MgrA Activates Staphylococcal Capsule via SigA-Dependent Promoter

The virulence of Staphylococcus aureus depends on the expression of various virulence factors, which is governed by a complex regulatory network. We have been using capsule as a model virulence factor to study virulence gene regulation in S. aureus. MgrA is one of the regulators of capsule and has a major effect on capsule production. However, how MgrA regulates capsule genes is not understood. In this study, we were able to define the mechanism involving MgrA regulation of capsule. In addition, we also delineated the role of MgrA in capsule regulatory pathways involving the key virulence regulators Agr and Arl. This study further advances our understanding of virulence gene regulation in S. aureus, an important human pathogen.

ABSTRACT

Staphylococcus aureus capsule polysaccharide is an important antiphagocytic virulence factor. The cap genes are regulated at the promoter element (Pcap) upstream of the cap operon. Pcap, which consists of a dominant SigB-dependent promoter and a weaker upstream SigA-dependent promoter, is activated by global regulator MgrA. How MgrA activates capsule is unclear. Here, we showed that MgrA directly bound to the Pcap region and affected the SigA-dependent promoter. Interestingly, an electrophoretic mobility shift assay showed that MgrA bound to a large region of Pcap, mainly downstream of the SigA-dependent promoter. We further showed that the ArlRS two-component system and the Agr quorum sensing system activated capsule primarily through MgrA in the early growth phases.

IMPORTANCE The virulence of Staphylococcus aureus depends on the expression of various virulence factors, which is governed by a complex regulatory network. We have been using capsule as a model virulence factor to study virulence gene regulation in S. aureus. MgrA is one of the regulators of capsule and has a major effect on capsule production. However, how MgrA regulates capsule genes is not understood. In this study, we were able to define the mechanism involving MgrA regulation of capsule. In addition, we also delineated the role of MgrA in capsule regulatory pathways involving the key virulence regulators Agr and Arl. This study further advances our understanding of virulence gene regulation in S. aureus, an important human pathogen.

INTRODUCTION

As a successful human pathogen, Staphylococcus aureus is capable of producing a large number of virulence factors needed for the pathogen to infect various human tissues (1, 2). Remarkably, the organism also possesses a plethora of transcriptional regulators (3, 4). How virulence genes are regulated is only partially understood. Capsule is one of the virulence factors that contribute to the virulence of S. aureus (1). Capsules are polysaccharides that protect S. aureus from the host immune system. However, they also mask the bacterial cell surface, thereby impeding surface molecules from interacting with host tissues or abiotic materials (5–9). Capsules are therefore regulated in response to environmental conditions. In fact, cap gene expression appears to be controlled by a complex network of regulators (10). Most strains of S. aureus produce either serotype 5 or serotype 8 capsule (11, 12). The cap loci for the production of both serotypes are allelic, each composed of 16 operonic genes with common genes flanking 4 type-specific genes (13). As the primary promoter (Pcap) region is located upstream of the first gene in the common region, the two capsule serotypes are regulated similarly.

We have previously mapped the promoter and identified an inverted repeat that is required for the transcription of cap genes (14). However, Keinhörster et al. (15) recently redefined the promoter region and concluded that cap gene expression was driven by two different promoters, a stronger SigB-dependent promoter and a weaker upstream SigA-dependent promoter. The −35 region of the SigB-dependent promoter is located within the cis-acting inverted repeat that we initially identified (14). In addition, Keinhörster et al. proposed that SigB could also regulate cap genes via several DNA-binding regulators, including MgrA (15). However, it is not clear how activation of cap genes by MgrA is affected by SigB. MgrA is a small transcriptional regulator with a DNA-binding motif that affects not only cap genes but also a large number of other genes (16, 17). Whether MgrA activates cap genes by direct binding is still unknown. To delineate how MgrA regulates capsule, we employed genetic and molecular analyses. Here, we found that MgrA activated capsule by direct binding to Pcap to affect the SigA-dependent promoter and played a major role in mediating Agr and Arl regulation of capsule in early growth phases.

RESULTS

MgrA activates capsule independently from SigB.

