Phosphorylation at the D53 but Not the T65 Residue of CovR Determines the Repression of rgg and speB Transcription in emm1- and emm49-Type Group A Streptococci

CovR/CovS is a two-component regulatory system in group A Streptococcus (GAS). The D53 residue of CovR is phosphorylated by CovS, and the phosphorylated CovR_D53 binds to the rgg-speB intergenic region and acts as the transcriptional repressor. Nonetheless, the transcription of rgg and Rgg-controlled speB is upregulated in the covR mutant but inhibited in the covS mutant. The present study showed that nonphosphorylated CovR_D53 protein inhibits rgg and speB transcription in the presence of the phosphorylated CovR_D53 in vivo, indicating that nonphosphorylated CovR_D53 has a dominant role in suppressing rgg transcription. These results reveal the roles of nonphosphorylated CovR_D53 in regulating rgg transcription, which could contribute significantly to invasive phenotypes of covS mutants.

ABSTRACT

CovR/CovS is a two-component regulatory system in group A Streptococcus and primarily acts as a transcriptional repressor. The D53 residue of CovR (CovR_D53) is phosphorylated by the sensor kinase CovS, and the phosphorylated CovR_D53 protein binds to the intergenic region of rgg-speB to inhibit speB transcription. Nonetheless, the transcription of rgg and speB is suppressed in covS mutants. The T65 residue of CovR is phosphorylated in a CovS-independent manner, and phosphorylation at the D53 and T65 residues of CovR is mutually exclusive. Therefore, how phosphorylation at the D53 and T65 residues of CovR contributes to the regulation of rgg and speB expression was elucidated. The transcription of rgg and speB was suppressed in the strain that cannot phosphorylate the D53 residue of CovR (CovR_D53A mutant) but restored to levels similar to those of the wild-type strain in the CovR_T65A mutant. Nonetheless, inactivation of the T65 residue phosphorylation in the CovR_D53A mutant cannot derepress the rgg and speB transcription, indicating that phosphorylation at the T65 residue of CovR is not required for repressing rgg and speB transcription. Furthermore, trans complementation of the CovR_D53A protein in the strain that expresses the phosphorylated CovR_D53 resulted in the repression of rgg and speB transcription. Unlike the direct binding of the phosphorylated CovR_D53 protein and its inhibition of speB transcription demonstrated previously, the present study showed that inactivation of phosphorylation at the D53 residue of CovR contributes dominantly in suppressing rgg and speB transcription.

IMPORTANCE CovR/CovS is a two-component regulatory system in group A Streptococcus (GAS). The D53 residue of CovR is phosphorylated by CovS, and the phosphorylated CovR_D53 binds to the rgg-speB intergenic region and acts as the transcriptional repressor. Nonetheless, the transcription of rgg and Rgg-controlled speB is upregulated in the covR mutant but inhibited in the covS mutant. The present study showed that nonphosphorylated CovR_D53 protein inhibits rgg and speB transcription in the presence of the phosphorylated CovR_D53 in vivo, indicating that nonphosphorylated CovR_D53 has a dominant role in suppressing rgg transcription. These results reveal the roles of nonphosphorylated CovR_D53 in regulating rgg transcription, which could contribute significantly to invasive phenotypes of covS mutants.

INTRODUCTION

Streptococcus pyogenes (group A Streptococcus [GAS]) is a Gram-positive human pathogen causing diseases from mild pharyngitis and tonsillitis to severe cellulitis, necrotizing fasciitis, and toxic shock syndrome. An increase in invasive GAS infections has been reported globally (1–6) and is associated with high mortality and economic burden (7, 8). Therefore, there is an urgent need to understand the pathogenesis of invasive GAS infections.

Mutations in covS and covR genes are detected more frequently in strains isolated from patients with severe manifestations than from patients with mild symptoms (9–12). CovR/CovS is a two-component regulatory system of GAS (13). CovS acts as a membrane-associated sensor kinase/phosphatase to modulate the phosphorylation levels of the intracellular response regulator CovR (14, 15). There are two residues of CovR that can be phosphorylated. The phosphorylation level of the D53 residue of CovR (CovR_D53) is modulated by CovS; however, the T65 residue of CovR (CovR_T65) is phosphorylated by a CovS-independent mechanism (16). In addition, in vitro analysis showed that phosphorylation of these two residues in the recombinant CovR protein is mutually exclusive (16). Inactivation of CovS results in decreased levels of phosphorylated CovR_D53 and the derepression of CovR-regulated virulence factors, including the hyaluronic acid capsule synthesis (has operon), streptolysin S (sls), streptolysin O (slo), NADase (nga), and DNase (sda1) (14, 17–19). In addition, although CovS is considered to regulate target genes solely through CovR (20), a strain that expresses nonphosphorylated CovR_D53 is more virulent than the covR deletion strain in a mouse infection model (16), suggesting that nonphosphorylated CovR_D53 in the covS mutant still has important roles in GAS pathogenesis. Nonetheless, how CovR contributes to GAS pathogenesis in the absence of CovS is still not clear.

