An amino-terminal signal peptide of Vfr protein negatively influences RopB-dependent SpeB expression and attenuates virulence in Streptococcus pyogenes

Authors

  • Samuel A. Shelburne III,

    1. Department of Infectious Diseases, MD Anderson Cancer Center, Houston, TX, USA
    Search for more papers by this author
  • Randall J. Olsen,

    1. Center for Molecular and Translational Human Infectious Diseases Research, The Methodist Hospital Research Institute, and Department of Pathology and Genomic Medicine, The Methodist Hospital System, Houston, TX, USA
    Search for more papers by this author
  • Nishanth Makthal,

    1. Center for Molecular and Translational Human Infectious Diseases Research, The Methodist Hospital Research Institute, and Department of Pathology and Genomic Medicine, The Methodist Hospital System, Houston, TX, USA
    Search for more papers by this author
  • Nicholas G. Brown,

    1. Department of Pharmacology, Baylor College of Medicine, Houston, TX, USA
    Search for more papers by this author
  • Pranoti Sahasrabhojane,

    1. Department of Infectious Diseases, MD Anderson Cancer Center, Houston, TX, USA
    Search for more papers by this author
  • Ebru M. Watkins,

    1. Center for Molecular and Translational Human Infectious Diseases Research, The Methodist Hospital Research Institute, and Department of Pathology and Genomic Medicine, The Methodist Hospital System, Houston, TX, USA
    Search for more papers by this author
  • Timothy Palzkill,

    1. Department of Pharmacology, Baylor College of Medicine, Houston, TX, USA
    Search for more papers by this author
  • James M. Musser,

    1. Center for Molecular and Translational Human Infectious Diseases Research, The Methodist Hospital Research Institute, and Department of Pathology and Genomic Medicine, The Methodist Hospital System, Houston, TX, USA
    Search for more papers by this author
  • Muthiah Kumaraswami

    Corresponding author
    1. Center for Molecular and Translational Human Infectious Diseases Research, The Methodist Hospital Research Institute, and Department of Pathology and Genomic Medicine, The Methodist Hospital System, Houston, TX, USA
      E-mail mkumaraswami@tmhs.org; Tel. (+1) 713 441 5252; Fax (+1) 713 441 3447.
    Search for more papers by this author

E-mail mkumaraswami@tmhs.org; Tel. (+1) 713 441 5252; Fax (+1) 713 441 3447.

Summary

Streptococcal pyrogenic exotoxin B (SpeB) is an extracellular cysteine protease that is a critical virulence factor made by the major human pathogen group A Streptococcus (GAS). speB expression is dependent on the regulator of proteinase B (RopB) and is upregulated with increasing cell density and during infection. Because computer modelling suggested significant structural similarity between RopB and peptide-sensing regulatory proteins made by other Gram-positive bacteria, we hypothesized that speB expression is influenced by RopB–peptide interactions. Inactivation of the gene (vfr) encoding the virulence factor related (Vfr) protein resulted in increased speB transcript level during the exponential growth phase, whereas provision of only the amino-terminal region of Vfr comprising the secretion signal sequence in trans restored a wild-type speB expression profile. Addition of the culture supernatant from a Vfr signal peptide-expressing GAS strain restored wild-type speB transcript level to a vfr-inactivated isogenic mutant strain. A distinct peptide in the Vfr secretion signal sequence specifically bound to recombinant RopB. Finally, overexpression of the Vfr secretion signal sequence significantly decreased speB transcript level and attenuated GAS virulence in two mouse models of invasive infection. Taken together, these data delineate a previously unknown small peptide-mediated regulatory system that controls GAS virulence factor production.

Introduction

Precise temporal regulation of virulence factor production is critical to microbial pathogenesis (Somerville and Proctor, 2009; Yoon et al., 2009; Liu et al., 2011). The genomic era has brought significant advances in understanding the range of genes influenced by specific transcriptional regulators (Goodman and Lory, 2004). However, information regarding the molecular mechanisms by which stand-alone transcription factors respond to environmental stimuli to fine tune expression of virulence-factor-encoding genes is limited (Withers et al., 2001; Declerck et al., 2007). A better understanding of how pathogenic microbes regulate virulence factor production is critical to the goal of designing novel antimicrobials that can interfere with such signalling pathways (Kreikemeyer et al., 2003; Hung et al., 2005; George et al., 2008; Rasko et al., 2008).

Group A Streptococcus (GAS) is a Gram-positive bacterium that has long served as a model pathogen for investigating bacterial virulence factor regulation (Kreikemeyer et al., 2003; McIver, 2009). GAS causes a broad spectrum of human infections ranging from mild pharyngitis and impetigo to life-threatening necrotizing fasciitis and streptococcal toxic shock syndrome (Olsen et al., 2009). Among the many virulence factors made by GAS, a secreted cysteine protease known as streptococcal pyrogenic exotoxin B (SpeB) is critical for virulence and dissemination of infection (Lukomski et al., 1997; 1998; 1999; Svensson et al., 2000). SpeB production is growth phase-dependent and increases as GAS transitions from the exponential to the stationary growth phase (Neely et al., 2003). Moreover, speB transcript is significantly increased during infection compared with growth in a standard laboratory medium (Loughman and Caparon, 2006). Given its significance in GAS pathogenesis, SpeB is subject to multiple layers of temporal and environmental regulation (Carroll and Musser, 2011).