To test whether MgrA functions downstream of SigB with regard to cap gene regulation, we employed real-time reverse transcription (RT)-quantitative PCR (qPCR) to assay the effect of MgrA and SigB on capA (the first gene of the cap operon) gene expression in strain CYL11481, an SaeS-restored Newman strain (18). We found that MgrA and SigB have effects of ∼20-fold and ∼60-fold, respectively, while the double mutant reduced the capA gene expression level to almost undetectable (Fig. 1), suggesting that the two regulators affect the cap genes independently. To test this further, we performed complementation tests by capsule assay. As shown in Fig. 2, sigB mutation resulted in an almost nondetectable amount of capsule, whereas mgrA mutation resulted in an ∼10-fold reduction in capsule production. The results were consistent with the qPCR assays, further showing that SigB has a stronger effect on capsule regulation. In addition, complementation experiments showed that mgrA and sigB could restore their cognate mutants, but not each other’s mutant, to the wild-type level (Fig. 2), confirming the notion that MgrA and SigB independently exert regulation on Pcap. To see if the independent regulation occurs in a different strain, we used community-acquired methicillin-resistant S. aureus strain JE2. As in most strains of USA300 clonal lineage, JE2 possesses a mutation in the −35 region of sigB promoter in Pcap and two other mutations within the cap operon that render it phenotypically capsule negative (19). As a result of the mutation in the SigB-dependent promoter, expression of cap genes in JE2 is drastically reduced (15). By measuring capA gene expression with qPCR, we found that mgrA mutation reduced the capA expression by ∼10-fold, whereas sigB mutation reduced expression only 50% (Fig. 3), suggesting that MgrA regulation of cap genes is largely independent of SigB in strain JE2. These results also confirmed that SigB in JE2 had only a small effect on cap expression most likely due to the mutation in the sigB-dependent promoter. Our results are consistent with a recent gene profiling of MgrA that showed that cap genes were affected by MgrA in S. aureus strain JE2 (17).

FIG 1.

Effect of MgrA and SigB on capA expression by quantitative PCR (qPCR). Expression levels are expressed relative to that of CYL11481. Statistical significance relative to the wild type was analyzed by Student’s t test (n = 3). ****, P < 0.0001.

FIG 2.

Complementation tests of sigB::bursa and mgrA::ermC by capsule assay. Bacterial cultures were grown in tryptic soy broth (TSB) without glucose in the presence of 10 μg/ml chloramphenicol for 2 h and then induced with 25 ng/ml (A) or 5 ng/ml (B) anhydrotetracycline for 2 h. Cultures were harvested and normalized with optical density at 600 nm (OD₆₀₀). Capsules were isolated, serially diluted 3-fold, spotted on nitrocellulose paper, and blotted with an anti-type 5 antibody.

FIG 3.

Effect of MgrA and SigB on capA expression in JE2 determined by qPCR. Expression levels are relative to that of JE2. Statistical significance relative to the wild type was analyzed by Student's t test (n = 2). *, P < 0.05; **, P < 0.01.

MgrA binds to the cap promoter region.

The MgrA protein possesses a DNA-binding motif and has been shown to bind to several promoters (17, 20–22). It is likely that MgrA also regulates cap genes by directly binding to Pcap. To map the potential MgrA DNA binding site in Pcap, we made serial 5′ deletions of the cap promoter region fused to the gene encoding superfolder green fluorescent protein (sfgfp) in a plasmid vector and compared the green fluorescent protein (GFP) activities in the wild type and the mgrA mutant at 3 h and 16 h of incubation times. As shown in Fig. 4, in the wild-type strain CYL11481 at 3 h, progressive deletions of the full-length fragment (449 bp upstream of the start codon of capA gene) had no or little effect on the promoter activities (within ∼20% of the activity of the full-length fragment) except at position −59, in which the −35 region of the SigB promoter was deleted. In the 16-h CYL11481 samples, similar results were obtained with slightly more effects in deletions at positions −103 and −78 upstream of ATG. In the mgrA mutant at 3 h and 16 h, progressive deletions up to position −153 were about the same as the full-length fragment. As MgrA has a strong effect of Pcap activity, these results suggest that there is no cis-acting site for MgrA binding upstream of position −153. However, because SigB is a much stronger activator than MgrA, SigB might interfere with the MgrA effect. To exclude this concern, we also carried out the experiments in the sigB mutant background. Similarly, we did not observe significant drop in GFP in deletions up to position −153. Interestingly, in the mgrA mutant, deletions with 5′ end points at positions −103 and −78 had much higher GFP activity than the full-length fragment or the upstream deletions in both 3-h and 16-h samples. The increases were also noticeable in the wild type at the 16-h time point, but the difference was much smaller compared to the full-length fragment. These results suggest that at least one cis-acting site for a repressor is located in this region and the repression could be more clearly detected under the mgrA mutant background. In the sigB mutant, the increases were not observed, suggesting that the repression is SigB dependent.

