fig.44 and in Table Table3,3, 7 of the 14 up-regulated genes were found up-regulated in Mu3 when graR*, but not the graRn gene, was overexpressed. isaB was enhanced in Fig. Fig.44 but was not included in Table Table33 because it was not enhanced in Mu50/Mu3. On the other hand, glnR and SAS091, found in the up-regulated gene list of the Mu50/Mu3 combination, were also found to be up-regulated in the Mu3(pgraR*)/Mu3(pgraRn) combination (Table (Table33).
Comparison of the two array data for Mu3(pgraRn)/Mu3(pYT3) and Mu3(pgraR*)/Mu3(pYT3) clearly showed that the intact graR gene was unable to increase the transcription of these genes [except for fmtC(mprF), which gave a value of 2.38] (Table (Table3).3). A similar repertoire of genes was increased in transcription in the N315(pgraR*)/N315(pgraRn) combination. However, the rates of increase of transcription were small with N315(pgraR*)/N315(pYT3) compared to what was seen for Mu3(pgraR*)/Mu3(pYT3), except that vraDE genes encoding a putative ABC transporter were significantly increased in both combinations. Therefore, the regulatory effect of graR* was considered much stronger in Mu3 than in N315. Deletion of graSR of N315 and Mu3 decreased the transcription of only two of the seven up-regulated genes by graR*. They were fmtC(mprF) and vraD (0.49- and 0.18-fold in Mu3 and 0.67- and 0.57-fold in N315, respectively), but the others (vraFG, vraE, SAS091, and glnR) were not affected much (0.79- to 1.40-fold) (Table (Table3).3). The data indicated that these genes were also regulated by some regulatory genes other than graSR.
The genetic mechanism of vancomycin resistance in VISA is not well understood. Several genes have been proposed as being involved in certain clinical VISA strains (2, 3, 17, 19, 22, 23). Since our first isolation of VISA strain Mu50 and hetero-VISA strain Mu3 from Juntendo University Hospital, we have been trying to understand the genetic process of vancomycin resistance acquisition in the Mu3-Mu50 lineage of strains. Mu3 and Mu50 have in common the characteristic of accelerated cell wall synthesis (11). Activation of the two-component system vraSR, which up-regulates the enzymes of the peptidoglycan synthesis pathway such as PBP2 and the sgtB and murZ products, etc., explains these features common to Mu3 and Mu50 (18). In fact, Mu3 and Mu50 share a common missense mutation in the vraS gene in the region corresponding to the N-terminal region (Asn5 → Ile), and its transcription is constitutively increased in both strains (18, 19) (K. Hiramatsu, unpublished observation). In this study, we identified a mutation in another two-component system, graSR. The mutation was located in the response regulator graR, and its expression in Mu3 caused remarkable cell wall thickening and an increase in the vancomycin MIC to the level seen for VISA strain Mu50. The overexpression of graR* in N315 did not cause that significant effect on the level of vancomycin resistance or cell wall thickening (data not shown). This host difference in the expression of vancomycin resistance may well be explained by the fact that the enzymes of cell wall synthesis, such as PBP2 and the sgtB and murZ products, etc., are more abundant in Mu3 than in N315 due to the activated vraSR system in Mu3. Therefore, the mutated graR seems to confer the VISA phenotype only to cells that are producing cell wall peptidoglycan at an enhanced rate.
With regard to the physiological function of graR*, our microarray experiments showed increased transcription of two sets of ABC transporter genes: vraDE and vraFG. It is therefore a tempting hypothesis to suggest that graR* activates the membrane traffic of substrates for the cell wall synthesis enzymes to help them produce thickened cell wall peptidoglycan. Alternatively, graR* might negatively regulate autolytic activity, thus allowing the cell wall peptidoglycan accumulate on the cell surface. This is also a plausible hypothesis, since deletion of graSR drastically increased autolytic activity in Mu3 (Fig. (Fig.22 and and3).3). It was also noted that fmtC(mprF) was among the seven genes whose transcription was significantly increased in Mu3(pgraR*) (Fig. (Fig.44 and Table Table33).
Enhanced expression of the gene product lysylphosphatydylglycerol synthetase might contribute to vancomycin resistance by increasing the cell membrane positive charge and repulsing the positively charged vancomycin molecules (24, 25). The decreased growth rate of Mu3(pgraR*) compared to Mu3 may be a consequence of the reduced autolytic activity mediated by graR* expression. More-detailed experiments would be needed to identify the regulatory role of the mutated as well as the intact graSR two-component system in the hetero-VISA-to-VISA conversion.