Transcription of speB is directly controlled by a global regulator known as regulator of proteinase B (RopB) (Lyon et al., 1998; Chaussee et al., 1999; Carroll and Musser, 2011). RopB belongs to the Rgg family of transcription regulators that are present in a diverse array of pathogenic low G+C Gram-positive bacteria (McIver, 2009). The speB promoter region has various cis-acting elements including two putative RopB binding sites with pseudo-palindromic sequences that are separated from each other by approximately ∼120 bp (Fig. 1A) (Neely et al., 2003). RopB is required for activation of speB transcription, but attempts to activate speB expression at exponential phase of growth by ectopic provision of RopB failed to cause early onset of speB expression (Neely et al., 2003). Thus, it appears that additional signals are required for temporal regulation of SpeB production, but the identity and nature of such signals remain poorly understood (Chaussee et al., 1997; Neely et al., 2003).

Figure 1.

Organization of the speB gene region and RopB model. A. Organization of the speB gene region in strain MGAS10870. The divergently transcribed speB and ropB genes are shown as rectangle boxes and labelled. The bent arrows above the line indicate two transcription start sites of speB and bent arrow below denote the transcription start site of ropB. Red arrows indicate putative palindromic sites that have the RopB binding site. Data regarding transcription start sites and RopB binding sites are derived from Neely et al. (2003). B. RopB structure predicted by I-TASSER protein modelling server (http://zhanglab.ccmb.med.umich.edu/I-TASSER). Ribbon representation of the predicted RopB homodimer is shown and the individual subunits of RopB dimer are coloured in blue and pink. The amino- and carboxy-terminus of one subunit are indicated as N and C, respectively, and those of the second subunit are indicated with a prime (′). The putative DNA-binding domain in the amino-terminus and the oligomerization/regulatory domain in the carboxy-terminus of one subunit are labelled. The model shown is modified from fig. 2 from Carroll et al. (2011).

To better understand the molecular mechanism of gene regulation by RopB, we recently performed three-dimensional computer modelling of RopB (Carroll et al., 2011). We discovered that RopB has significant predicted structural homology with PlcR from Bacillus cereus and PrgX from Enterococcus faecalis, the founding members of the Rap/Npr/PlcR/PrgX (RNPP) family of quorum-sensing regulators (Fig. 1B) (Rocha-Estrada et al., 2010). Members of the RNPP family use secreted peptides as regulatory signals and alter gene expression in response to changes in bacterial cell density (Declerck et al., 2007; Fontaine et al., 2010; Mashburn-Warren et al., 2010; Rocha-Estrada et al., 2010; Fleuchot et al., 2011). Given its predicted structural homology with PlcR and PrgX, it is possible that RopB uses secreted peptides as regulatory signals to control speB expression. In this regard, it is noteworthy that 15 years ago it was reported that inactivation of the GAS genes encoding oligopeptide permeases (Opp) or dipeptide permeases (Dpp) drastically reduced speB expression (Podbielski et al., 1996; Podbielski and Leonard, 1998) suggesting that peptide transport plays a role in speB regulation.

Ma et al. (2009) recently reported that inactivation of the gene (vfr) encoding virulence factor related (Vfr) protein increased speB expression and decoupled speB expression from growth phase dependence (Ma et al., 2009). Vfr lacks typical characteristics of a DNA-binding protein and the molecular mechanism by which it alters SpeB production is not known. Vfr contains a putative amino-terminal secretion signal sequence (Ma et al., 2009). After processing by peptidases, signal peptides in the extracellular milieu can be imported to the cytosol where they interact with transcriptional regulators (Lazazzera, 2001; Antiporta and Dunny, 2002; Kozlowicz et al., 2006). In this study we tested the hypothesis that a peptide derived from the amino-terminal region of the Vfr secretion signal sequence directly interacts with RopB to regulate SpeB production. Our findings confirmed this hypothesis and elucidated a new mechanism of GAS virulence factor regulation.

Results

Kinetics of vfr, ropB and speB transcript levels in a serotype M3 GAS strain

The only previous study of the vfr gene reported that its transcript peaks in the exponential growth phase and gradually decreases in stationary phase (Ma et al., 2009). Given that there is significant strain-to-strain variation in gene expression profiles among GAS isolates (Dmitriev et al., 2008), we first measured the transcript levels of vfr, ropB and speB in a serotype M3 strain (MGAS10870) at different growth stages (Fig. 2A). MGAS10870 was chosen for this study because it is an invasive isolate that has been fully sequenced, and it contains the most common RopB isoform (Beres et al., 2010; Carroll et al., 2011). Consistent with previous observations (Neely et al., 2003; Carroll et al., 2011), in standard laboratory medium ropB transcript reached a maximum level in late exponential phase, and the speB transcript level peaked in the stationary phase (Fig. 2B). The kinetics of the vfr and ropB transcripts paralleled one another, reaching their highest levels at late exponential phase and returning to a basal level in stationary phase (Fig. 2B). These data suggest that the presence of Vfr in the late exponential growth phase may be inhibiting RopB-mediated speB transcription, and this inhibition is subsequently relieved in the stationary phase of growth.

Figure 2.

vfr inactivation alters speB expression profile. A. Growth curve of indicated strains in THY broth. Samples were collected at early exponential (EE, A600 ∼ 0.3), mid-exponential (ME, A600 ∼ 0.6), late exponential (LE, A600 ∼ 1.0) and stationary phase (SP, A600 ∼ 1.7) for transcript and protein analysis. Arrows indicated the sampling time points. B. Transcript levels of speB, ropB and vfr in strain MGAS10870 measured by Taqman qRT-PCR. C. speB transcript level in strains MGAS10870 and isogenic mutant strain Δvfr. For (B) and (C) duplicate biological replicates were grown on two separate occasions and analysed in duplicate. Data graphed are mean ± standard deviation. D. Representative western immunoblot analysis of SpeB in filtered growth media from strain MGAS10870 and isogenic mutant strain Δvfr. Each experiment was performed on three different occasions with essentially identical results. Growth media collected at indicated sampling points from strains MGAS10870 and Δvfr were analysed using anti-SpeB polyclonal rabbit antibody and chemiluminescence. The bands corresponding to the mature form of SpeB are labelled (SpeBm) and indicated by an arrow. E. Assay of SpeB protease activity by azocasein hydrolysis assay. Filtered growth media collected at indicated growth phases from strains MGAS10870 and Δvfr were assayed for secreted SpeB protease activity. Azocasein hydrolysis by SpeB was determined by measuring the absorption at A360.

