Carbon Catabolite Repression on the Rgg2/3 Quorum Sensing System in Streptococcus pyogenes is Mediated by PTS^Man and Mga

Summary

Streptococcus pyogenes, also known as Group A Streptococcus or GAS, is a human-restricted pathogen causing a diverse array of infections. The ability to adapt to different niches requires GAS to adjust gene expression in response to environmental cues. We previously identified the abundance of biometals and carbohydrates led to natural induction of the Rgg2/3 cell-cell communication system (quorum sensing, QS). Here we determined the mechanism by which the Rgg2/3 QS system is stimulated exclusively by mannose and repressed by glucose, a phenomenon known as carbon catabolite repression (CCR). Instead of CcpA, the primary mediator of CCR in Gram-positive bacteria, CCR of Rgg2/3 requires the PRD-containing transcriptional regulator Mga. Deletion of Mga led to carbohydrate-independent activation of Rgg2/3 by down-regulating rgg3, the QS repressor. Through phosphoablative and phosphomimetic substitutions within Mga PRDs, we demonstrated that selective phosphorylation of PRD1 conferred repression of the Rgg2/3 system. Moreover, given the carbohydrate specificity mediating Mga-dependent governance over Rgg2/3, we tested mannose-specific PTS components and found the EIIA/B subunit ManL was required for Mga-dependent repression. These findings provide newfound connections between PTS^Man, Mga, and QS, and further demonstrate that Mga is a central regulatory nexus for integrating nutritional status and virulence.

Abbreviated Summary

Here we describe the mechanism by which a quorum sensing system in Streptococcus pyogenes is exclusively stimulated by mannose. Selective phosphorylation of the virulence regulator Mga by the PTS^man system governs the expression of quorum sensing regulator Rgg3. Our study provides newfound connections between nutritional status, quorum sensing, and virulence regulation.

Graphical Abstract

Introduction

Carbon catabolism is one of several essential biochemical processes that sustain all life forms. For these biochemical reactions to proceed, bacteria first need to import extracellular carbohydrates to initiate catabolism. The Phosphoenolpyruvate-dependent Transferase System (PTS) is a means for bacteria to monitor and regulate the import of carbohydrates through dedicated membrane transporters with subsequent phosphorylation of sugars by sugar-specific protein kinases (Deutscher et al., 2006). The PTS is comprised of Enyzme I (EI) and HPr, both of which are cytosolic proteins that initiate a phosphorelay to carbohydrate specific Enzyme II (EII) complexes comprised of EIIA, EIIB, EIIC, and sometimes EIID. These EII complexes can be grouped into four superfamilies (Saier Jr, 2015), of which the mannose family (PTS^Man) has garnered special interests due to its structural and biological peculiarities (reviewed in Jeckelmann & Erni, 2020). Briefly, in conversion of phosphoenolpyruvate (PEP) to pyruvate, a phosphate group is transferred to EI and subsequently to HPr, the phosphor-carrier protein, on the histidine-15 residue. Phosphorylated HPr (HPr-His15~P) transfers the phosphate to an EIIA subunit and further transfer to EIIB. EIIB catalyzes transfer of the phosphate to the incoming sugar brought into the cell by the EIIC/D membrane components (Deutscher et al., 2006).

Besides participating in the uptake of carbohydrates, the phosphor-carrier protein HPr is involved in regulation of genes that encodes catabolic enzymes and carbohydrate transports, including PTS systems. This is achieved in conjunction with Carbon Catabolite Protein A (CcpA), primarily a transcriptional repressor but can also function as a transcriptional activator of a small subset of genes within its regulon (Henkins, 1996, DebRoy et al., 2016, DebRoy et al., 2021). This regulatory network is unique to Gram-positive Firmicutes, which requires phosphorylation of HPr on a serine residue, catalyzed by a kinase (HprK) with ATP as a phospho-donor (reviewed in Deutscher et al., 2005). The phosphorylation of the serine residue (HPr-Ser46-P) allows it to be a cofactor to CcpA, which increases its DNA-binding activity for binding to cre (Görke & Stülke, 2008) or cre2 (Willenborg et al., 2014) elements to control gene transcription (Figure 1) (Henkins, 1996, DebRoy et al., 2016, DebRoy et al., 2021). The regulon of CcpA is species specific but typically includes genes encoding enzymes and transporters, such as the PTS system, for the import and catabolism of specific and less preferred carbohydrates. In the presence of a preferred sugar, such as glucose, these genes are repressed by CcpA, while in the absence of glucose but in the presence of a less preferred sugar, genes encoding the transporter and enzymes for the abundant sugar are derepressed. This regulatory phenomenon is also known as Carbon Catabolite Repression (CCR) (Görke & Stülke, 2008).

Figure 1: CcpA-dependent and CcpA-independent Carbon Catabolite Repression (CCR).

Left: During growth in the presence of a preferred carbon source (e.g. glucose), HPrK will phosphorylate HPr at the serine residue (HPr-Ser46-P), which increases its binding affinity to CcpA. The protein complex will proceed to bind DNA to initiate repression on target genes that encode gene products facilitating transport and utilization of less preferred carbon sources.

Right: During growth in the absence of glucose but in the presence of a PTS-dependent sugar, HPr will be phosphorylated at the histidine-15 residue through EI enzyme, with phosphoenolpyruvate (PEP) as a phosphate donor. It will act as a phosphoryl donor to EIIA to prime the cells for the immediate uptake of any available carbohydrates. In some cases, these PTS transport systems are regulated by a transcriptional regulator that harbor PTS Regulatory Domains (PRD). These PRD-containing regulators can be phosphorylated by cognate EIIB, or by HPr-His15~P itself, to modulate expression of catabolic genes.