Rgg is a positive regulator of the cysteine protease SpeB (21, 22). The emm3-type isolates with mutations in rgg express a higher level of virulence factors, such as slo, nga, and ska, than do isolates with intact rgg (11). In the emm49-type NZ131 strain, Rgg acts as a transcriptional repressor of genes that are involved in virulence, metabolism, and oxidative stress responses (23–25). Both rgg mutants of emm3- and emm49-type strains showed higher lethality in the mouse infection models than did wild-type strains (11, 23), indicating that inactivation of Rgg contributes to the increase of bacterial virulence.

The phosphorylated CovR_D53 protein has better DNA-binding activity than the nonphosphorylated CovR_D53 protein (26–28). The phosphorylated CovR_D53 protein binds to the rgg-speB intergenic region to inhibit speB transcription (28–30). Nonetheless, the transcription of rgg is repressed in the covS mutant and therefore resulted in the inactivation of speB transcription (19, 30). These results indicate that the covS mutant and covR mutant are not identical. In addition, unlike the direct binding of the phosphorylated CovR_D53 protein and its inhibition of speB transcription, which have been extensively demonstrated previously, whether nonphosphorylated CovR_D53 is capable of repressing rgg transcription and subsequently preventing SpeB production has not been verified.

Phosphorylation of the CovR D53 and T65 residues is mutually exclusive (16). Inactivation of the CovR protein phosphorylation by a D53A or T65A amino acid substitution does not result in aberrantly folded protein in vitro (16). Nonetheless, the CovR_T65A mutant was hypervirulent compared to the wild-type strain (16), suggesting that the phosphorylated CovR_T65 protein contributes significantly to the regulatory networks of GAS. The present study aims to analyze how phosphorylated CovR_D53 and CovR_T65 proteins mediate the repression of rgg and speB transcription. The results showed that phosphorylation at the T65 residue of CovR was not required for suppressing rgg and speB transcription. In addition, the nonphosphorylated CovR_D53 had the more dominant roles in inhibiting rgg and speB transcription than did the phosphorylated CovR_D53 protein.

RESULTS

Transcription of rgg was inhibited in the CovR_D53A mutant but derepressed in the CovR_T65A mutant.

Phosphorylation of CovR_D53 promotes its dimerization and increases its activity to bind to the rgg-speB intergenic region (28). CovR acts as a transcriptional repressor of rgg; however, the deletion of covS and inactivation of the D53 residue phosphorylation of CovR result in the repression of rgg transcription (19, 30), suggesting that the repression is unrelated to the phosphorylation at D53. In addition to the D53 residue, the T65 residue of CovR is phosphorylated by a CovS-independent mechanism (16). Horstmann et al. (16) showed that phosphorylation at the D53 and T65 residues of CovR is mutually exclusive; therefore, the phosphorylated CovR_T65 could mediate the repression of rgg in the covS mutant. In order to elucidate whether the phosphorylated CovR_T65 is involved in the repression of rgg transcription, the isoallelic GAS strains that were unable to be phosphorylated at D53 [A20_CovR(D53A) mutant] and T65 [A20_CovR(T65A) mutant] were constructed, and the levels of CovR protein expression and phosphorylation in these mutants were analyzed by Western blotting and Phos-tag Western blotting, respectively. The results showed that the expression of CovR protein in the CovR_D53A mutant was decreased compared to those in the wild-type A20 strain and CovR_T65A mutant (Fig. 1A). The phosphorylated CovR_D53 protein was detected in the wild-type A20 strain and its CovR_T65A mutant (Fig. 1A). Nonetheless, the phosphorylated CovR_T65 protein cannot be observed in the covS and CovR_D53A mutants (Fig. 1A), suggesting that low levels of phosphorylated CovR_T65 were presented in those mutants or that the phosphorylated CovR_T65 cannot be detected by Phos-tag Western blotting effectively. Next, the transcription levels of rgg in the wild-type A20 strain, A20_CovR(D53A) mutant, and A20_CovR(T65A) mutant were analyzed. Compared to the wild-type A20 strain, the transcription of rgg was inhibited in the A20_CovR(D53A) mutant (Fig. 1B). Nonetheless, the transcription of rgg in the A20_CovR(T65A) mutant was restored to a level similar to that in the wild-type A20 strain (Fig. 1B).

FIG 1.