Vfr negatively influences speB expression in the late exponential growth phase

To investigate the role of vfr in regulation of speB expression, we used non-polar insertional mutagenesis to construct an isogenic mutant derivative of strain MGAS10870 in which the entire vfr gene was replaced by a spectinomycin-resistance cassette. For the purpose of clarity, strain 10870Δvfr will be referred to hereafter as Δvfr. The growth kinetics of strains MGAS10870 and Δvfr did not differ significantly in rich laboratory media (Fig. 2A). Next, we compared speB transcript levels of the wild-type and mutant strains at different phases of growth. Very little speB transcript was expressed in the mid-exponential growth phase by either strain MGAS10870 or strain Δvfr (Fig. 2C). However, compared with wild type, strain Δvfr had an approximately 64-fold increase in the speB transcript level at the late exponential growth phase (Fig. 2C). The speB transcript level in mutant strain Δvfr remained elevated in the stationary growth phase, but the difference at this point was only about twofold compared with wild type (Fig. 2C).

To determine if these differences in transcript level resulted in variation in the amount of SpeB made by the strains, we assessed the presence of secreted SpeB in the extracellular medium by Western immunoblotting and SpeB protease activity via by an azocasein-cleavage assay. Consistent with the transcript data, immunoreactive SpeB was detected only in the growth medium of strain Δvfr at the late exponential growth phase, whereas SpeB was detected in the growth medium of both strains at the stationary phase (Fig. 2D). Next, we measured the SpeB protease activity by a chromogenic azocasein hydrolysis assay. Strain Δvfr had increased SpeB protease activity in the late exponential and stationary growth phase compared with wild type, although the differences observed at the stationary growth phase were more pronounced compared with those observed at the late exponential growth phase (Fig. 2E). Taken together, these data indicate that, directly or indirectly, Vfr negatively influences speB expression and SpeB production, with the majority of the regulatory effect occurring in the late exponential growth phase.

Overexpression of Vfr negatively influences speB expression

To test the hypothesis that ectopic expression of vfr by a plasmid would complement the vfr-inactivated isogenic mutant strain, we constructed plasmid pDC-vfr containing the entire vfr gene and its promoter region. The complemented mutant strain (strain Δvfr:pDC-vfr), strain Δvfr with empty vector (strain Δvfr:pDC) and the wild-type strain with empty vector (strain 10870:pDC) did not differ significantly in growth (Fig. 3A). To ensure that vfr in the pDC-vfr construct constituted a functional transcriptional unit, we measured the level of vfr transcript in the complemented strain by qRT-PCR. The pattern of the relative abundance of vfr transcripts at different growth stages was preserved in the complemented strain. We also found an overall increase (≥ 16-fold) in vfr transcript level in the complemented strain relative to wild-type strain MGAS10870, indicating that the complemented strain overexpressed vfr transcript (Fig. 3B). Importantly, the presence of the vfr-overexpressing plasmid in the vfr-inactivated strain restored a speB phenotype similar to that of wild type at the late exponential growth phase (Fig. 3C). Suppression of speB transcription by pDC:vfr was extended well into the stationary phase of growth, as the increased speB transcript level characteristic of stationary phase was absent in strain Δvfr:pDC-vfr (Fig. 3C). Consistent with the qRT-PCR data, no detectable immunoreactive SpeB was made by the complemented mutant strain (Fig. 3D). Similarly, negligible SpeB protease activity was detected in culture supernatants prepared from strain Δvfr:pDC-vfr samples with levels being comparable to those of the isogenic mutant SpeB-inactivated strain (strain 10870ΔspeB) (Fig. 3E). Thus, the speB phenotype of the Vfr-inactivated strain can be reversed by trans-complementation with vfr. In addition, the negative regulatory influences of Vfr on speB can be extended into stationary growth phase by vfr overexpression.

Figure 3.

The secretion signal sequence of Vfr restores a wild-type speB expression profile to strain Δvfr. A. Growth pattern of indicated strains in THY broth. Samples were collected at mid-exponential (ME A600 ∼ 0.6), late exponential (LE A600 ∼ 1.0) and stationary phase (SP A600 ∼ 1.7) for transcript and protein analysis at time points designated by arrows. B. vfr transcript levels as measured by Taqman qRT-PCR at indicated growth phase. C. speB transcript levels in complemented strains at indicated growth phases. For (B) and (C), biological replicates were analysed in duplicate on two separate occasions with data graphed being mean ± standard deviation. D. Western immunoblot analysis of SpeB in filtered growth media prepared from indicated strains at indicated growth phases. E. Assay of SpeB protease activity in the culture supernatant by azocasein hydrolysis. Azocasein hydrolysis by SpeB was assessed by absorption at A360. Samples were analysed in triplicates on two separate occasions with data graphed are mean ± standard deviation.