Although CcpA and HPr-Ser46-P are the primary mediators for CCR, some catabolic operons are regulated in a CcpA-independent manner. In these cases, HPr-His15~P serves as the central processing unit by influencing the activity of dedicated transcriptional activators that modulate the expression of less-preferred catabolic systems (Figure 1). These regulators typically harbor a PTS Regulatory Domain (PRD), which is subject to phosphorylation on conserved histidine residues and consequently influences their DNA-binding activity (Deutscher et al., 2014). For example, in Bacillus subtilis, MtlR activates the transcription of the mtl operon, which encodes EIIABC^Mtl components for transport and a mannitol-6-phosphate dehydrogenase for the catabolism of mannitol through glycolysis. MtlR contains two PRDs (PRD1 and PRD2), followed by an EIIB^Gat-like domain and an EIIA^Mtl-like domain, all of which partake in regulating its activity. Phosphorylation of PRD2 by EI/HPr and protein-protein interactions with the EIIA^Mtl via the MtlR-EIIA- and -EIIB-like domains have a stimulatory effect, whereas phosphorylation of the EIIB-like domain has an inhibitory effect (Henstra et al., 2000, Joyet et al., 2010, Bouraoui et al., 2013). Similarly, in Listeria monocytogenes, ManR is a transcriptional regulator that induces the expression of PTS^Man during growth in mannose or glucose as the carbohydrate, and has two PRDs flanking an EIIA^Man-like domain and EIIB^Gat-like domain. Instead of interaction with its cognate EIIA/B^Man, the activity of ManR is influenced by EIIB^mpo, a non-homologous PTS that belongs to the same family of PTS^Man, which serves as a carbohydrate sensor (Aké et al., 2011). For ManR, phosphorylation of the EIIA^Man-like domain by EI/HPr, in parallel with protein-protein interactions of the EIIB^Gat-like domain and PRD2 with EIIB^mpo, have stimulatory effects. Phosphorylation of PRD2 by EIIB^mpo abolishes the protein-protein interaction with EIIB^mpo, and confers an inhibitory effect on ManR (Zébré et al., 2015). Further examples are reviewed by Deutscher and colleagues (Deutscher et al., 2014). These studies highlight the complex interplay between transcriptional regulators and PTS components to mediate CCR. Furthermore, intricacies of regulation exemplify a lack of consensus as to which domains have stimulatory or inhibitory effects upon phosphorylation and/or protein-protein interactions. These studies also underscore the multi-faceted functions of PTS components that involve both regulatory and carbohydrate transport roles.

The human host is the only known reservoir of Streptococcus pyogenes, also known as Group A Streptococcus (GAS). GAS colonizes the upper respiratory tract or skin with subclinical consequence, but frequently cause mild infections such as strep throat or impetigo. It can also cause severe, invasive infections such as septicemia and necrotizing fasciitis (Cunningham, 2000). For GAS to manifest such a wide array of clinical pathologies throughout the host, the bacteria need to adjust gene expression to survive and adapt to the various niches during infection. As GAS only possess one alternative sigma factor that modulates genes associated with competence, it relies on various stand-alone regulators, two-component signaling systems and quorum sensing systems to regulate gene expression in response to environmental stimuli (Kreikemeyer et al., 2003).

One such stand-alone regulator, conserved in all GAS strains, is Mga, which controls expression of a core regulon that includes virulence factors, such as M protein and C5 peptidase, as well as serotype-specific accessory genes (Ribardo & McIver, 2006, Hondorp & McIver, 2007). The Mga protein possesses a modular structure; at its N-terminus is a helix-turn-helix (HTH) DNA-binding domain, followed by two PRDs (PRD1 and PRD2) and a C-terminal EIIB^Gat-like domain (Hondorp et al., 2012, Hondorp et al., 2013). Each of these domains provide a regulatory function: EI/HPr-mediated phosphorylation of two histidine residues in PRD1 inactivates Mga, while phosphorylation of a sole histidine in PRD2 amplifies its transcriptional activity (Hondorp et al., 2013). The EIIB^Gat-like domain is critical for the oligomerization of Mga (Hondorp et al., 2012). As the EI/HPr phosphorelay is active during growth in the presence of less-preferred carbohydrates, it facilitates Mga to adjust the expression of its regulon based on the metabolic status of the cell, providing a link between metabolism and virulence.

GAS also utilizes cell-cell signaling systems, also referred to as quorum sensing (QS), to coordinate gene expression. QS is mediated through peptide pheromones that bind to cytosolic transcriptional regulators of the RRNPP family that are widely present in Firmicutes (Neiditch et al., 2017). Four RRNPP regulators are conserved in GAS (RopB, Rgg2, Rgg3, and ComR) and prior studies have demonstrated the Rgg2/Rgg3 system is regulated by multifaceted means that include availability of pheromones, metals, and carbohydrates (Chang et al., 2011, LaSarre et al., 2012, Chang et al., 2015, Wilkening et al., 2016). Rgg2 and Rgg3 work together, in response to their cognate pheromones, SHP2 and SHP3, to control gene activities related to lysozyme resistance, biofilm formation, and immune suppression (Chang et al., 2015, Gogos et al., 2018, Rahbari et al., 2021). Activity of this system is silent in peptide-free chemically defined media (CDM) containing glucose, but becomes activated if mannose is the primary carbon source, demonstrating it is subject to CCR and is induced by a specific PTS sugar (Chang et al., 2015).

In this study, we elucidate the regulatory elements and pathways that confer CCR on the Rgg2/3 QS system in GAS. We demonstrate that CCR on Rgg2/3 occurs in a CcpA-independent manner and instead requires Mga. Through genetic manipulation, we identify the phosphorylated form of Mga that represses the Rgg2/3 QS system during growth in different carbohydrates. We also report that the carbohydrate-dependent repression of the Rgg2/3 QS system is mediated by a mannose PTS system, likely also through Mga. These findings provide new insights in understanding how Mga is a critical nexus between metabolism and virulence during pathogenesis of GAS.

Materials and Methods

Bacterial strains and growth condition

Streptococcus pyogenes NZ131 (M49 Serotype) was routinely grown in Todd Hewitt Broth (BD) supplemented with 0.2% yeast extract (Amresco) (THY) at 37°C. When appropriate, antibiotics were added at the following concentration: erythromycin (Erm; 0.5 μg ml⁻¹). Chemically Defined Medium (CDM) was prepared as previously described (Chang et al., 2011) and glucose (CDM-Glu) was added in to a final concentration of 1% v/v. For mannose (CDM-Man) or sucrose (CDM-Suc), either carbohydrate was added to a final concentration of 1% v/v, in addition to glucose at a final concentration of 0.02% v/v, unless otherwise stated, to support robust growth. Escherichia coli cloning strain BH10C (Howell-Adams & Seifert, 2000) was routinely maintained in Luria broth (LB) or on Luria agar and if necessary, supplemented with erythromycin (500 μg ml⁻¹).

Synthetic peptides

Synthetic peptides as described previously (Chang et al., 2011) were purchased from Neo-peptide (Cambridge, MA). The peptides were resuspended in DMSO at a concentration of 1 mM and stored at −80°C in aliquots. Working stock of peptides (100 μM) were diluted in DMSO and stored at −20°C.