Phosphorylation levels of CovR and the expression of CovR, rgg, and speB in the wild-type A20 strain (emm1 type) and its CovR_D53A and CovR_T65A mutants. Bacterial strains were cultured to the exponential (for detecting CovR) and stationary (for detecting rgg and speB) phases of growth. RNAs, culture supernatants, and total proteins were collected for quantitative real-time PCR (RT-qPCR), Western blotting, and Phos-tag Western blot analyses. (A) Levels of phosphorylation and expression of CovR protein in A20, CovR_D53A mutant, and CovR_T65A mutant. CovR∼P, phosphorylated CovR; CovR, nonphosphorylated CovR. Total protein serves as the loading control. (B) Transcription of rgg in A20, CovR_D53A mutant, and CovR_T65A mutant. (C) The transcription of speB and secretion of SpeB protein in A20, CovR_D53A mutant, and CovR_T65A mutant. Biological replicate experiments were performed using three independent preparations. The expression of rgg and speB was normalized to that of gyrA. Wt, wild-type strain; *, P < 0.05.

Rgg is the positive regulator of streptococcal pyrogenic exotoxin B (SpeB) (21). Therefore, whether the expression of SpeB was derepressed in the A20_CovR(T65A) mutant was further analyzed with quantitative PCR (qPCR) and Western blotting. The results showed that the transcription of speB was partially restored in the A20_CovR(T65A) mutant compared to the A20_CovR(D53A) mutant (Fig. 1C). Furthermore, both the zymogen (42 kDa) and mature (28 kDa) forms of SpeB proteins were detected in the supernatants from the wild-type A20 strain and A20_CovR(T65A) mutant (Fig. 1C). These results support the idea that the nonphosphorylated CovR_D53 acts as the repressor for rgg transcription; in addition, the phosphorylated CovR_T65 may also participate in the regulation of rgg transcription.

Inactivation of CovR_D53 phosphorylation in the emm49-type NZ131 strain inhibits rgg and rgg-controlled speB transcription.

The results shown in Fig. 1 indicate that the transcription of speB was inhibited in the A20_CovR(D53A) mutant but derepressed in the A20_CovR(T65A) mutant. Nonetheless, in the emm3-type strain, the transcription of speB was derepressed both in CovR_D53A and CovR_T65A mutants (16). The emm3-type strain is rarely found in Taiwan. Therefore, to elucidate how CovR_D53A and CovR_T65A mutations influence the transcription of rgg and speB, the CovR_D53A and CovR_T65A mutants of another reference strain, NZ131 (emm49 type), were constructed, and the transcription of rgg and speB was analyzed. The levels of expression and phosphorylation of CovR proteins in NZ131 and its covR mutants were similar to those observed in A20 and its covR mutants (Fig. 2A). Compared to the wild-type NZ131 strain, the transcription of rgg and speB was repressed in the NZ131_CovR(D53A) mutant but derepressed in the NZ131_CovR(T65A) mutant (Fig. 2B and C). These results further support the idea that the nonphosphorylated CovR_D53 or phosphorylated CovR_T65 could mediate the repression of rgg and speB transcription.

FIG 2.

Phosphorylation levels of CovR and the expression of CovR, rgg, and speB in the wild-type NZ131 strain (emm49 type) and its CovR_D53A and CovR_T65A mutants. Bacterial strains were cultured to the exponential (for detecting CovR) and stationary (for detecting rgg and speB) phases of growth. RNAs, culture supernatants, and total proteins were collected for RT-qPCR, Western blotting, and Phos-tag Western blot analyses. (A) Levels of phosphorylation and expression of CovR protein in NZ131, CovR_D53A mutant, and CovR_T65A mutant. CovR∼P, phosphorylated CovR; CovR, nonphosphorylated CovR. Total protein serves as the loading control. (B) Transcription of rgg in NZ131, CovR_D53A mutant, and CovR_T65A mutant. (C) Transcription of speB and secretion of SpeB protein in NZ131, CovR_D53A mutant, and CovR_T65A mutant. Biological replicate experiments were performed using three independent preparations. The expression of rgg and speB was normalized to that of gyrA. Wt, wild-type strain; *, P < 0.05.

Inactivation of CovR_T65 phosphorylation in the CovR_D53A mutant cannot derepress rgg and speB transcription.