Overexpression of only the amino-terminal secretion signal sequence of Vfr negatively influences speB expression

To test the hypothesis that the element in Vfr responsible for regulating SpeB resides in the amino-terminal secretion signal sequence, we cloned the vfr gene region that encodes only the amino-terminal 39 amino acid residues of Vfr and the vfr promoter region into pDC123 to create plasmid pDC-pep. Consistent with our hypothesis, the complemented strain Δvfr:pDC-pep had a speB transcript level profile essentially identical to that observed for the wild-type strain containing empty vector (Fig. 3C). The decreased speB transcript level observed in the late exponential growth phase for strain Δvfr:pDC-pep was comparable to that of full-length vfr complementation. These results indicate that the secretion signal peptide of Vfr alone is sufficient to negatively influence speB expression at the late exponential growth phase (Fig. 3C). Similarly, results from western immunoblotting analysis showed that SpeB production in strain Δvfr:pDC-pep was comparable to wild type in the late exponential phase but there was decreased production of the mature SpeB isoform in strain Δvfr:pDC-pep in the stationary phase (Fig. 3D). The azocasein cleavage assay was too insensitive to reveal differences in SpeB protease activity in exponential-phase samples (Fig. 2E). Thus, we performed the assay only with stationary-phase samples. Compared with the vfr-inactivated strain containing empty vector, SpeB protease activity was significantly reduced at stationary phase in both complemented strains, with full-length complementation causing more pronounced regulation (Fig. 3E). In the aggregate, these data show that the amino-terminal secretion signal sequence of Vfr is sufficient for the negative influence of Vfr on speB expression in the late exponential phase of growth.

A factor in the culture supernatant of a Vfr-expressing strain restores a wild-type speB phenotype to strain Δvfr

Multiple attempts to synthesize peptides corresponding to regions of the Vfr secretion signal sequence failed to yield synthetic peptide of reliable sequence, high purity, high yield and solubility, likely due to marked hydrophobicity of the peptides. Thus, we could not investigate whether direct addition of a synthetic peptide containing the Vfr secretion signal sequence to the growth medium inhibited speB expression. Rather, we performed culture supernatant-swapping experiments designed to determine if the Vfr secretion signal peptide is present in the extracellular environment and then internalized. Strain Δvfr was grown to mid-exponential phase (A600 ∼ 0.6), cells were collected by centrifugation, and the cell pellet was suspended in the filtered growth media prepared from late exponential phase cultures (A600 ∼ 0.8) from one of the following strains: Δvfr:pDC, strain Δvfr:pDC-vfr or Δvfr:pDC-pep. No immunoreactive SpeB was present in the pre-swap samples (data not shown). Samples were collected 1 h post swapping (A600 ∼ 1.0), and speB transcript level and secreted SpeB were measured by qRT-PCR and Western immunoblotting respectively. Consistent with the hypothesis that the secretion signal sequence of Vfr is present in the extracellular milieu and subsequently imported into the cytoplasm to mediate speB regulation, significantly less speB transcript level was observed in strain Δvfr grown in the medium derived from strain Δvfr:pDC-vfr or strain Δvfr:pDC-pep compared with the same strain exposed to the medium derived from strain Δvfr-pDC (Fig. 4A and B). Similarly, culture supernatant derived from strain Δvfr:pDC-vfr or strain Δvfr:pDC-pep, but not strain Δvfr:pDC, drastically reduced the production of immunoreactive SpeB (Fig. 4B). Thus, speB expression is negatively influenced by Vfr or a product derived from it that is present in the culture supernatant.

Figure 4.

Vfr represses SpeB production by a product present in the culture supernatant. A. speB transcript level analysis of mutant strain Δvfr grown in the culture supernatant the indicated strains for 1 h as described in Experimental procedures. Two biological replicates were performed on two separate days and analysed in duplicate with data graphed being mean ± standard deviation. P-value is derived from a two-sample t-test. B. Western immunoblot analysis of SpeB in the growth medium of mutant strain Δvfr grown for 30 min and 60 min in the culture supernatant derived from indicated strains as described in (A). C. Western immunoblot analysis of SpeB in the growth medium of mutant strain Δvfr grown for 60 min in either the total culture supernatant or culture supernatant filtered with YM-3 (< 3 kDa) membrane derived from indicated strains. Proteinase K-treated culture supernatants (Prot K) used for swapping are indicated.

To determine the chemical nature and the likely molecular mass of the regulatory factor, we performed a swap assay with culture supernatants of Δvfr:pDC-vfr that is sieved through YM-3 (∼3 kDa molecular weight cut-off filter) membrane filter and/or treated with proteinase K. In accordance with the hypothesis that the secreted regulatory factor has low molecular mass, strain Δvfr grown in the low-molecular-weight component (≤ 3 kDa) of the culture supernatant obtained from Δvfr:pDC-vfr strain had negligible levels of secreted SpeB compared with that of strain Δvfr grown in the low-molecular-weight component of supernatant derived from strain Δvfr:pDC (Fig. 4C). Consistent with the hypothesis that the regulatory factor present in the culture supernatant of Δvfr:pDC-vfr is proteinaceous, strain Δvfr grown in the proteinase K-treated culture supernatant of Δvfr:pDC-vfr failed to repress SpeB production (Fig. 4C). Together, these data indicate that the secreted factor present in the culture supernatant of Δvfr:pDC-vfr that negatively regulates speB expression is a low-molecular-weight peptide.

A distinct region of the amino-terminal secretion signal sequence of Vfr interacts with RopB

Inasmuch as the predicted structural homologues of RopB interact with secreted peptides to mediate gene regulation (Shi et al., 2005; Declerck et al., 2007), we next investigated whether a distinct region of the amino-terminal secretion signal sequence of Vfr directly interacts with RopB using a synthetic peptide array. Overlapping 17-amino-acid synthetic peptides that span the entire predicted signal sequence of Vfr (amino acid residues 1–49) were generated by SPOT synthesis on a cellulose membrane. The resulting synthetic peptide array encompassed all three components commonly found in secretion signal sequences, namely a basic n-region, a hydrophobic h-region and a c-region with the peptidase cleavage site (Fig. 5A). Interaction between the synthetic peptides and purified recombinant hexa-histidine-tagged RopB (Fig. 5B) was assessed. The data show that RopB interacts with a distinct region of the amino-terminal secretion signal sequence of Vfr (Fig. 5C). The amino acids required for RopB binding are located in the h-region and c-region fragment of the Vfr signal sequence (Fig. 5C). Omission of amino acid residues R39 at the carboxy-terminus and L32 at the amino-terminus markedly decreased RopB binding, indicating the likely boundaries of the RopB binding site in the Vfr signal sequence (Fig. 5C). Thus, we conclude that a region of the amino-terminal secretion signal sequence of Vfr interacts directly with RopB suggesting a mechanism by which Vfr alters speB expression.