Generation of ΔccpA mutant

To generate markerless ccpA mutant, upstream fragment of ccpA was amplified with JJ289 and JJ292 while downstream fragment of ccpA was amplified with JJ290 and JJ291. Amplicons were purified and ligated to generate a deletion fragment, which was subsequently cloned into pJC162 to generate pJC162-ccpA. The construct was transformed into competent GAS through electroporation and plated onto THY-Erm plates. Plates were incubated at 30°C and positive clones were verified by PCR using JWP0007 and JWP0008. Upon positive verification for the plasmid, the clones were grown in THY-Erm broth at 37°C overnight and plated onto THY-Erm plates. Verification of plasmid integration within the genome of GAS at the ccpA locus was performed using JWP0011 and JWP0008, and JWP0012 and JWP0007. Positive integrants were stored as glycerol stocks at −80°C. A single colony of the integrant was cultured in THY overnight, and subsequently serially passaged in CDM supplemented with 0.02% glucose and 1% mannose for three days at 30°C. The passaged culture was serially diluted and plated onto THY plates. Putative isolates were patched onto both THY-Erm plates and THY and isolates that were Erm sensitive were verified for the loss of ccpA using JWP0011 and JWP0012. Positive mutants were stored as glycerol stocks at −80°C.

Generation of Δmga, ΔptsA, ΔptsB and ΔptsC mutant

To generate markerless mga mutant, the coding sequence of mga, together with 500 bp upstream from the translational start codon, were amplified with JWP0047 and JWP0048. The amplicon was cloned into pJC162 to generate pJW107. Using pJW107 as a template, inverse PCR was performed with 5’-phosphorylated primers JWP0071 and JWP0072 and subsequently ligated to generate pJW108. For ptsA mutant, the upstream fragment was amplified with JWP0037 and JWP0038, while downstream fragment was amplified with JWP0039 and JWP0040. For ptsB mutant, the upstream fragment was amplified with JWP0037 and JWP0041 while downstream fragment was amplified with JWP0042 and JWP0043. Single overlapping PCR was performed to fuse the upstream and downstream fragments, and subsequently, the amplicon was cloned into pJC162 to generate pJW102 for ptsA and pJW103 for ptsB. For ptsC mutant, the coding sequence of ptsABCD operon was amplified with JJ273 and JJ277 and cloned into pJC162 to generate pJC162-ptsABCD. Subsequently, inverse PCR was performed with JJ278 and JJ279 to generate pJC162-ptsC. The resulting plasmids were transformed into competent GAS cells for integration and passaged in THY at 30°C as described in above section. The loss of target genes was verified; JWP0029 and JWP0030 for mga, JWP0035 and JWP0036 for both ptsA and ptsB, and JWP0035 and JJ277 for ptsC. All positive mutants were stored as glycerol stocks at −80°C.

Generation of manL mutant

To generate markerless manL mutant, the upstream fragment of manL was amplified with JWP0094 and JWP0095 and downstream fragment was amplified with JWP0096 and JWP0097. Subsequently, single-overlapping extension PCR was performed with both upstream and downstream fragment to generate a deletion fragment, which was cloned into pJC162 to generate pJW128. The resulting plasmid was transformed into competent GAS cells for integration and passaged in CDM-Suc at 30°C as described in above section. The loss of manL was verified with PCR using JWP0092 and JWP0093. All positive mutants were stored as glycerol stocks at −80°C.

Generation of mga point mutation complements

All constructs (pJW138 – pJW145) were made through inverse PCR using pJW107 as template and targeted mutations for the coding sequence of the histidine residue within the primer listed in Table S3. After PCR, the constructs were ligated and sequenced to ensure the correct mutations were obtained. The constructs were transformed into competent Δmga mutants for integration and passaged in THY at 30°C as described in above section. The restoration of mga allele was confirmed with PCR using JWP0029 and JWP0030. All positive variants were stored as glycerol stocks at −80°C.

Luciferase assay

Strains of interests were grown overnight in THY broth at 37°C. The next morning, cultures were diluted 1:10 in fresh THY broth and incubated at 37°C until mid-exponential phase (OD₆₀₀ = 0.2 – 0.4). Appropriate volumes of the cultures were washed twice with phosphate-buffered saline (PBS) and resuspended in 200 μl of PBS and inoculated into fresh CDM at a starting OD₆₀₀ of 0.05. Cultures were incubated at 37°C without shaking and the OD600 was monitored hourly using Genesys30 (Thermo Scientific). At the corresponding time-point, 50 μl of culture was transferred to a 96-well white opaque plate (Greiner Bio-one), exposed to decyl aldehyde fumes for 30 s and counts per second (CPS) were quantified using a Veritas microplate luminometer (Turner Biosystems). Relative Light units (RLUs) were calculated by normalizing CPS to OD₆₀₀. Each experiment was performed at least thrice on independent days.

Results

Carbon catabolite repression of Rgg2/3 quorum sensing system is not mediated by CcpA

We previously reported that mannose is a unique carbohydrate signal for the induction of Rgg2/3 quorum sensing (QS) system in Streptococcus pyogenes NZ131 (GAS). To monitor for Rgg2/3 activity, we employed a luciferase reporter driven by the promoter of shp3 (P_shp3-luxAB), which encodes the pheromone of the QS system (Chang et al., 2011). Activity of the QS system, determined by the transcriptional status of shp promoters (P_shp), was also subject to carbon catabolite repression (CCR) when bacterial cultures were titrated with glucose even in the presence of excess mannose (Chang et al., 2015) (Figure 2A). In low G+C Gram-positive bacteria, including GAS, CCR is often mediated by the transcription regulator CcpA (Deutscher et al., 2006 and Figure 1), which predominantly serves as a transcriptional repressor of target genes (DebRoy et al., 2016, Willenborg et al., 2014, Carvalho et al., 2011). To investigate if CcpA mediates CCR of the Rgg2/3 QS system, we generated a ΔccpA mutant and determined if the deletion resulted in QS induction during growth in a chemically defined medium supplemented with 1% glucose (CDM-Glu). We anticipated that if CcpA were the mediator of CCR, then deletion of ccpA would result in induction of Rgg2/3 QS system during growth in CDM-Glu.

Figure 2: Carbon catabolite repression on Rgg2/3 quorum sensing induction is mediated by a CcpA-independent mechanism.

Induction of the Rgg2/3 QS circuitry is monitored by the expression of luciferase genes driven by the promoter of the pheromone encoding gene, shp3. Relative Light Units (RLU) were obtained by normalizing the abundance of luminescence units per second to the OD₆₀₀ of the corresponding sample and plotted on the Y axis on a logarithmic scale while the X axis is the OD₆₀₀ value of the sample. A) The Rgg2/3 quorum sensing circuit is induced when cells were grown in CDM supplemented with 1% mannose, but is subjected to CCR with increasing concentrations of glucose provided together with mannose. B) Repression of the Rgg2/3 quorum sensing circuitry when cells are grown in the presence of glucose is independent of CcpA. C) The ΔccpA mutant harbors an intact Rgg2/3 QS circuit.