In order to further verify whether phosphorylation at the T65 residue of CovR is required for suppressing rgg and speB transcription, the T65A amino acid substitution of CovR in the A20_CovR(D53A) mutant was constructed [A20_{CovR(D53A-T65A)} mutant]. As expected, results from the Phos-tag Western blot analysis showed that the phosphorylated CovR proteins cannot be detected in the A20_{CovR(D53A-T65A)} mutant (Fig. 3A). In addition, different from the A20 and its CovR_D53A mutant, the expression level of CovR in CovR_T65A and CovR_D53A-T65A mutants was similar (Fig. 1A and 3A). Although the phosphorylated CovR_T65 protein in the CovR_D53A mutant cannot be detected by Phos-tag Western blotting, these results support the idea that the phosphorylated CovR_T65 contributes to regulating CovR expression (16). In line with the results shown in Fig. 1B, the wild-type A20 strain and A20_CovR(T65A) mutant had similar levels of rgg transcription (Fig. 3B). Nonetheless, the transcription of rgg was still suppressed in the A20_{CovR(D53A-T65A)} mutant (Fig. 3B). In addition, the expression of SpeB was inhibited both in the transcriptional and translational levels in A20_{CovR(D53A-T65A)} mutant compared to the wild-type A20 strain and A20_CovR(T65A) mutant (Fig. 3C). These results indicate that phosphorylation at the T65 residue of CovR is not required for repression of rgg and speB transcription.

FIG 3.

Phosphorylation levels of CovR and the expression of CovR, rgg, and speB in emm1-type A20 and its CovR_T65A and CovR_D53A-T65A mutants. Bacterial strains were cultured to the exponential (for detecting CovR) and stationary (for detecting rgg and speB) phases of growth. RNAs, culture supernatants, and total proteins were collected for RT-qPCR, Western blotting, and Phos-tag Western blot analyses. (A) Levels of phosphorylation and expression of CovR protein in A20, CovR_D53A mutant, and CovR_D53A-T65A mutant. CovR∼P, phosphorylated CovR; CovR, nonphosphorylated CovR. Total protein serves as the loading control. (B) Transcription of rgg in A20, CovR_T65A mutant, and CovR_D53A-T65A mutant. (C) Transcription of speB and secretion of SpeB protein in A20, CovR_T65A mutant, and CovR_D53A-T65A mutant. Biological replicate experiments were performed using three independent preparations. The expression of rgg and speB was normalized to that of gyrA. Wt, wild-type strain; *, P < 0.05.

Expression of the CovR_D53A protein in CovR_T65A mutants inhibits rgg and speB transcription.

Phosphorylated CovR_D53 has better DNA-binding activity than does the nonphosphorylated CovR_D53 protein (28); therefore, whether the nonphosphorylated CovR_D53 protein could inhibit rgg and speB transcription in the presence of phosphorylated CovR_D53 was further verified. The covR gene encoding the D53A amino acid substitution was carried by the pTRKL2 vector and was electroporated into the strain that expresses the phosphorylated CovR_D53 protein [A20_CovR(T65A)] (see Fig. S1 in the supplemental material). The transcription of rgg in the A20_CovR(T65A) mutant, its vector control strain, and the CovR_D53A trans-complemented strain were further compared. The results showed that trans complementation of the CovR_D53A protein in the A20_CovR(T65A) mutant suppressed the transcription of rgg compared to the wild-type A20 and its vector control strains (Fig. 4A). Next, the expression levels of SpeB in the wild-type A20 strain, A20_CovR(T65A) mutant, its vector control strain, and the CovR_D53A trans-complemented strain were analyzed by qPCR and Western blotting. Compared to the wild-type strain, the expression of SpeB in both transcriptional and translational levels was decreased in the vector control strain (Fig. 4B). Nonetheless, the expression of SpeB cannot be detected in the CovR_D53A trans-complemented strain (Fig. 4B). Furthermore, the transcription of rgg and speB was similarly analyzed in NZ131 strains. In line with results shown in the A20 strains, the transcription of rgg and speB was inhibited in the CovR_D53A trans-complemented strain of NZ131_CovR(T65A) (Fig. 4C and D). These results suggest that the nonphosphorylated CovR_D53A protein has a more dominant role in repressing the transcription of rgg and speB than does the phosphorylated CovR_D53 protein.

FIG 4.

Expression of rgg and speB in CovR_T65A mutants and their vector control and CovR_D53A trans-complementary strains. Bacterial strains were cultured to the stationary phase of growth. RNAs and culture supernatants were collected for RT-qPCR and Western blot analyses. (A and B) Transcription of rgg and the transcriptional and translational levels of SpeB expression in the emm1-type A20 strain, CovR_T65A mutant, and its vector control (Vector) and CovR_D53A trans-complementary (pCovR_D53A) strains. (C and D) Transcription of rgg and the transcriptional and translational levels of SpeB expression in the emm49-type NZ131 strain, CovR_T65A mutant, and its vector control and CovR_D53A trans-complementary strains. Biological replicate experiments were performed using three independent preparations. The expression of rgg and speB was normalized to that of gyrA. Wt, wild-type strain; *, P < 0.05.