Figure 5.

Recombinant RopB binds the Vfr amino-terminus signal sequence. A. Primary sequence alignment of Vfr signal sequence with characterized Gram-positive bacterial pheromones, PapR from Bacillus cereus; and cCF10, cAD1 and iAD1 from Enterococcus faecalis. Regions corresponding to the three major components of signalling peptides, n-region, h-region and c-region, are underlined and labelled. Amino acid sequences corresponding to mature peptides of each pheromone and the putative Vfr mature peptide were underlined. Amino acid sequences of Vfr included in the peptide array are boxed. B. SDS-polyacrylamide gel with increasing concentrations of purified recombinant hexa-histidine tagged RopB. The corresponding positions of molecular weight markers (in kilodaltons) are indicated. C. Peptide array of overlapping 17-mer peptides that spans the entire putative Vfr signal sequence was generated using SPOT synthesis on a cellulose membrane. Membrane was incubated with purified recombinant hexa-histidine-tagged RopB, and bound protein was detected with anti-hexahistidine tag antibodies and chemiluminescence. The red arrows indicate the likely amino-terminal and carboxy-terminal boundaries of the RopB binding site in the Vfr secretion signal sequence, located at L32 and R39 amino acid residues respectively. The numbering of amino acid residues indicates their position within the Vfr secretion signal sequence.

Overexpression of the vfr secretion signal sequence significantly decreases GAS virulence in mouse models of invasive infection

To test the hypothesis that the vfr gene and its influence on speB expression contribute to GAS virulence, we compared strains MGAS10870, ΔspeB, Δvfr, Δvfr:pDC-vfr and Δvfr:pDC-pep in two different mouse models of invasive infection. First, mice were injected intraperitoneally with 1 × 107 colony-forming units (cfu) of each strain, and the ability to cause near mortality was monitored for 7 days. The strains could be divided into two groups based on virulence phenotype. Wild-type strain MGAS10870 and mutant strain Δvfr caused significantly higher mortality compared with strains Δvfr:pDC-vfr, Δvfr:pDC-pep or ΔspeB (P < 0.01 when comparing any two strains in different groups) (Fig. 6A). There was no significant difference in mortality among strains within either of the two groups.

Figure 6.

Overexpression of the vfr signal sequence decreases GAS virulence in mouse models of infection. A. Fifteen CD-1 mice were inoculated with each indicated strain intraperitoneally and the near mortality was monitored. Kaplan-Meier survival curve with P-values derived by log rank test. B. Ten CD-1 mice were injected intramuscularly with 1 × 107 cfu of each strain. Kaplan-Meier survival curve with P-values derived by log rank test. C. Histopathological analysis of muscular lesions from mice infected with each indicated strain. Areas of host tissue damage are circled, whereas confined, less destructive lesions are indicated by arrows. D and E. TaqMan qRT-PCR analysis of speB and vfr gene transcript levels from mouse tissue samples infected with indicated strains. Samples were collected from the lesions of three mice per infecting strain and analysed in duplicate. Resulting data were graphed as mean ± standard deviation with P-values derived from two-sample t-test.

GAS cause a wide range of human infections with different virulence factors having varying impact depending on the site of infection (Engleberg et al., 2004; Olsen et al., 2009). Therefore, we also tested the role of vfr in GAS virulence using an intramuscular infection model that mimics human necrotizing fasciitis. Consistent with the results obtained from the intraperitoneal injection experiments, strains MGAS10870 and Δvfr were significantly more virulent than strains Δvfr:pDC-pep, Δvfr:pDC-vfr and ΔspeB (P < 0.01 when comparing any two strains in different groups) (Fig. 6B). To better understand the role of vfr in GAS pathogenesis at the cellular level, we performed histopathological analysis of intramuscular tissue lesions. The strains from the two groups caused contrasting pathological characteristics. Lesions from the mice infected with strains MGAS10870 and Δvfr showed severe host-tissue damage and disseminated infection (Fig. 6C). In comparison, lesions from mice infected with strains Δvfr:pDC-pep, Δvfr:pDC-vfr and ΔspeB were small, circumscribed and bordered by healthy muscle tissue (Fig. 6C).

To gain insight into the molecular mechanism by which overexpression of the Vfr secretion signal sequence reduces GAS virulence in vivo, we performed qRT-PCR analysis for speB and vfr transcript levels in infected mouse tissue. Similar to our observations for the stationary phase of growth in standard laboratory medium (Fig. 2C), there was no significant difference in in vivo speB transcript level between wild-type strain MGAS10870 and mutant strain Δvfr (Fig. 6D). In contrast, there was a significant decrease in speB transcript level in strains Δvfr:pDC-vfr and Δvfr:pDC-pep (Fig. 6D). As expected, no detectable vfr transcript was made by strain Δvfr, whereas vfr transcripts were present at a relatively high level in both complemented strains, Δvfr:pDC-vfr and Δvfr:pDC-pep (Fig. 6E). Interestingly, vfr transcript level was nearly undetectable in strain MGAS10870 suggesting strong repression of vfr expression during infection (Fig. 6E). Together, these data indicate that inactivation of Vfr does not significantly affect GAS virulence but overexpression of the Vfr secretion signal sequence significantly reduces speB transcript level in the host and attenuates GAS virulence.