FIG 4.

Deletion analysis of upstream of Pcap-sfgfp. The 5′ end points of deletion relative to the ATG codon of capA (+1) are shown in the map in the top panel. Relative fluorescence values were reported by normalizing with OD₆₀₀. Statistical significance relative to the full-length fragment (−449) was analyzed by Student’s t test (n ≥ 2). *, P < 0.05; **, P < 0.01.

The results from the above deletion analyses did not clearly define the potential MgrA binding site on Pcap. To further map the MgrA binding site in the Pcap region, we performed an electrophoretic mobility shift assay (EMSA). As shown in Fig. 5, we found that a 242-bp DNA fragment probe containing Pcap could be retarded by MgrA. The shifted band could be competed away by the presence of nonlabelled fragment A at 500-fold excess, suggesting the binding is specific. Various deleted DNA fragments in this region were then used to compete for the MgrA binding to the 242-bp DNA probe. We found the fragment with a 5′ deletion to position −207 upstream of the capA ATG (fragment C in Fig. 5) could still effectively compete away the shifted band, whereas further deletion to position −170 was less competitive, and deletion to position −103 totally eliminated the competition. On the other hand, deletions of the 207-bp fragment (fragment C) from the 3′ end up to the SigB-dependent promoter (Pcap/SigB) at position −62 (fragment F) abolished the ability to compete. These results suggest that MgrA binding site spans well over 100 bp, which, surprisingly, includes the region between capA ATG and the Pcap/SigB that is located mainly downstream of the SigA-dependent promoter (Pcap/SigA).

FIG 5.

Mapping of MgrA binding to Pcap by electrophoretic mobility shift assay (EMSA). EMSA was performed with a 6-carboxyfluorescein (FAM)-labeled probe (arrow) and increasing amounts of His-MgrA, indicated below the gel. DNA fragments used as probe or cold competitors are shown in the lower panel. Cold competitors were generated by PCR using primers shown to the left of each fragment and added at 500-fold excess. The end points and the sizes of the DNA fragments are also shown. The EMSA gel was directly imaged with a fluorescence imager using a 530/28-nm filter.

MgrA mediates regulation of capsule by Agr and Arl.

We have previously reported that RNAIII of the Agr regulatory system and Arl two-component system affected capsule production through MgrA (23, 24). To determine whether MgrA is the main regulator mediating the regulation of cap gene expression by Agr or Arl, we compared the GFP activities of the Pcap-gfp fusions with various deletions described above. We found that the effects of MgrA, RNAIII, and ArlR had very similar patterns at both 3 h and 16 h (Fig. 6). Furthermore, the degrees of the mutational effect were also very similar for the 3-h samples. In the 16-h samples, RNAIII mutation resulted in much reduced GFP activities, whereas arlR mutation resulted in slightly lower GFP activity compared to that with the mgrA mutation. These results suggest that Pcap is activated by RNAIII and ArlR primarily through MgrA at the 3-h time point, whereas additional regulatory pathways are also involved at the 16-h time point, particularly for RNAIII.

FIG 6.

Promoter activities of Pcap deletions fused with sfgfp were assayed in mgrA, RNAIII, or arlR mutants. The 5′ end points of deletion are the same as those shown in Fig. 4. Relative fluorescence values were reported by normalizing with OD₆₀₀. Statistical significance relative to the full-length fragment (−449) was analyzed by Student’s t test (n ≥ 2). *, P < 0.05; **, P < 0.01, ***, P < 0.001.