Expression of phosphorylated CovR_D53 protein in CovR_D53A mutants cannot restore rgg and speB transcription.

The transcription of speB is repressed in the covS mutant, suggesting that CovS acts as a positive regulator of speB (20, 31). In addition, it has been considered that CovS regulates downstream genes solely through CovR and is required for the CovR phosphorylation (20). Therefore, whether the phosphorylated CovR_D53 could activate the rgg and speB transcription in A20_CovR(D53A) was analyzed. The covR gene with the T65A amino acid substitution was carried by the pTRKL2 vector and was electroporated into CovR_D53A mutants. The D53 residue of the CovR_T65A protein can be phosphorylated by CovS, and expression of the phosphorylated CovR_T65A protein in CovR_D53A mutants was confirmed by Phos-tag Western blotting (Fig. 1; Fig. S1). Next, the transcription of rgg in the A20_CovR(D53A) mutant, its vector-control strain, and the CovR_T65A trans-complemented strain were further compared. The results showed that expression of the CovR_T65A protein in the A20_CovR(D53A) mutant had no effect on activating rgg and speB transcription (Fig. 5A and B). In addition, expression of the SpeB protein can only be detected in the wild-type A20 strain but not in A20_CovR(D53A) and CovR_T65A trans-complementary mutant strains (Fig. 5B). Furthermore, similar results were observed in NZ131 strains (Fig. 5C and D). These results indicate that the phosphorylated CovR_D53 cannot attenuate the repressor activity of the CovR_D53A protein to rgg and speB transcription.

FIG 5.

Expression of rgg and speB in CovR_D53A mutants and their vector control and CovR_T65A trans-complementary strains. Bacterial strains were cultured to the stationary phase of growth. RNAs and culture supernatants were collected for RT-qPCR and Western blot analyses. (A and B) Transcription of rgg and the transcriptional and translational levels of SpeB expression in the emm1-type A20 strain, CovR_D53A mutant, and its vector control (Vector) and CovR_T65A trans-complementary (pCovR_T65A) strains. (C and D) Transcription of rgg and the transcriptional and translational levels of SpeB expression in the emm49-type NZ131 strain, CovR_D53A mutant, and its vector control and CovR_T65A trans-complementary strains. Biological replicate experiments were performed using three independent preparations. The expression of rgg and speB was normalized to that of gyrA. Wt, wild-type strain; *, P < 0.05.

DISCUSSION

The phosphorylated CovR_D53 protein has better activity for binding to the rgg-speB intergenic region than does nonphosphorylated CovR_D53 (28). In addition, the transcription of rgg and speB is upregulated in the covR mutant, suggesting that CovR inhibits rgg and speB transcription. Nonetheless, the transcription of speB is repressed in the covS mutant. In the present study, how phosphorylated and nonphosphorylated CovR_D53 proteins contribute to regulating rgg and speB expression was elucidated. The results showed that the nonphosphorylated CovR_D53 protein has a more dominant role in suppressing rgg and speB transcription than does the phosphorylated CovR_D53 protein. Furthermore, inactivation of the phosphorylation at D53 but not the T65 residue of CovR is crucial for repressing rgg and speB transcription.

The D53 (CovR_D53) and T65 (CovR_T65) residues of CovR can be phosphorylated, and the phosphorylation at these two residues is mutually exclusive (16). CovR_D53 is phosphorylated by CovS; however, CovR_T65 is phosphorylated by the serine/threonine kinase (16). The strain unable to be phosphorylated at CovR_T65 was hypervirulent compared to the wild-type strain (16), suggesting that the phosphorylated CovR_T65 could have an important role in modulating the regulatory activity of CovR. The present study showed that the transcription of rgg and speB was repressed in strains unable to be phosphorylated at CovR_D53 (CovR_D53A mutants) but restored to levels similar to the wild-type strains in CovR_T65A mutants (Fig. 1B and C and 2B and C), suggesting that phosphorylation at CovR_T65 could contribute to the repression of rgg and speB transcription. Nonetheless, inactivation of CovR_T65 phosphorylation in the CovR_D53A mutant cannot derepress rgg and speB transcription (Fig. 3B and C). Although these results indicate that phosphorylation at the CovR_T65 is not required for inhibiting rgg and speB transcription, whether the phosphorylated CovR_T65 contributes to regulating other regulatory systems still needed to be verified.