Discussion

Many bacterial pathogens sense changes in population density in their environment via secreted molecules and respond by altering expression of genes involved in diverse cellular processes such as competence, conjugation, biofilm formation and pathogen–host interaction (Bassler and Losick, 2006). Virulence gene expression at high cell density can result in host tissue damage and promote microbial dissemination and survival (Bassler, 1999; De Kievit and Iglewski, 2000; Van Delden and Comte, 2001). Conversely, pathogens also have mechanisms to downregulate virulence gene expression at low cell density to avoid premature elicitation of host defence responses which could promote bacterial clearance from the site of infection (Kozlowicz et al., 2006). The data presented herein provide new insights into virulence gene regulation in response to changes in the GAS growth phase. We have shown that at low cell density and in a growth phase-dependent manner, the amino-terminal secretion signal sequence of Vfr negatively influences expression of a key virulence factor, SpeB, by interfering with RopB-dependent speB transcription.

A previous study reported that genetic inactivation of vfr relieves growth phase-dependent repression of speB expression (Ma et al., 2009). It was proposed that the Vfr influence occurs through protein–protein interactions with molecules in other regulatory networks (Ma et al., 2009). Our data show that the SpeB-related phenotype in Vfr inactivated strain can be reversed with only the region of vfr encoding the secretion signal sequence. This finding indicates that the speB regulatory function of vfr resides in the amino-terminal 40 amino acids of Vfr. Known regulatory peptides in Gram-positive bacteria are generally derived either from a small peptide-encoding open reading frame or from the amino-terminal secretion signal sequence of lipoproteins (Lazazzera, 2001; Antiporta and Dunny, 2002). In this regard, we note that the Vfr propeptide region has many of the characteristics of Gram-positive pheromones including a basic amino acid-rich amino-terminus region (n-region) and a hydrophobic central region (h-region). However, the Vfr secretion signal sequence lacks a carboxy-terminal cysteine residue commonly required for signal peptidase II cleavage and secretion (Von Heijne, 1990; Paetzel et al., 2002) (Fig. 5A). Interestingly, recently identified pheromone propeptides from Streptococcus mutans and Streptococcus thermophilus also lack the characteristic cysteine residue at the cleavage site and are believed to be processed and secreted by unknown secretion systems (Fontaine et al., 2010; Mashburn-Warren et al., 2010). Additionally, Chang et al. recently reported that short-hydrophobic peptides (SHP2 and SHP3) interact with RopB-like regulators (designated Rgg2 and Rgg3 in that study) in GAS and mediate gene regulation (Chang et al., 2011). The Vfr–RopB interaction seems fundamentally distinct from those of the SHP-Rgg as shp2 and shp3 are encoded immediately upstream of the respective rgg gene and are not part of a larger open reading frame as is the case with the Vfr amino-terminal signal sequence. Thus, in concert with recently published data, our findings add to the rapidly expanding appreciation of the role of the diverse role of small peptide molecules in GAS pathophysiology.

Similar to many other bacterial transcription regulators, RopB may function as a dimer, and two RopB binding sites with inverted repeats have been identified in the speB promoter region that are separated from each other by ∼130 bp (Fig. 1B) (Neely et al., 2003). For both PrgX and PlcR (structural homologues of RopB) peptide binding triggers structural changes in their respective effector molecule that leads to altered DNA binding and/or oligomerization, thereby altering gene expression (Shi et al., 2005; Kozlowicz et al., 2006; Declerck et al., 2007). PlcR–promoter interactions and the consequent transcription activation by PlcR occur only in the presence of the activating peptide PapR (Slamti and Lereclus, 2002). On the other hand, binding of either inhibitory peptide, iCF10, or activating peptide, cCF10, to PrgX controls the regulatory properties of PrgX. Interactions between PrgX and iCF10 promote PrgX tetramerization, resulting in the occupation of both PrgX binding sites by DNA looping and repression (Kozlowicz et al., 2006). Conversely, interactions between cCF10 and PrgX disrupt the tetramerization interface and the dimeric PrgX activates gene expression (Shi et al., 2005). Thus, the Vfr regulatory peptide may influence speB expression by modulating oligomerization of RopB, thereby interfering with or otherwise altering DNA binding at one or both of the speB promoter binding sites.

The negative regulation of speB expression by Vfr is more pronounced during the exponential phase of growth, with repression being relieved in the stationary growth phase. Our data indicate that removal of Vfr-mediated negative regulation of speB expression by downregulation of vfr gene expression at stationary phase contributes to transcriptional activation of speB (Fig. 2B). Similarly, repression of vfr transcript level during infection may help to explain the high levels of speB transcript that are typically observed following GAS infection (Fig. 6D and E). However, inactivation of vfr alone cannot fully account for the rapid increase in speB expression observed in MGAS10870 at stationary growth phase as the Vfr-inactivated strain had the same speB transcript level as strain MGAS10870 at the mid-exponential growth phase (Fig. 2C). Moreover, the speB transcript level in strain Δvfr at late exponential phase remained significantly less than that observed for strain MGAS10870 at the stationary phase of growth. Thus it is likely that an activation element, possibly a secreted peptide that is produced at high cell density is responsible for the activation of speB expression observed at stationary phase. Similar dynamics between inhibiting and activating peptides were observed in the regulation of PrgX function (Shi et al., 2005; Kozlowicz et al., 2006). The finding of such an activation peptide would explain the long-standing observation that elimination of transporter systems involved in the uptake of small peptides reduces SpeB production and might help explain how GAS is able to increase SpeB production during infection (Podbielski et al., 1996; Podbielski and Leonard, 1998).

In conclusion, we show that the amino-terminal secretion signal sequence of Vfr is an integral part of the system used by GAS to regulate expression of the gene encoding SpeB, a critical virulence factor. These data describe the existence of a pheromone-like regulatory circuit that controls virulence gene expression in a growth phase-dependent manner. Complete elucidation of peptide-controlled regulatory networks such as this could serve as the platform for the development of novel intervention strategies.