MgrA regulates capsule through SigA-dependent promoter.

Since MgrA did not affect capsule through the SigB-dependent promoter, we tested whether MgrA affected the upstream Pcap/SigA promoter. To this end, we constructed two plasmids with sfgfp reporter fusion to Pcap/SigA, pMLE201 with Pcap between positions −344 and −103 and pMLE200 with Pcap between −344 and −78 upstream from cap5A. The reporter plasmids were introduced to CYL11481 and the isogenic mgrA and sigB mutants. GFP assays were performed from 3-h and 16-h cultures. For the pMLE201 reporter, the GFP activities were reduced by ∼50% in the mgrA mutant compared to the wild type (after subtracting the baseline level of vector control) at both time points (Fig. 7A). For the pMLE200 reporter, the expression level was much lower than that of pMLE201, indicating that sequence between −103 and −78 negatively impacts the promoter activity (Fig. 7B). Nonetheless, the GFP activities in the mgrA mutant with pMLE200 were also lower than those in the wild type, at ∼25% and ∼42% of 3-h and 16-h cultures, respectively. There was no significant difference with 3-h cultures between the mgrA mutant and the wild type, which was most likely due to very low GFP expression from pMLE200 at this time point. In the sigB mutant, the GFP activity from either reporter plasmid was virtually the same as those of the wild type. These results suggest that MgrA affects the SigA-dependent promoter of the Pcap. However, Pcap/SigA is not totally controlled by MgrA, as there was only a partial reduction in GFP activity in the mgrA mutant compared to that in the wild type.

FIG 7.

Promoter activities of Pcap/SigA-sfgfp reporter plasmids. (A) pMLE201 with Pcap/SigA insert from −344 to −103. (B) pMLE200 with Pcap/SigA insert from −344 to −78. The activities were assayed in wild-type CYL11481 and isogenic mgrA and sigB mutants. GFP assay was carried out with 3-h and 16-h cultures. Relative fluorescence values were reported by normalizing with OD₆₀₀ and then subtracting the values of the reporter plasmid controls. Statistical significance relative to the wild type was analyzed by Student’s t test (n = 2). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

DISCUSSION

Staphylococcal capsule is regulated by more than a dozen of regulators forming a complex regulatory network. Among these regulators, SigB, XdrA, CodY, RbsR, Rot, SaeR, SpoVG, AirR, and KdpE have been shown to directly bind to Pcap (15, 25–31). In addition, MsaB/CspA and CcpE have also been implicated for binding, but the findings were inconsistent with those of other studies (15, 32–35). Most of these binding studies were performed by EMSA, with only a few by DNase I footprinting. In this report, we demonstrated that MgrA was another Pcap binding regulator. To our knowledge, the number of regulators capable of binding to Pcap is the highest among staphylococcal promoters studied to date. In addition, a handful of other regulators, some of which are upstream regulators, have also been reported (11, 12). The large number of regulators involved in capsule regulation is surprising, reflecting that capsule is subject to a high degree of regulation.