The transcription of covR is negatively regulated by the CovR protein (28, 32). Compared to the wild-type A20 strain, the transcription of covR increased 2.95-fold ± 0.23-fold in its CovR_D53A mutant. Nonetheless, the decreased expression of the CovR protein was observed in the CovR_D53A mutants compared to that in the wild-type A20 and NZ131 strains (Fig. 1A and 2A). In addition, we also found that the expression level of CovR protein in the CovR_D53A mutants was lower than that in the CovR_T65A mutants (Fig. 1A and 2A). The underlying mechanisms of the decreased CovR protein in the CovR_D53A mutants remain to be further studied; however, our results showed that the transcription of rgg and speB was not repressed in the CovR_T65A mutants even though those mutants had higher levels of CovR protein than the CovR_D53A mutants (Fig. 1 and 2). In addition, the present study showed that trans complementation of nonphosphorylated CovR_D53 in strains that express phosphorylated CovR_D53 resulted in the repression of rgg and speB transcription (Fig. 4). These results support the idea that the phosphorylated CovR_D53 has minor effects on repressing rgg and speB transcription. Furthermore, it would be reasonable to suggest that the nonphosphorylated CovR_D53 inhibits rgg and speB transcription through interacting with unknown regulatory systems but not through direct binding to the rgg-speB intergenic region.

In the emm3-type strain, the transcription levels of speB in covR deletion, CovR_D53A, and CovR_T65A mutants were all upregulated compared to those of the wild-type strain (16). The present study showed that trans complementation of phosphorylated CovR_D53 in CovR_D53A mutants cannot derepress rgg and speB transcription (Fig. 5), further supporting the notion that the CovR protein acts as the negative regulator for rgg and speB. Nonetheless, in the emm1- and emm49-type strains used in this study, the transcription of speB was repressed in the CovR_D53A mutants but not in the CovR_T65A mutants (Fig. 1C and 2C). In addition, trans complementation of the CovR_D53A protein in strains that have the endogenous phosphorylated CovR_D53 protein resulted in the repression of rgg and speB transcription (Fig. 4). These results indicate that the nonphosphorylated CovR_D53 protein is crucial to mediate the repression of rgg and speB transcription. The different phenotypes observed in emm1-, emm49-, and emm3-type strains could be caused by the variations in the CovR/CovS phosphorelay signal among these strains (14). Nonetheless, how these variations influence the CovR/CovS regulatory activity to its target genes still needs to be further elucidated.

Rgg is required for triggering the expression of the cysteine protease SpeB (21, 33). In addition, inactivation of Rgg results in the increased resistance to oxidative stresses as well as the overproduction of bacterial virulence factors, such as DNase and streptolysin O (SLO) (11, 23, 34); these phenotypes may lead to the increase in bacterial virulence. Spontaneous mutations in rgg were also frequently found in clinical isolates from patients with invasive manifestations (11). Nonetheless, Friães et al. (17) showed that acquisition of the covS null allele but not the rgg null allele is associated with invasive infections. The present study showed that nonphosphorylated CovR_D53 enhances its repression of rgg transcription (Fig. 1 to 4). These results suggest that inactivation of covS may not only derepress the expression of CovR-controlled genes but also upregulate the expression of Rgg-suppressed virulence genes. These phenomena may provide an explanation as to why the covS but not the rgg mutation is associated with invasive infections.

CovR/CovS is a two-component regulatory system; however, the covS and covR null mutants are not identical. The present study demonstrated that nonphosphorylated CovR_D53 has a more dominant role in repressing rgg and speB transcription than does the phosphorylated CovR_D53 protein. Whether the nonphosphorylated CovR_D53 represses rgg transcription directly or indirectly through binding to the rgg-speB intergenic region still needs to be verified; however, these results reveal the potential role of CovR in regulating its controlled genes in invasive covS mutants.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

GAS A20 (emm1 type) and NZ131 (emm49 type) were described previously (30). GAS strains (Table 1) were cultured on Trypticase soy agar with 5% sheep blood or in tryptic soy broth (Becton, Dickinson and Company, Sparks, MD) supplemented with 0.5% yeast extract (TSBY). Escherichia coli DH5α was purchased from Yeastern Biotech Co., Ltd. (Taipei, Taiwan) and was cultured in Luria-Bertani (LB) broth at 37°C with vigorous aeration. When appropriate, the antibiotics chloramphenicol (25 µg/ml and 3 µg/ml for E. coli and GAS, respectively), spectinomycin (100 µg/ml), and erythromycin (150 µg/ml and 5 µg/ml for E. coli and GAS, respectively) were used for selection.

TABLE 1.