The graR mutation replaces Asn197 of GraR with Ser. In view of the signal transmission mechanism of the bacterial two-component system, the conformational change of the GraR protein by the mutation might have activated its response regulator function without signal input from the signal transducer GraS. This is predicted from the complementation experiment with graSR knockout strains (Table (Table2).2). Since the function of graR* in altering antibiotic MICs was observed in the absence of the signal transducer GraS, it is probable that at least a part of the GraR* proteins was already active without activation through phosphorylated GraS proteins. The activation may occur either via an increased chance of autophosphorylation of GraR* or by a structural mimicry of the conformational change of GraR brought about by GraS-mediated phosphorylation. Another possibility is that GraR* may be activated by some of the other 15 two-component signal transducers known to be present in the S. aureus chromosome (20). Detailed studies on the graSR activation mechanism are under way to solve the question.
Reduced daptomycin susceptibility is highly correlated with the vancomycin resistance of VISA clinical strains (10). The expression of graR* in Mu3 increased not only vancomycin resistance but also daptomycin resistance. This provided solid evidence that at least a part of the mechanism of vancomycin resistance is shared by the mechanism of daptomycin resistance.
It was also noticed that graR* affected oxacillin resistance negatively when introduced in N315 (Table (Table2).2). This phenomenon reminds us of the “see-saw phenomenon,” which occurs in certain stages of vancomycin resistance promotion (26). It is possible that graSR activation is the underlying genetic event for the phenomenon. N315 is called pre-MRSA, since its mecA gene expression is strongly repressed by the presence of the intact mecA repressor gene mecI (21). The mecI gene was among the down-regulated genes in our microarray data with the N315ΔgraSR/N315 combination (the ratio of transcription was 0.26). The phenotypic expression of the raised oxacillin resistance in N315ΔgraSR was not of an Eagle type (defined by resistance to high but not to low concentrations of oxacillin, a phenotypic expression caused by the presence of an intact mecI gene) (21). Therefore, the raised resistance in N315ΔgraSR was consistent with the lifting of mecI gene-mediated repression of the mecA gene. However, the contribution of graSR on mecI gene regulation seemed complex. For example, the overexpression of graR* in N315 did not increase the mecI gene transcript level (data not shown). Therefore, the effect of graSR on mecI gene transcription does not seem to occur in a simple colinear fashion. Further study is required to elucidate the relationship between mecA gene expression and graR* expression.
Although many questions arose and remain to be answered, we provided the first glimpse of the genetic alterations underlying the hetero-VISA-to-VISA promotion of vancomycin resistance naturally occurring in hospitals. The mechanism may not apply to all VISA strains. However, the physiological outcome of the genetic alteration underlying the VISA phenotype is expected to be similar, i.e., increased peptidoglycan synthesis and accumulation of cell wall materials on the cell surface. To achieve this conversion at high frequency within the cell population exposed to vancomycin, S. aureus cells seem to utilize “regulator gene mutations” (17). Regulator gene function can be affected by mutation in many different ways. Some mutations may cause complete inactivation of the regulator function or may partially inactivate it to various degrees depending on the position and nature of mutation (substitution to different amino acids) in the coding region. In this study, we found a curious example of regulator mutation that seems to activate or alter its regulatory function. Compared to VRSA, where a simple intake of the genetic system (vanA, vanH, vanY, and vanX) completes the drastic alteration of cell wall peptidoglycan composition, the acquisition of vancomycin resistance by VISA is based on spontaneous mutation. Dozens of mutations would be needed to achieve a VISA phenotype by activating transcription of each of the many enzymes and transporters involved in the cell wall synthesis pathway. This may be the reason why regulator gene mutations have a significant biological meaning. Microarray experiments revealed that a single mutation in vraS or graR alters the expression of more than 100 genes, thus causing an extreme perturbation of cell homeostasis towards the generation of cells with novel equilibrium that fits the new adverse environment. The Mu3-Mu50 lineage of strains used mutations in two sets of two-component regulatory systems, vraSR and graSR, to generate a VISA genotype to survive vancomycin chemotherapy.
We thank Taeok Bae for providing us the plasmid pKOR1.
This work was supported by a grant-in-aid for 21st century COE research and a grant-in-aid for scientific research (18590438) from The Ministry of Education, Science, Sports, Culture and Technology of Japan.
Published ahead of print on 22 October 2007.