Experimental procedures

Bacterial strains, plasmids and growth conditions

Bacterial strains and plasmids used in this study are listed in Table 1. Strain MGAS10870 is a previously described invasive serotype M3 isolate whose genome has been fully sequenced (Beres et al., 2010). MGAS10870 is representative of serotype M3 strains that cause invasive infections and has a wild-type sequence for all major regulatory genes including RopB (Beres et al., 2010). GAS was grown routinely on Trypticase Soy agar containing 5% sheep blood (BSA; Becton Dickinson) or in Todd–Hewitt broth containing 0.2% (w/v) yeast extract (THY; DIFCO). When required, spectinomycin or chloramphenicol was added to a final concentration of 150 µg ml−1 and 5 µg ml−1 respectively. All GAS growth experiments were performed in triplicate on three separate occasions for a total of nine replicates. Overnight cultures were inoculated into fresh medium to achieve an initial OD600 of 0.03. Growth was monitored by measuring A600. The Escherichia coli strain used for protein overexpression was grown in Luria–Bertani broth (LB broth; Fisher Scientific) and when appropriate ampicillin was added to a final concentration of 40 µg ml−1.

Table 1.  Bacterial strains and plasmids used in this study.
Strain or plasmidDescriptionReference
Strains  
 MGAS10870Invasive isolate, serotype M3Beres et al. (2010)
 MGAS10870:pDCMGAS10870 with empty vector, Cm+This study
 ΔvfrMGAS10870Δvfr::aad9This study
 Δvfr:pDCMGAS10870Δvfr::aad9, empty vector, Cm+This study
 Δvfr-pDC:vfrMGAS10870Δvfr::aad9, vfr+, Cm+This study
 Δvfr-pDC:pepMGAS10870Δvfr::aad9, Vfr peptide+, Cm+This study
 MGAS10870ΔspeBMGAS10870ΔspeB::aad9This study
Plasmids  
 pSL60-1Vector containing aad9 gene encoding spectinomycin resistanceLukomski et al. (2000)
 pDC123Low-copy-number plasmid capable of replication in GAS and Escherichia coli, Cm+Li et al. (1997)
 pDC:vfrpDC123 with entire vfr gene plus promoter, Cm+This study
 pDC:peppDC123 with Vfr amino-terminal secretion signal plus promoter, Cm+This study

Creation of vfr- and speB-inactivated mutant strains

Insertion inactivation of the speB and vfr genes in wild-type strain MGAS10870 was performed by methods described previously (Lukomski et al., 2000; Kuwayama et al., 2002). Briefly, a PCR fragment containing a spectinomycin-resistance (spc) cassette with the fragment of gene to be deleted on either side was generated in a three-step PCR process. Subsequently, the plasmid with the spc gene-disruption cassette was introduced into the parent strain by electroporation and the gene was disrupted through homologous recombination. The isogenic mutant strains were selected by growth on spectinomycin-containing medium. Inactivation of the gene was confirmed by DNA sequencing. Primers used for the construction of strain 10870ΔspeB and Δvfr are listed in Table S1.

Construction of Vfr full-length and Vfr-peptide complementation plasmids

To complement isogenic mutant strain Δvfr, the coding sequence of the full-length vfr gene or the region encoding the amino-terminus secretion signal sequence (amino acid residues 1–39) were cloned into the E. coli–GAS shuttle vector pDC123 (Chaffin and Rubens, 1998): Using the primers listed in Table S1, the respective fragments were amplified by PCR from GAS genomic DNA, digested with BglII and NdeI, and ligated into digested vector pDC123. The inserts were verified by DNA sequencing and electroporated into the appropriate mutant derivatives of strain MGAS10870.

Measurement of transcript levels by quantitative RT-PCR

GAS strains were grown to the indicated OD600 and incubated with two volumes of RNAprotect (Qiagen) for 10 min at room temperature. Bacteria were harvested by centrifugation and the cell pellets were snap frozen with liquid nitrogen. RNA isolation and purification were performed using an RNeasy kit (Qiagen). Purified RNA was analysed for quality and concentration with an Agilent 2100 Bioanalyser. cDNA was synthesized from the purified RNA using Superscript III (Invitrogen) and Taqman quantitative RT-PCR was performed with an ABI 7500 Fast System (Applied Biosystems). Comparison of transcript levels was done using ΔCT method of analysis using tufA as the endogenous control gene (Virtaneva et al., 2005). The Taqman primers and probes used are listed in Table S1.

Western immunoblot analysis of SpeB in the culture supernatant

Cells were grown to the indicated A600 and harvested by centrifugation. The culture supernatant was filtered and the filtrate was concentrated twofold by speed-vac drying. Equal volumes of the samples were resolved on a 15% SDS-PAGE gel, transferred to a nitrocellulose membrane and probed with polyclonal anti-SpeB rabbit antibodies (Shelburne et al., 2010). SpeB was detected by secondary antibody conjugated with horseradish peroxidase and visualized by chemiluminescence using SuperSignal West Pico Rabbit IgG detection kit (Thermo Scientific).

Casein hydrolysis assay for SpeB protease activity

Cysteine protease activity in the culture supernatant was analysed by methods described previously (Collin and Olsén, 2000). Briefly, GAS strains were grown to indicated OD600 and harvested by centrifugation. The culture supernatant was filtered with a 0.22 µm membrane filter and the filtrate was used for the assay. Culture supernatant (200 µl) was mixed with an equal volume of activation buffer (0.1 M acetate buffer pH 5.0, 1 mM EDTA and 20 mM DTT) and incubated at 40°C for 30 min. The activated sample was incubated with 400 µl of 2% azocasein (Sigma) in activation buffer for 60 min at 40°C. The reaction was stopped by adding trichloroacetic acid (TCA) to a final concentration of 10% v/v. After 5 min incubation at room temperature, the samples were vortexed and centrifuged at 15 000 r.p.m. for 5 min. Casein hydrolysis was assessed by measuring the absorbance at 360 nm. A control sample in which TCA was added to the activated reaction before the addition of azocasein was used as blank.