Surprisingly, deletion of ccpA did not cause induction of the Rgg2/3 QS system during growth in CDM-Glu, as the expression of shp3 remained on a similar level to that of WT as well as the Δrgg2 strain, which lacks the transcriptional activator of the QS circuitry (Figure 2B). When synthetic pheromone (SHP3) was added exogenously to cultures, the ΔccpA mutant induced P_shp3, indicating the QS sensory system remained intact (Figure 2C). Additionally, in silico analysis of both rgg2 and rgg3 promoter regions determined that cre (TGWAARCGYTWNCW) (Görke & Stülke, 2008) or cre2 (TTTTYHWDHHWWTTTY) (Willenborg et al., 2014) elements are absent, further supporting the conclusion that CCR on the Rgg2/3 QS system is not mediated by CcpA.

Carbon catabolite repression of Rgg2/3 quorum sensing system is mediated by Mga

Carbon catabolite repression can also be achieved in a CcpA-independent manner, one that involves the PTS components of EI and HPr phosphorylating a catabolic enzyme or a transcriptional regulator (Deutscher et al., 2014). Prior studies determined that an HPr deletion is not tolerated by GAS (Le Breton et al., 2015, DebRoy et al., 2021) and our attempts to generate an EI (ptsI) mutant were also unsuccessful. We therefore focused on the end point of this regulatory network, transcriptional regulators subject to PTS modulation that contain phosphorylatable histidine residues within PTS Regulatory Domains (PRD). In GAS, Mga is the sole transcriptional regulator that harboring a PRD, and it has been shown that Mga’s activity is impacted by phosphorylation from EI/HPr (Hondorp et al., 2013). The Mga regulon has yet to be defined in NZ131 but is well characterized in other serotypes (Ribardo & McIver, 2006, Sanson et al., 2015, Valdes et al., 2018). Interestingly, transcriptomic data indicates that Mga is a repressor of rgg3 in both M1 and M6 serotypes, while in an M4 and M1T1 serotypes, Mga had no effect on rgg3 expression, indicating the differential regulation of Rgg2/3 QS system is serotype specific (Ribardo & McIver, 2006). Inferring from this study, we hypothesized that Mga could be regulating the Rgg2/3 QS system in a carbohydrate-specific manner in NZ131 by modulating the transcription of rgg3.

To address this hypothesis, we generated a Δmga mutant and determined if the Rgg2/3 QS system was induced during growth in CDM-Glu. Integration of the P_shp3-luxAB reporter into the Δmga mutant enabled the observation that induction of the QS system in CDM-Glu conditions occurred like that of the positive control, Δrgg3, which is devoid of the repressor for the QS system (Figure 3A). Similarly, when Δmga was grown in CDM supplemented with 1% mannose and 0.02% glucose (CDM-Man), QS induction was observed like that of WT. Moreover, when cells were grown in CDM supplemented with 1% sucrose and 0.02% glucose (CDM-Suc), a carbohydrate that does not induce the QS system in WT, induction did occur in the Δmga mutant (Figure 3B). This suggested that Mga represses the Rgg2/3 QS system in a carbohydrate-specific manner.

Figure 3: Mga mediates carbon catabolite repression of Rgg2/3 quorum sensing circuit by regulating rgg3 expression.

A) Deletion of mga, a PRD-containing transcriptional regulator, results in activation of the Rgg2/3 circuit, B) regardless of carbon source, indicating that Mga mediates CCR on Rgg2/3 quorum sensing circuitry. C) Mga promotes transcription of rgg3, which is a repressor of the quorum sensing circuit. Addition of SHP3 pheromone or mannose reduces expression of rgg3 in WT.

As Mga is primarily a transcriptional activator, we speculated whether it incurs CCR on Rgg2/3 QS directly by activating expression of rgg3, a repressor of the Rgg2/3 QS system. To validate this hypothesis, we integrated a luciferase reporter construct driven by the rgg3 promoter (P_rgg3-luxAB) into both WT and Δmga strains and monitored luciferase expression during growth in CDM-Glu (Figure 3C). We found that the Δmga strain exhibited a 10-fold reduction of rgg3 expression compared to WT when grown in CDM-Glu. Diminished expression of rgg3 was also seen in WT cultures when QS was induced either by exogenous addition of synthetic SHP3 pheromone or by growth in CDM-Man. This suggested that down-regulation of rgg3 is a contributing factor in the natural activation of the QS system. However, the level to which rgg3 expression was reduced in WT under SHP3 or mannose conditions did not reach levels seen in the Δmga strain, which were 2- to 3-fold lower. Collectively, the results strongly suggest that Mga regulates the Rgg2/3 QS circuitry directly or indirectly by modulating the expression of rgg3.

Mga represses Rgg2/3 quorum sensing system through phosphorylation of PRD1

Mga, together with AtxA of Bacillus anthracis (Tsvetanova et al., 2007), MgaSpn of Streptococcus pneumoniae (Solano-Collado et al., 2016) and MafR of Enterococcus faecalis (Ruiz-Cruz et al., 2016), belongs to a unique class of transcriptional regulators known as PRD Containing Virulence Regulators (PCVR) (Hondorp et al., 2013). Modulation of PCVR activity occurs by phosphorylation of histidine residues by PTS components EI and HPr (Hondorp et al., 2013, Tsvetanova et al., 2007). In Mga, three histidine residues have been identified as phosphorylated by EI/HPr; two in PRD1 (H204 and H270) and one in PRD2 (H324) (Hondorp et al., 2013) (Figure 4A). It was revealed that in the M1 serotype, phosphorylation of the histidine residues in PRD1 inactivates Mga, although unphosphorylated Mga appeared to have diminished activity as well, albeit to a lower magnitude compared to the phosphorylated form (Hondorp et al., 2013). This suggests that Mga activity is highly dynamic in nature, and multiple phosphorylatable forms of Mga potentially contribute towards the activation and repression of its regulon in response to carbohydrate availability. To determine which phosphorylated form of Mga represses the Rgg2/3 QS system in NZ131, we exchanged the Δmga deletion with mga alleles that harbor point mutations. These allelically exchanged variants were designed to produce either phosphoablative (histidine to alanine) or phosphomimetic (histidine to aspartic acid) substitutions in various combinations (Figure 4A). These strains all contain the mga allele in single copy, expressed at the native locus, rather than utilizing a multi-copy plasmid as a complementation strategy as it was previously reported that multi-copy complementation could provide experimental artifacts (Hondorp et al., 2013). Each variant’s ability to repress the Rgg2/3 QS system using the P_shp3-luxAB reporter was assessed.