Plasmids and strains used in this study

Plasmid or strain	Parent strain	Descriptiona	Reference or source
Plasmids
pTRKL2		E. coli-Streptococcus shuttle vector	30
pCN143		Temperature-sensitive vector	30
pMW506		pSF152::covRΔcm	30
pCN114		pSF152::covSΔcm	This study
pCN129		pCN143::CovRD53A	30
pCN160		pCN143::covSΔcm	This study
pCN167		pCN143::CovRT65A (covR with A203G substitution)	This study
pCN169		pTRKL2::CovRD53A	This study
pCN176		pCN143::CovRD53A-T65A	This study
pCN180		pTRKL2::CovRT65A	This study
Strains
A20		emm1-type wild-type strain	40
NZ131		emm49-type wild-type strain	40
A20 ΔcovS	A20	A20 covS isogenic mutant	This study
A20 ΔcovR	A20	A20 covR isogenic mutant	30
NZ131 ΔcovS	NZ131	NZ131 covS isogenic mutant	This study
NZ131 ΔcovR	NZ131	NZ131 covR isogenic mutant	This study
A20CovR(D53A) mutant	A20	CovRD53A mutant strain	30
NZ131CovR(D53A) mutant	NZ131	CovRD53A mutant strain	This study
A20CovR(T65A) mutant	A20	CovRT65A mutant strain	This study
NZ131CovR(T65A) mutant	NZ131	CovRT65A mutant strain	This study
A20CovR(D53A-T65A) mutant	A20	CovRD53A and CovRT65A mutant strain	This study

^acm, chloramphenicol cassette.

DNA and RNA manipulations.

GAS genomic DNA extraction, RNA extraction, and reverse transcription were performed as described previously (35). Real-time PCR was performed in a 20-µl mixture containing 1 µl of cDNA, 0.8 µl of primers (10 µM), and 10 µl of SensiFAST SYBR Lo-ROX premixture (Bioline Ltd., London, UK), according to the manufacturer's instructions. Biological replicate experiments were performed from three independent RNA preparations in duplicate. The expression level of each target gene was normalized to that of gyrA and analyzed using the ΔΔC_T method (7500 software version 2.0.5; Applied Biosystems, Thermo Fisher Scientific, Inc.). In addition, all values of control and experimental groups were divided by the mean of control samples before statistical analysis (36). Primers used for real-time PCR analysis (Table 2) were designed by Primer3 (version 0.4.0; http://frodo.wi.mit.edu) according to the MGAS5005 sequence (NCBI RefSeq accession no. NC_007297.1).

TABLE 2.

Primers used in this study

Primer	Use	Sequence (5′ to 3′)a	Reference or source
covR-BamHI-F7	Construction	GCGGGATCCAAATGACAAAGAAAATTTTA	This study
covR-BamHI-R6	Construction	GCGGGATCCTAGGCACTTTCTTCTCAGAT	This study
covR-T65A-F	Construction	GATGGTTTTGAAGTGgCCCGTCGTTTGCAAACC	16
covR-T65A-R	Construction	GGTTTGCAAACGACGGGcCACTTCAAAACCATC	16
covS-F-2	Construction	CGGGATCCTCGTGAAGGGTTAGAAACTG	This study
covS-R-2-2	Construction	CCCCGGGCCATATGACTTATTTCTCACGAA	This study
covS-F-3-2	Construction	CCCCGGGGGCCCAGTCTAAAGAGAGTTAGAG	This study
covS-R-3-2	Construction	CCGGAATTCTTCGATTTCATCAACGTCCT	This study
Vec78_cat-F-1	Construction	CGCCGCGTTAACGATAGATTTATGATATAG	This study
Vec78_cat-R-1	Construction	CGCCGCGTTAACATTTATTCAGCAAGTCTT	This study
CovR/S-F-3	Construction	GCGGATCCGCTTGCAAGGGTTGTTTGATG	30
CovR-R-5	Construction	GCGGGATCCATGCAAGCCAGGAGATGATT	This study
Rgg-F-3	qPCR	TTTGAATGCCGAAACATAGAAGGTT	30
Rgg-R-2	qPCR	CTAATAACACCTTGACCAAGGCAAA	30
speB-F-2	qPCR	TGCCTACAACAGCACTTTGG	30
speB-R-2	qPCR	GGTAAAGTAGGCGGACATGC	30
gyrA-F-3	qPCR	CGTCGTTTGACTGGTTTGG	30
gyrA-R-3	qPCR	GGCGTGGGTTAGCGTATTTA	30

^aUnderlining indicates restriction enzyme sites; lowercase letters denote mutated nucleotides.

Construction of covR deletion, covS deletion, CovR_D53A, CovR_T65A, and CovR_D53A-T65A mutants.