Culture supernatant swapping experiments

Overnight cultures of strains Δvfr, Δvfr:pDC, Δvfr:pDC-vfr and Δvfr:pDC-pep were inoculated into fresh THY media. The complemented strains were grown to A600 ∼ 0.8, cells were pelleted by centrifugation, and culture supernatants were prepared by filtering through a 0.22 µm membrane filter. The cell pellets of strain Δvfr grown to mid-exponential phase (A600 ∼ 0.6) were suspended in the culture supernatants prepared from one of the three strains tested, Δvfr:pDC or Δvfr:pDC-vfr or Δvfr:pDC-pep and incubated at 37°C for 1 h (A600∼ 1.0). RNA analysis and Western immunoblotting were then performed as described above.

To characterize the regulatory factor present in the culture supernatant of Δvfr:pD-Cvfr, the complemented strain was grown at 37°C to late exponential phase (A600 ∼ 0.8) and the culture supernatant was collected by centrifugation and prepared as described below. To obtain the low-molecular-weight component of the culture supernatant, the samples were filtered using YM-3 (3 kDa cut-off filter) membrane and the flow-through was used for the culture supernatant swap experiments. Proteinase K-treated culture supernatants were prepared by incubating the filtered growth medium with proteinase K (0.15 mg ml−1) for 1 h at 37°C and the enzyme was subsequently removed by filtering through YM-10 (10 kDa cut-off filter) membrane. The filtered samples were used for the swapping. Swapping experiments were performed as described above.

Protein overexpression and purification

The ropB gene of strain MGAS10870 was cloned into plasmid pET-15b and protein was overexpressed in E. coli strain BL21* (DE3). Cells were grown at 37°C till the A600 reaches 0.5 and induced with 0.5 mM IPTG at 13°C overnight. Cell pellets were suspended in 50 ml of buffer A [20 mM Tris HCl pH 8.5, 100 mM NaCl, 10% glycerol and 1 mM Tris 2-carboxyethyl phosphine hydrochloride (TCEP)] supplemented with one protease inhibitor cocktail pellet and DNase I to a final concentration of 5 µg ml−1. Cells were lysed by a microfluidizer M-110L device (Microfluidics) and cell debris was removed by centrifugation at 15 000 r.p.m. for 30 min. RopB was purified by affinity chromatography using a Ni-NTA agarose column followed by size exclusion chromatography with Superdex 200G column. The protein was purified to > 95% homogeneity by Ni-affinity chromatography (Fig. 5B).

SPOT-peptide array

Cellulose-bound peptides were prepared by automated SPOT synthesis (MultiPep RS, Intavis, Bergisch Gladbach, Germany) as described previously (Frank and Overwin, 1996; Kramer et al., 1999). The SPOT membranes were rehydrated in Tris-buffered saline containing 0.05% Tween-20 (TBS-T) and blocked overnight at 4°C with 4% non-fat dry milk in TBS-T. After two washes with TBS-T, the membrane was incubated with purified hexa-histidine-tagged purified RopB at a final concentration of 2 µg ml−1 for 1 h at room temperature. RopB–peptide binding was assessed by chemiluminescence with anti-hexa-histidine tag antibodies conjugated to horseradish peroxidase (Clontech). Binding of the probing antibodies to the peptides was eliminated by control reactions with anti-hexahistidine tag antibody alone (data not shown).

Mouse infection studies

Virulence of the isogenic mutant GAS strains was assessed using two mouse models. For intraperitoneal infection, 15 female 3- to 4-week-old CD1 mice (Harlan Laboratories) were used for each GAS strain. Animals were inoculated intraperitoneally with 1 × 107 cfu and survival was monitored daily. Data were graphically displayed as a Kaplan-Meier survival curve and analysed using the log-rank test. For intramuscular infection, 10 female 3- to 4-week-old CD1 mice (Harlan Laboratories) were inoculated in the right hindlimb with 1 × 107 cfu of each strain and monitored for near mortality. Results were graphically displayed as a Kaplan-Meier survival curve and analysed using the log-rank test. For histopathology, infected hindlimbs were examined at 24 and 48 h post inoculation. Tissues from excised lesions were fixed in 10% phosphate-buffered formalin, decalcified, serially sectioned and embedded in paraffin using automated standard instruments. Haematoxylin and eosin and Gram's stained sections were examined in a blinded fashion with a BX5 microscope and photographed using a DP70 camera (Olympus). Micrographs of tissue taken from the inoculation site that showed pathology characteristic of each strain were selected for publication. All animal protocols were approved by the Institutional Animal Care and Use Committee at The Methodist Hospital Research Institute.

Transcript analysis from infected tissue

To assay in vivo transcript levels, skin lesions from three mice per infecting strain were collected 24 h post infection and the tissue samples were incubated with RNAlater (Qiagen). Samples were snap frozen with liquid nitrogen and stored at −80°C until use. RNA was isolated and purified using an RNeasy fibrous tissue mini kit (Qiagen). The quality and concentration of RNA were assessed with an Agilent 2100 Bioanalyser. cDNAs were prepared using Superscript III (Invitrogen) and transcript levels were measured by Taqman qRT-PCR. Data were analysed using the ΔCT method.

Acknowledgements

We thank Gary Dunny for critical reading of the manuscript. This work was supported by an American Heart Association Grant 09GRNT2280109 (S.A.S.) and a National Institute Allergy and Infectious Diseases K08 Career Development Award AI-064564 (S.A.S.).

Ancillary