Figure 4: Combinatorial phosphorylation or unphosphorylation of PRD sites in Mga varied the suppression of Rgg2/3 circuitry.

A) A schematic depicting the location of phosphorylatable histidine residues in Mga and the nomenclature of phosphoablative and phosphomimetic variants in subsequent studies. B) Induction of Rgg2/3 quorum sensing circuit (P_shp3-luxAB) when complemented with phosphoablative variations of the mga allele grown in glucose, or C) mannose as the sole carbon source. D) Induction of Rgg2/3 quorum sensing circuit when complemented with phosphomimetic variations of mga allele grown in glucose, or E) mannose as the sole carbon source.

When phosphoablative variants were cultured in CDM-Glu (Figure 4B), we observed that variants with single substitutions in PRD1, Mga^HAA and Mga^AHA, caused the QS system to be repressed. However, combined phosphoablative substitution on both PRD1 residues, Mga^AAH, abrogated the repression of the QS system. Similarly, when all three histidine residues were substituted with alanine (Mga^AAA), the variant failed to repress the QS system. This indicates that when GAS is grown in CDM-Glu, Mga is phosphorylated on either PRD1 histidine residue to mediate repression. When these variants were grown in CDM-Man, all showed WT levels of induction except Mga^AHA, which showed reduced induction of the QS system (Figure 4C). These results suggest that Mga-dependent repression of the Rgg2/3 QS system during growth in glucose is dependent on the phosphorylated state of PRD1, but not PRD2.

When phosphomimetic variants were grown in CDM-Glu (Figure 4D), we observed that substitutions in both histidine residues of PRD1 (Mga^DDH) resulted in the derepression of the Rgg2/3 QS system. This corroborates published observations in the M1 serotype that phosphorylation of PRD1 histidine residues inactivate Mga (Hondorp et al., 2013). Interestingly, it was noticed that variants possessing phosphomimetic substitutions in PRD2 (Mga^HDD, Mga^DHD and Mga^DDD) all exhibited a growth defect in CDM-Glu and CDM-Man in comparison to WT. This indicates that phosphorylation of PRD2 negatively impacts the cellular fitness of NZ131. None of these PRD2 variants demonstrated induction of Rgg2/3 system in either CDM-Glu or CDM-Man but induction of QS was observed in the presence of synthetic SHP3 pheromone (data not shown), indicating that these variants harbor an intact QS circuitry for import and sensing of the pheromone. However, the Mga^DHD variant displayed a level of QS activity slightly above the basal level of Δrgg2 during growth in CDM-Man (Figure 4E), suggesting that phosphorylation of the first histidine residue of PRD1 prevented the induction of Rgg2/3 QS system when cells are grown in mannose. The results also lend confidence that phosphorylation of either histidine of PRD1 facilitate the Mga-dependent repression of Rgg2/3 QS system.

Carbon catabolite repression of Rgg2/3 quorum sensing system is dependent on ManL

The activity of PRD containing transcriptional regulators are often modulated through phosphorylation by the general PTS components EI and HPr, but also by sugar-specific EIIA and/or EIIB proteins (Martin-Verstraete et al., 1998, Henstra et al., 2000, Tsvetanova et al., 2007, Hondorp et al., 2013). A previous study showed that Mga is subject to phosphorylation by EI and HPr and demonstrated the link between carbon metabolism and virulence factor synthesis. However, if repression of the Rgg2/3 QS system only requires Mga to be phosphorylated at PRD1 by EI/HPr during growth in glucose, why is the Rgg2/3 QS system induced in the presence of mannose but not by sucrose, given that EI and HPr are essential for PTS-mediated transport of both carbohydrates? This discrepancy, that different PTS-dependent carbohydrates impact the regulation of the QS system distinctly raises the possibility that an intermediary component affects the activity of Mga.

Several studies have demonstrated that PRD-containing transcriptional regulators could be modulated by phosphorylation and/or through protein-protein interactions with EIIA or EIIB components in multiple bacterial species (Henstra et al., 2000, Bouraoui et al., 2013, Joyet et al., 2013, Zébré et al., 2015). Given these precedents, we hypothesized that the EIIA or EIIB component of a PTS system could modulate Mga activity. Since the induction of Rgg2/3 QS system is only observed in mannose, we focused our attention on the mannose PTS system. In GAS, two non-homologous PTS^Man transporters are present within the genome; manLMN and ptsABCD (Sundar et al., 2017). For the former, EIIA and EIIB components are fused as a single protein encoded by manL, while for the latter, EIIA and EIIB are encoded by ptsA and ptsB, respectively. We generated ΔmanL, ΔptsA, and ΔptsB mutants to determine if the absence of these EII components would result in the activation of Rgg2/3 QS when grown in non-permissive condition. We previously observed that a ΔptsC mutant, lacking a transmembrane component of the translocation complex, was unable to induce QS during growth in mannose (Chang et al., 2015); thus, we anticipated that the ΔptsA and ΔptsB mutant would phenocopy that of ΔptsC.

When grown in CDM-Glu (Figure 5A), none of the mutants showed induction of Rgg2/3 QS, indicating that none were involved in the CCR of the QS system when glucose is the primary carbohydrate. It is likely that during growth in CDM-Glu, EI/HPr is the primary phosphor-donor of Mga. When grown in CDM-Man (Figure 5B), we observed that both the ΔmanL and ΔmanM mutants showed poor growth in this medium, confirming that the ManLMN PTS system is a mannose transporter. In contrast, both the ΔptsA and ΔptsB mutants showed WT levels of growth, corroborating our previous observation that PtsABCD is not the primary transporter for mannose (Chang et al., 2015). However, we observed that both ΔptsA and ΔptsB mutants, as well as the ΔptsC mutant, showed induction of Rgg2/3 QS system in this condition. In addition, we also generated a marker-less mutant of ptsC for further validation of the previous result and yet we observed that the markerless ΔptsC mutant showed induction as well (data not shown). As induction was observed in these mutants, we conclude that PtsABCD is not required for the induction of Rgg2/3 QS system in the presence of mannose.

Figure 5: ManL facilitates CcpA-independent CCR of Rgg2/3 quorum sensing circuit.

A) Induction of the Rgg2/3 quorum sensing circuit in PTS^Man EIIA and EIIB mutants when grown in glucose, B) mannose, or C) sucrose as the sole carbon sources. D) Induction of the Rgg2/3 quorum sensing circuit in ΔmanL grown in CDM Suc supplemented with a range of glucose concentrations.