Consistent with a previous report of SigB being the dominant regulator (15), we showed that sigB mutation resulted in about 60-fold reduction of capA transcription by qPCR (Fig. 1). It is therefore not surprising that CodY, SaeR, and Rot were shown to affect Pcap through SigB and that MgrA, SpoVG, and RbsR were thought to be SigB-dependent activators as well (15). However, in this report, we found that MgrA affected Pcap independently from SigB. We also found that MgrA had a strong effect on Pcap, as mgrA deletion resulted in reduction of capA transcription by about 20-fold. MgrA has been shown to affect a large number of genes besides cap genes (16, 17). How MgrA activates Pcap has not been delineated. Here, we showed that MgrA activated cap genes by direct binding to the Pcap region and affected the Pcap/SigA promoter upstream of the Pcap/SigB promoter. Interestingly, our results showed that the MgrA binding site spanned a large region located mostly downstream of Pcap/SigA. In prokaryotes, DNA-binding activators typically affect regulation by binding to upstream of the promoter (36). The cis-acting site of an activator located downstream of the promoter is unusual, with only a few examples have been reported (37–40). Two mechanisms have been proposed. One mechanism is by antagonizing repression, and the other is by altering local DNA structure thereby promoting RNA polymerase binding to the promoter (41). Previously, we reported that CodY and XdrA repressed capsule by binding to the Pcap that overlaps with the MgrA binding region (26). However, CodY has been shown to affect the SigB-dependent promoter (15). Whether MgrA could activate the SigA-dependent promoter by antagonizing XdrA awaits further studies. It should be noted here that two different MgrA DNA binding consensus sequences have been proposed (42, 21). In a recent study, it was found that a DNA fragment containing a mutation in one of the proposed consensus sequences in the ebh promoter region was unable to compete with MgrA binding (44), suggesting that this consensus sequence, (A/T)GTTGT, is the specific binding site for MgrA. However, we found no such consensus sequence in the MgrA binding region of Pcap.

Previously, we showed that the ArlRS two-component system activated capsule through MgrA (24). Later, Crosby et al. (44) demonstrated that phosphorylated ArlR bound to the mgrA distal P2 promoter to activate mgrA expression. We also reported that RNAIII of the Agr quorum sensing system activated MgrA by stabilizing mgrA mRNA transcribed from the P2 promoter (23). In this study, by employing serially deleted Pcap fragments fused to sfgfp, we showed that RNAIII and ArlR regulated capsule primarily through MgrA to affect Pcap/SigA at the 3-h time point. The growth phase-dependent regulation is consistent with our previous observation that MgrA strongly affected capsule in early growth phases (16). At 16 h, besides the MgrA-mediated pathway, we found that additional pathways were also involved in capsule regulation by RNAIII and Arl. RNAIII has been shown to repress Rot mainly in late exponential growth phases (45, 46). As Rot has been shown to affect Pcap/SigB rather than Pcap/SigA (15), it is most likely that Rot is the regulator that mediates capsule regulation by Agr/RNAIII through Pcap/SigB in late growth phases.

The results in Fig. 7 show that the −103 to −78 region of Pcap had a negative effect on Pcap/SigA. Interestingly, this region could also serve as a repressor binding site affecting Pcap/SigB (Fig. 4). Although the precise boundary has yet to be determined, our results suggest that this short stretch of DNA sequence could serve as the binding site for a repressor or repressors affecting both upstream and downstream promoters. It has been reported that SaeR represses both Pcap/sigA and Pcap/sigB promoters (15). Furthermore, a possible SaeR binding site (GTTTAN₆ATTAA) with an 80% identity to the consensus (GTTAAN₆GTTAA) located just upstream of Pcap/sigB between −95 to −79 has been proposed (47). However, whether SaeR binds to this region to repress both promoters awaits further study.

Although we showed that MgrA affected capsule through Pcap/SigA rather than the Pcap/SigB, our results also suggest that MgrA is not the only positive regulator affecting the SigA-dependent promoter (Fig. 7). As Pcap is regulated by a large number of regulators, it is not surprising that additional positive regulators could also target this distal SigA-dependent promoter independently from MgrA. Recently, we have identified several potential Pcap-binding regulators (25). Whether these regulators affect the SigA- and/or SigB-dependent promoters remained to be investigated.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. JE2, NE1109, and NE1684 were obtained from the Nebraska Transposon Mutant Library collection (52) distributed by BEI Resources (beiresources.org) through the Network of Antimicrobial Resistance in S. aureus (NARSA) program. Culture conditions were essentially as previously described (25).

TABLE 1.