The covR gene of NZ131 was interrupted by pMW506 with a double-cross homologous recombination, as described previously (30). To construct the covS isogenic mutants, the 592-bp covS upstream region and 659-bp covS downstream region were amplified with primers covS-F-2, covS-R-2-2, covS-F-3-2, and covS-R-3-2 and ligated into the vector pSF152 (30) to generate pCN114 (Tables 1 and 2). The chloramphenicol cassette of vector 78 (37) was amplified using primers Vec78_cat-F-1 and Vec78_cat-R-1 and ligated into pCN114 with the SmaI site. The upstream and downstream regions of covS and the chloramphenicol cassette in pSF152 were amplified by primers covS-F-2 and covR-3-2 and ligated into TA Cloning vector (Yeastern Biotech Co., Ltd., Taipei, Taiwan) and then transferred to the temperature-sensitive vector pCN143 (30) with the BamHI site (pCN160, Table 1). The amino acid residue substitution of CovR protein was generated by overlap PCR, as described previously (30). The T65A amino acid substitution of CovR was generated by primers CovR-BamHI-F7, CovR-T65A-R, CovR-BamHI-R6, and CovR-T65A-F from the wild-type A20 strain and its CovR_D53A mutant. PCR products were ligated into the TA Cloning vector (Yeastern Biotech Co., Ltd.), and the amino acid substitution was confirmed by Sanger sequencing. The mutated covR genes were transferred to pCN143 with the BamHI site. The constructed plasmids (pCN160, pCN167, and pCN176) were transfected into strain A20 or NZ131 by electroporation, as described previously (38), and cultured at 30°C with the spectinomycin selection. Transformants were then transferred to 37°C to force the plasmid integration. Finally, transformants in which plasmid excised from the chromosome via recombination were selected by chloramphenicol (for covS and covR mutants) or in the blood agar plate (for CovR_D53A, CovR_T65A, and CovR_D53A-T65A mutants) at 30°C. The covS and covR deletion mutants and GAS strains with the D53A substitution or D53A and T65A substitutions of CovR were confirmed by Sanger sequencing. All primers used are shown in Table 2.

Construction of vector control and covR trans-complemented strains.

The E. coli-Streptococcus shuttle vector pTRKL2 was described previously (30). To construct covR trans-complemented strains, the genomic DNAs extracted from A20 CovR_D53A and CovR_T65A mutants served as the templates, and the covR gene, including its native promoter (1.3 kb), was amplified with primers CovR/S-F-3 and CovR-R-5 (Table 2). The PCR products were ligated into the BamHI site of pTRKL2, and constructed plasmids (pCN169 and pCN180) were transformed into CovR_T65A or CovR_D53A mutants with electroporation, as described previously (39).

Western blotting for SpeB and CovR proteins.

Bacteria were cultured in TSBY broth for 8 h, and the culture supernatant was collected and sterilized with a 0.22-µm-pore-size filter for detecting SpeB protein. Thirty microliters of bacterial culture supernatants was mixed with 6× protein loading dye and separated by 12% SDS-PAGE gels. In order to analyze the expression of CovR protein, bacteria that were grown to exponential phase (6 h of incubation) were collected, and total proteins were extracted by the bead beater (Mini-beadbeater; BioSpec Products, Inc., Bartlesville, OK) (30). Ten micrograms of total protein was subjected to analysis with 12% SDS-PAGE gels. Separated proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore, Billerica, MA), and membranes were blocked by 5% skim milk in PBST buffer (PBS containing 0.2% Tween 20) at 37°C for 1 h. SpeB and CovR proteins were detected by the anti-SpeB antibody (Toxin Technology, Inc., Sarasota, FL) and anti-CovR serum (30), respectively. After hybridization, the membrane was washed with PBST buffer and hybridized with a secondary antibody, the peroxidase-conjugated goat anti-rabbit IgG (1:10,000 dilution; Cell Signaling Technology, Inc., Danvers, MA), at room temperature for 1 h. The blot was developed using Pierce ECL Western blotting substrate (Thermo Fisher Scientific, Inc., Rockford, IL), and the signal was detected using the Gel Doc XR+ system (Bio-Rad).

Phos-tag Western blotting.

Bacterial protein was extracted according to a previously described method (39). Ten micrograms of bacterial proteins was mixed with 6× protein loading dye (without boiling) and separated by 10% SDS-PAGE gels containing 10 µM Phos-tag (Wako Pure Chemical Industries Ltd., Richmond, VA) and 0.5 µM MnCl₂. Phosphorylated and nonphosphorylated proteins were separated on Phos-tag SDS-PAGE gels for 120 to 140 min at 100 V at 4°C. Protein transfer, membrane blocking, hybridization, and signal detection were performed as described previously (39).

Statistical analysis.

Statistical analysis was performed by using the Prism software, version 5 (GraphPad, San Diego, CA). Significant differences in multiple groups were determined using analysis of variance (ANOVA). Posttest for ANOVA was analyzed by Tukey’s honestly significant difference test. A P value of <0.05 was taken as significant.