As the ΔmanL and ΔmanM mutants showed poor growth in CDM-Man, thus preventing us to determine its involvement in mediating CCR in the presence of a non-glucose, PTS-dependent substrate, as well as the confounding results of the ΔptsA, ΔptsB and ΔptsC mutants, we assessed their QS induction phenotype in CDM-Suc. We chose sucrose as our carbohydrate as it was previously shown that with the exception of sucrose PTS, the other PTS systems are highly redundant in the transport of non-cognate carbohydrates (Sundar et al., 2017). Moreover, we have established that the Rgg2/3 QS system is not induced in the presence of sucrose (Figure 3B). We expected that if either the ManLMN or the PtsABCD system is involved in mediating CCR, the corresponding mutant of that system will exhibit induction of Rgg2/3 QS system, as EI/HPr will be actively involved in transporting sucrose for utilization, instead of modulating Mga activity.

When grown in CDM-Suc (Figure 5C), none of the PTS mutants exhibited any pronounced growth defect, corroborating with the previous observation that there is minimal redundancy among the PTS systems in GAS with regards to sucrose transport (Sundar et al., 2017). In this medium, we observed that both ΔptsA and ΔptsB mutants behaved like WT and Δrgg2, where the Rgg2/3 QS system remain repressed. On the contrary, we observed the induction of QS in ΔmanL but not in the ΔmanM mutant, providing evidence that ManL mediates carbohydrate-specific CCR on the Rgg2/3 QS system, plausibly through phosphorylation of Mga. To further validate this statement, we repeated the experiment with cells grown in sucrose with a range of glucose concentrations. We anticipated that QS would be induced in ΔmanL in the presence of sucrose due to the lack of phosphorylation of Mga. However, the presence of glucose would phosphorylate Mga by way of EI/HPr phosphorylation. This expectation was supported by the result in Figure 5A, where QS remained repressed in the ΔmanL mutant in the presence of glucose. Indeed, we observed an inverse correlation between glucose concentration and the rate of QS induction in ΔmanL mutant in sucrose (Figure 5D), supporting our conclusion that ManL is a mediator for carbohydrate-specific CCR.

A proposed model depicting how ManL and Mga converge to modulate the induction of quorum sensing

The collective data presented here allowed us to establish a working model on how CCR on Rgg2/3 QS is achieved through both Mga and ManL when grown in the presence of glucose, mannose or sucrose (Figure 6). In the presence of glucose, translocation of glucose is independent of PTS, and instead, relies on a glucose permease, GlcU (spy49–1811c). Though not tested in GAS, the function of a GlcU homolog has been verified in Lactococcus lactis (Castro et al., 2009). Translocated glucose is then phosphorylated by putative glucokinases (Spy49–1181c and/or Spy49–0217) and glycolysis initiated for the generation of phosphoenolpyruvate (PEP) and subsequent energy for cellular proliferation. At this stage of growth, the metabolic state of the cell is reflected by the PEP:pyruvate ratio, where metabolically active cells have a low ratio (Hogema et al., 1998). This, in turn, will result in HPr-(Ser46-P) as the major phosphorylated form to serve as a partner protein for CcpA to mediate CCR in a CcpA-dependent manner. HPr-His15~P (or potentially HPr-(Ser46-P)-His15~P) will be available as a minor population and the phosphate will be transferred to Mga on the first histidine residue in PRD1 to mediate CcpA-independent CCR on the Rgg2/3 QS system by enhanced transcription of rgg3, encoding the repressor of the QS system.

Figure 6: Model depicting the interaction between PTS and Mga during growth in different carbon sources that dictate the induction or suppression of Rgg2/3 quorum sensing circuit.

Grey dashed arrows indicate the pathway procession of the metabolic intermediate. Black solid arrows indicate the flow of phosphate.

Left: In glucose, glycolysis results in an increase of PEP. As glucose uptake is independent of PTS, EI and Hpr phosphorylate Mga PRD1 at H204. Phosphorylated Mga elevates rgg3 transcription, which suppress induction of the QS circuitry.

Center: When mannose is the primary carbon source, EI and Hpr will phosphorylate ManL for the uptake of mannose. This results in an unphosphorylated population of Mga, which in turn reduces the expression of rgg3, and consequently, induces the QS circuitry.

Right: In sucrose, EI and Hpr will phosphorylate PTS^Suc components for the uptake of sucrose. Metabolism of sucrose will feed into glycolysis to increase PEP levels, providing phosphate return through EI and Hpr for subsequent uptake of sucrose. EI and Hpr will phosphorylate ManL; however, as mannose is absent, the phosphate is transferred to Mga PRD1 at H270. Phosphorylated Mga will activate transcription of rgg3 and suppress the induction of the QS circuitry.

In the presence of mannose, import of mannose is dependent on the ManLMN PTS system. Upon phosphorylation by ManL, phosphomannose isomerase (PMI; Spy49–1412c) will convert mannose-6-phosphate to fructose-6-phosphate to feed into the glycolysis pathway. As glucose is absent from the medium, cells will be less active metabolically, resulting in elevated PEP:pyruvate ratio. Consequently, PEP will phosphorylate the EI enzyme and sequentially phosphorylate HPr-His15 to facilitate further import of mannose through ManLMN for catabolism. Neither EI/HPr nor ManL will phosphorylate Mga, thus preventing Mga to initiate transcription of rgg3. The reduced transcription of rgg3 translates to a reduced level of protein, thus allowing QS to proceed during growth in mannose.

In the presence of sucrose, import of sucrose is dependent on PTS^Suc (ScrA; Spy49–1415c), with EIIA-EIIC-EIIB domains fused as a single polypeptide chain in this configuration. Upon its translocation, sucrose-6-phosphate hydrolase (ScrB; Spy49–1416) will cleave sucrose to generate glucose-6-phosphate and fructose, while fructokinase (Spy49–1413c) will phosphorylate the fructose moiety to generate fructose-6-phosphate. Both intermediates will proceed to undergo glycolysis for generation of PEP. For further import of sucrose, PEP will phosphorylate EI and subsequently HPr-His15. The resulting HPr-His15~P will also phosphorylate other EIIAs and EIIBs to prime cells for uptake of other available carbohydrates (Hogema et al., 1998, Vadeboncoeur et al., 1991). In this case, ManL is phosphorylated but owing to the absence of mannose, ManL will instead phosphorylate Mga at the second histidine residue of PRD1, resulting in the transcription of rgg3. This allows the repression of Rgg2/3 QS in the presence of sucrose.