Strains and plasmids

Strain or plasmid	Relevant characteristic(s)	Reference or source
S. aureus
RN4220	Restriction-negative laboratory strain	J. Iandolo
CYL11481	Newman saeS(P18L)	18
JE2	USA300 strain	NARSA
NE1109	JE2 sigB::bursa	NARSA
NE1684	arlR::bursa	NARSA
CYLA273	CYL11481 ΔmgrA::cat	This study
CYLA543	CYL11481 ΔmgrA::ermC	This study
CYLA695	CYL11481 sigB::bursa	This study
CYLA696	CYL11481 ΔmgrA::cat sigB:: bursa	This study
CYLA270	JE2 ΔmgrA::cat	48
CYLA687	JE2 sigB::bursa	This study
CYLA226	Newman RNAIII::ermC	This study
CYLA230	CYL11481 RNAIII::ermC	This study
CYLA546	CYL11481 arlR::bursa	This study
AL2530	RN6390 ΔmgrA::ermC	22
E. coli
XL1-Blue	Host strain	Strategene
CYL2961	BL21(λDE3)(pLysS)(pTL2961)	This study
Plasmids
pML100	Shuttle vector, with Pxyl/tetO	49
pLI50	Shuttle vector	50
pET15b	E. coli His tag protein expression vector	Novagen
pMLE195	pML100 with sigB	This study
pMLE197	pML100 with mgrA	This study
pMLE57	pLI50 with sfgfp	This study
pTL2961	pET15b with mgrA	This study
pCM11	sfgfp expression vector	51
pCL3292	pCL15 with mgrA	24
pMLE200	pMLE57-Pcap(−344 to −78)	This study
pMLE201	pMLE57-Pcap(−344 to −103)	This study

Strain and plasmid construction.

Primers used for strain and plasmid construction are listed in Table 2. Transduction between S. aureus strains were carried out by phage 52A as previously described (25). The transposon mutants, CYLA695 (sigB::bursa) and CYLA546 (arlR::bursa) were constructed by chromosomal transduction from NE1109 and NE1684 to CYL11481, respectively. To construct the sigB-mgrA double mutant, CYLA273 was transduced with sigB::bursa from NE1109. To construct Newman ΔRNAIII::ermC (CYLA226), a DNA fragment with an RNAIII internal deletion was first constructed by overlapping PCR using the primer pairs rna3dF1/rna3dR1 and rna3dF2/rna3dR2 and then cloned into pJB38 (53). The ermC gene of pE194 amplified by erm11 and erm12 primers was inserted into the cloned insert by restriction digestion and ligation such that the RNAIII region to be deleted was replaced with the ermC gene. The PCR-amplified fragments were verified by sequencing. Allelic replacement was performed as described previously (54). The ΔRNAIII::ermC allele was then moved to CYL11481 by phage transduction to yield CYL230.

TABLE 2.

Oligonucleotide primers used in this study

Primer	Sequence
sigBsal	GATGATATCGTCGACGAAAATAGGAGGAAATAATGGCGAAAGAGTCG
sigBe1	GATGATATCGAATTCTTAAAAATACCCCTCGATTTC
rna3df1	GATGATATCGTCGACCGCATAATATTTTTTTCGTTTAATAAGTCG
rna3dr1	CTCATCCCTTCTTCATTACGCGGCCGCCTAGTTATATTAAAACATGCTAAAAGC
rna3df2	CATGTTTTAATATAACTAGGCGGCCGCGTAATGAAGAAGGGATGAGTTAATC
rna3dr2	GATGATATCGAATTCGCTGAACCCTTTCAATTGTCTGAC
pcp449	GATGATATCAAGCTTGCAGAGCTCGCATTTGAAGATCA
pcp344	GATGATATCAAGCTTTCTATCTGATAATAATCATCTAACT
pcp242	GATGATATCAAGCTTATTGCCAAATCTTCAGAGAGAT
pcp207	GATGATATCAAGCTTCATTTTTTTAAATAAAGAAAATATTAAGA
pcp170	GATGATATCAAGCTTCATTTCACAAACTATTACTACTTTAGA
pcp153	GATGATATCAAGCTTCTATTACTACTTTAGAGTATAATTA
pcp103	GATGATATCAAGCTTAATAATGCGGTTTAAAAGTAATTAATTG
pcp78	GATGATATCAAGCTTGTTTAAACGATATGTAATATGTAAATAC
pcp59	GATGATATCAAGCTtATGTAAATACTATATATATAATACCAATTTTAATG
cp8bla.rk	GATGATATCGGTACCCTTAGTTTGATTCACTAAAATTTG
FAM-fp7	FAM-5′-ATTGCCAAATCTTCAGAGAGAT
FAM-fp3	FAM-5′-CCATTATTTACCTCCCTTAAAAATTTTC
Erm11	CCCGGGAATTGAATGAGACATGCTAC
Erm12	AAGCTTAAAACTGGTTTAAGCCGAC
SGcap8A1	ACTAAGGGTGACAATCCTCAG
SGcap8A2	AAGTCCTTTGACACCTCATCTA
SGhu2	ATCCAAAACTCACTTGCTAAAGG
SGhu3	ACCAGCTTTGAATGCTGGAAC