Discussion

It is well established that environmental signals sensed by bacterial cells dictate their gene expression as a response for survival or proliferation. Among the most critical of these signals is the availability of preferred carbohydrates and the corresponding orchestration of gene expression used to optimize metabolism is known as carbon catabolite repression or activation. We previously identified mannose as an environmental cue that induces the Rgg2/3 QS system and demonstrated that the induction of QS is subjected to CCR in GAS NZ131 (Chang et al., 2015). In this study, we sought to understand the underlying molecular mechanism that regulates this phenomenon. We report that CCR of the Rgg2/3 QS system is mediated through a CcpA-independent mechanism, where the Mannose PTS system converges with Mga, a transcriptional regulator of virulence genes, to exert a tight regulatory control in a carbohydrate-specific manner.

Although transcriptional regulators that harbor PRDs are widely studied in several bacterial species, there is no consensus whether phosphorylation of the PRD will have a stimulatory or inhibitory effect on the activity of the regulator (reviewed in Deutscher et al., 2014). It has been shown that the activity of Mga is dictated by the PTS components EI/HPr phosphorylating the PRDs, indicating its pivotal role in modulating gene expression according to the metabolic state of the cell (Hondorp et al., 2013). In the M1 serotype, it was demonstrated that the phosphorylation of either histidine residues of PRD1 had minimal impact on the expression of virulence genes arp and sof, while concurrent phosphorylation of both histidines inactivated Mga and reduced the expression these genes (Hondorp et al., 2013). This regulatory mode is consistent to a certain extent with the results presented here, where both the phosphomimetic Mga^DHD and Mga^HDD variants of PRD1 were able to restore repression of Rgg2/3 QS, while the Mga^DDH variant was unable to repress QS when grown with glucose as the sole carbon source (Figure 3D). However, the phosphomimetic variant Mga^DDD was also able to repress Rgg2/3 QS induction, while in the M1 serotype, the corresponding variant failed to initiate expression of virulence genes. Moreover, in this study, variants that harbor a phosphomimetic substitution in H324 (Mga^DHD, Mga^HDD and Mga^DDD) exhibited poor growth regardless of carbohydrate, suggesting that phosphorylation of PRD2 appears to have a detrimental impact on cellular growth. This strongly suggests that PRD2 has a dominant effect on the overall activity of Mga when it is phosphorylated in NZ131 (M49). That is in contrast to that of M4 serotype, where phosphorylation of PRD2 appears to amplify the activity of Mga with regards to virulence gene expression (Hondorp et al., 2013), while in M59 serotype, phosphorylation of PRD2 negatively influence the transcription of virulence genes (Sanson et al., 2015). It is unclear what the underlying cause of the growth defect in the M49 serotype is due to, but we speculate that in those variants, the phosphorylation of PRD2 results in the dysregulation of Mga accessory regulon. The Mga regulon of M49 remains to be determined but it likely includes ccpA, as well as genes that encode enzymes that catabolized less-preferred sugars as observed in M1 and M4 serotypes (Ribardo & McIver, 2006). Although it was reported that ccpA is essential for viability in M49 but not in other serotypes like M1 and M14 (Le Breton et al., 2015, Paluscio et al., 2018), we were able to obtain a viable ccpA deletion, but has a growth defect in CDM-Glu. This is consistent with the findings reported in other species such as Streptococcus mutans (Abranches et al., 2008), Streptococcus pneumoniae (Carvalho et al., 2011) and Bacillus subtilis (Henkins, 1996). It was proposed that the growth defect is due to intracellular accumulation of phosphorylated glycolytic intermediates that reduce cellular growth and glycolytic flux and could potentially be toxic to the cell (Andersen et al., 2001, Fraenkel, 1968). Therefore, it is plausible that those Mga variants that exhibit poor growth, could have dysregulated ccpA expression and/or activity which consequently affects their central metabolic flux. Alternatively, it is plausible that Mga being a growth phase regulator (McIver & Scott, 1997), the phosphorylation of PRD2 could potentially be an intracellular signal for stationary phase adaptation upon depletion of carbohydrates (see below). Currently, we are investigating the impact on global gene expression in these variants to determine a correlation with the growth phenotype.

In this study, we identified ManL as a PTS component that appears to play a regulatory role in QS induction in a carbohydrate dependent manner on top of its primary role as a mannose transporter. The regulatory role of ManLMN PTS appears to be a conserved mechanism in Streptococcus sp. such as S. mutans (Abranches et al., 2003, Abranches et al., 2006, Zeng & Burne, 2009, Zeng & Burne, 2010), S. pneumoniae (Fleming & Camilli, 2016) and S. gordonii (Tong et al., 2011) where ManL provides CCR independent of CcpA. Besides Streptococcus species, it was reported that ManL mediates CCR on catabolic operons (Vu-Khac & Miller, 2009), as well as influencing the expression of virulence genes belonging to the regulon of PrfA in Listeria monocytogenes (Aké et al., 2011). Thus far, the regulatory network that governs CcpA-independent CCR via ManL appears to be species-specific, but the actual mechanisms remain to be elucidated. In S. mutans, ManL appears to mediate CCR on fructose (Zeng & Burne, 2010) and cellobiose (Zeng & Burne, 2009) operons, presumably through protein-protein interaction to allosterically modulate the activity of their cognate transcriptional regulator LevR and CelR, respectively (Abranches et al., 2006). Although it remains plausible that ManL could allosterically interfere with these transcriptional regulators through protein-protein interaction, it was also proposed that ManL acts in concert with HPr-Ser46-P to mediate CCR of these metabolic operons through shunting of the phosphorelay by PTS components (Zeng & Burne, 2009). This is achieved by sensing the influx of preferred carbohydrate phosphorylatable by ManL and in parallel, HPr monitors the metabolic status of the cell based on the intracellular level of the glycolytic intermediate fructose-1,6-bisphosphate, which is a cofactor for HPrK/P to phosphorylate HPr-Ser46 (Deutscher & Saier, 1983). In the absence of glucose, HPr-Ser46-P will be dephosphorylated and instead, PEP will phosphorylate the histidine for the formation of HPr-His15~P, which will phosphorylate other EIIA proteins to prime the cells for the uptake of less preferred sugar (Hogema et al., 1998, Vadeboncoeur et al., 1991). The proposed mechanism is in line with our model on how ManL and HPr provides CCR on Rgg2/3 system through Mga in GAS (Figure 6), although it remains a possibility that ManL could modulate Mga activity through allosteric interactions as well. The only difference is that no homologs of Mga exist in S. mutans, while Mga homologs are present in both S. gordonii and S. pneumoniae. It remains to be determined if both latter species utilize Mga homologs to govern CCR as well as virulence factors expression. In L. monocytogenes, it is less clear how ManL exerts CCR control on PrfA-dependent expression of virulence factors, owing to the involvement of another PTS^Man system (EIIABCD^mpo) in addition to ManLMN (Aké et al., 2011, Zébré et al., 2015), as well as the lack of PRDs in PrfA (Scortti et al., 2007).

The work presented here also indicates that the presence of different carbohydrates is able to mediate repression of Rgg2/3 QS system which could be achieved through different phosphorylated forms of Mga. In the presence of abundant glucose, a likely situation when GAS is infecting the bloodstream of the human host, HPr exists predominantly as HPr-Ser46-P, serving as a cofactor to CcpA to mediate CCR, while HPr-His15~P is present in very low amounts (Thevenot et al., 1995, Vadeboncoeur et al., 1991). As glucose transport in GAS is PTS-independent (Sundar et al., 2017), HPr-His15~P thus serves a regulatory role by phosphorylating Mga, consequently repressing the Rgg2/3 QS system by transcribing rgg3, which encodes the repressor of the QS system. We surmise that HPr-His15~P phosphorylates the first histidine (H204) in PRD, evident in the phosphomimetic variant Mga^DHD, which represses the induction of QS even when grown in mannose (Figure 4E), suggesting that Mga is in the active state when PRD1 is phosphorylated. Although evaluating the virulence gene expression is beyond the scope of this study, we predict that this variant will present comparable expression of virulence genes to WT, as indicated by the active state of Mga in repressing the Rgg2/3 QS system.

On the contrary, when GAS is grown in mannose, QS induction is observed. When colonizing the mucosal surface of the human host, which is a glucose-deplete environment, GAS will need to utilize alternative carbon sources for proliferation. As mannose is typically available in glycans of host proteins decorating the surface of cells, as well as a component of host mucins, it could be available as a carbon source to sustain cellular viability upon degradation (Andreassen et al., 2020) by GAS that are colonizing the mucosal surface. Thus, the presence of mannose could signal to GAS that it is in a host environment that is suitable for colonization, resulting in the promotion of biofilm formation mediated through the QS system (Gogos et al., 2018). Indeed, the Rgg2/3 QS system was observed to be induced upon colonization of murine nasopharynx (Gogos & Federle, 2020), which putatively benefited GAS through suppression of the host immune system (Rahbari et al., 2021) and defense against host antimicrobial agents such as lysozyme (Chang et al., 2015).

Although it remains to be determined empirically, we favor the notion that in M49 GAS, ManL is a phosphoryl donor to Mga, plausibly at the second histidine of PRD1 (H270) to elicit repression of Rgg2/3 QS system by transcribing rgg3 (Figure 6). This is supported by our observation where the Mga^AHA variant was able to restore repression of Rgg2/3 QS system in CDM-Glu, a non-permissive condition (Figure 4B). As the Mga^AAA variant failed to repress the QS system in CDM-Glu, it is therefore assumed that the second histidine of PRD1 is phosphorylated in the Mga^AHA variant. These collective phenotypes suggest the following scenario; when M49 GAS is in a glucose-deplete environment with an abundance of a PTS-dependent carbohydrate that is not imported through ManLMN (ie; sucrose), ManL and other PTS EII components will be phosphorylated, primed for the immediate uptake of any available carbohydrate (Vadeboncoeur et al., 1991, Hogema et al., 1998). If mannose is present in addition to other less preferred carbon source such as sucrose, QS remains induced as we have previously shown that mannose-mediated induction of QS has a dominant effect over other carbohydrates, bar glucose (Chang et al., 2015). In the absence of mannose but abundance of PTS-dependent carbohydrate that GAS is able to catabolize, QS remains repressed due to ManL phosphorylating Mga. Therefore, it will appear that ManL is a carbohydrate sensor for GAS and during exponential growth in the absence of mannose, phosphorylates the second histidine of PRD1 maintaining CCR on Mga regulon. We are currently elucidating if ManL is a phosphoryl donor of Mga and if ManL also affects Mga through allosteric interactions.

It remains to be determined under what biologically relevant condition is Mga dually phosphorylated at PRD1 or at PRD2. As GAS only possess one alternative sigma factor, and instead relies on two-component signaling systems and stand-alone regulators to govern global gene expression during different growth phases (Kreikemeyer et al., 2003), perhaps the phosphorylation state of Mga is correlated with the growth phase of GAS, by acting as a phosphate sink per se for the PTS system and consequently adjusting global gene expression according to the available carbohydrate. It is tempting to speculate that depending on which histidine residue of PRD1 is phosphorylated, coupled with the abundance of phosphorylated Mga, it could potentially be a means to achieve a graded response of Mga activity, as observed with the CCR on Rgg2/3 QS system in both WT (Figure 2A) and ΔmanL mutant (Figure 5D). Other nutrient-sensing transcriptional regulators such as CodY (Brinsmade et al., 2014, Waters et al., 2016) and CcpA (Paluscio et al., 2018) have been shown to exhibit a spectrum of activity, where gene expression occurs as a hierarchical model. Application of this theory could explain the growth phenotypes of the phosphomimetic Mga variants (Figure 4D). Based on the model proposed here, it will appear that during exponential phase of growth and depending on which carbohydrate is available, Mga could be phosphorylated at PRD1 on a single histidine residue, thus allowing discriminate expression or repression of its regulon (Ribardo & McIver, 2006, Sanson et al., 2015, Valdes et al., 2018). Upon the depletion of nutrients, PRD1 could be dually phosphorylated, presumably by EI/HPr at H204 and ManL at H270. Subsequently at stationary phase, any remaining HPr-His15~P could phosphorylate PRD2, as a signal to indicate the complete exhaustion of nutrients and thus reduce the growth rate of the bacterium. Work is currently evaluating how the phosphorylation of each histidine residues in both PRD1 and PRD2 of Mga impact global gene expression in GAS M49 serotype which will provide valuable insight if a hierarchy of Mga gene expressions exist and does it correlate with the phosphorylation status of each histidine.

In conclusion, we have determined that the induction of Rgg2/3 QS system is influenced by carbon availability, mediated by Mga, a virulence gene regulator and ManL, a PTS transporter that has a regulatory role as a carbohydrate sensor. Although it remains to be empirically determined, we predict that ManL is an interacting partner of Mga to facilitate the repression of QS, plausibly either through post-translational modification of histidine residues within PRDs of Mga, or through allosteric interactions, or in combination. This is of particular significance as to date, although PRD-containing transcriptional regulators have been shown to interact with EII components of the PTS system, these regulators are solely dedicated to the regulation of a genes that encode catabolic enzymes and carbohydrate transporters, instead of a global transcriptional regulator. The results presented here also highlighted the intricate interplay between carbohydrate catabolism and virulence of GAS and provided valuable insight on the mechanism of this frequently observed by poorly understood phenomenon that is conserved in other pathogens.