For mgrA complementation, pMLE197 containing the Newman mgrA gene under the control of Pxyl/tetO was constructed by ligating a HindIII-digested 565-bp fragment in pCL3292 (24) into HindIII-digested pML100, then screened for correct orientation. For sigB complementation, pMLE195 containing the sigB gene was constructed by ligating a SalI/EcoRI-digested 864-bp PCR fragment amplified with the primer pair sigBsal/sigBe1 into similarly digested pML100. To express the recombinant His₆-MgrA protein in Escherichia coli, a 460-bp fragment containing the mgrA gene from Newman was amplified using primers mgr38 and mgr39 and ligated to NheI/BamHI-digested pET-15b (Novagen, Madison, WI). All clones were validated by restriction mapping and sequencing of the inserts. Superfolder green fluorescent protein gene (sfgfp) reporter plasmid pMLE57 was constructed by inserting the sfgfp fragment from pCM11 into the KpnI and EcoRI sites of pLI50. Full-length Pcap (449 bp upstream of the capA start codon) and its various deletions with 344 bp, 242 bp, 207 bp, 170 bp, 153 bp, 103 bp, 78 bp, and 59 bp (numbers refer to the distance from capA start codon) were generated by paring 5′ primers pcp449, pcp344, pcp242, pcp207, pcp170, pcp153, pcp103, pcp78, and pcp59, respectively, with 3′ primer cp8bla.rK and cloned to the HindIII and KpnI sites of pMLE57. All inserts were verified by sequencing.

Recombinant protein expression and purification.

To express His₆-MgrA proteins, pTL2961 was transformed into E. coli BL21(λDE3)(pLysS) (Novagen). Procedures for protein expression and purification were essentially as described previously (25). The purity of the His₆-MgrA protein was estimated by SDS-PAGE to be greater than 95%(not shown).

Electrophoretic mobility shift assay.

Electrophoretic mobility shift assay (EMSA) was performed as described previously (26). The 242-bp DNA probe (shown in Fig. 5) was generated from strain Newman chromosome DNA by PCR using primers FAM-fp3 and FAM-fp7. The EMSA gel was directly imaged with a fluorescence imager using a 530/28-nm filter.

Fluorescence reporter assay.

S. aureus strains harboring sfgfp reporter fused to various Pcap deletions in pMLE57 were grown in tryptic soy broth (TSB) medium without glucose (TSB-0G) with 10 μg/ml of chloramphenicol. Cultures were centrifuged and suspended in phosphate-buffered saline (PBS), and green florescence was measured in quadruplet in black 96-well microtiter plates essentially as described previously (26). Relative fluorescence values of promoter-reporter fusions were calculated by normalizing the average fluorescence of each sample with the corresponding absorbance of optical density at 600 nm (OD₆₀₀).

Capsule immunoblotting.

To measure capsule production, capsules were prepared as described previously (18) from cultures grown in TSB-0G. Serially diluted samples (1.5 μl each) were applied directly to nitrocellulose membrane by using a pipette. Membranes were treated with specific anticapsule antibody and detected as described previously (18).

RNA isolation and reverse transcription-quantitative PCR.

RNA samples were isolated as previously descried (49). qPCR was performed using primers SGcapA1 and SGcap8A2 to detect capA expression and normalized with the hu gene using the primers SGhu2 and SGhu3 as described previously (49).

Statistics.

Statistical analyses for comparison between means were analyzed by GraphPad Prism (San Diego, CA) using paired Student t tests.