The ChrSA and HrrSA Two-Component Systems Are Required for Transcriptional Regulation of the hemA Promoter in Corynebacterium diphtheriae

ABSTRACT

Corynebacterium diphtheriae utilizes heme and hemoglobin (Hb) as iron sources for growth in low-iron environments. In C. diphtheriae, the two-component signal transduction systems (TCSs) ChrSA and HrrSA are responsive to Hb levels and regulate the transcription of promoters for hmuO, hrtAB, and hemA. ChrSA and HrrSA activate transcription at the hmuO promoter and repress transcription at hemA in an Hb-dependent manner. In this study, we show that HrrSA is the predominant repressor at hemA and that its activity results in transcriptional repression in the presence and absence of Hb, whereas repression of hemA by ChrSA is primarily responsive to Hb. DNA binding studies showed that both ChrA and HrrA bind to the hemA promoter region at virtually identical sequences. ChrA binding was enhanced by phosphorylation, while binding to DNA by HrrA was independent of its phosphorylation state. ChrA and HrrA are phosphorylated in vitro by the sensor kinase ChrS, whereas no kinase activity was observed with HrrS in vitro. Phosphorylated ChrA was not observed in vivo, even in the presence of Hb, which is likely due to the instability of the phosphate moiety on ChrA. However, phosphorylation of HrrA was observed in vivo regardless of the presence of the Hb inducer, and genetic analysis indicates that ChrS is responsible for most of the phosphorylation of HrrA in vivo. Phosphorylation studies strongly suggest that HrrS functions primarily as a phosphatase and has only minimal kinase activity. These findings collectively show a complex mechanism of regulation at the hemA promoter, where both two-component systems act in concert to optimize expression of heme biosynthetic enzymes.

IMPORTANCE Understanding the mechanism by which two-component signal transduction systems function to respond to environmental stimuli is critical to the study of bacterial pathogenesis. The current study expands on the previous analyses of the ChrSA and HrrSA TCSs in the human pathogen C. diphtheriae. The findings here underscore the complex interactions between the ChrSA and HrrSA systems in the regulation of the hemA promoter and demonstrate how the two systems complement one another to refine and control transcription in the presence and absence of Hb.

INTRODUCTION

Corynebacterium diphtheriae, a strict human pathogen, is the etiologic agent of diphtheria (1–3). After colonizing the upper respiratory tract, actively dividing cells secrete the potent diphtheria toxin (DT), which disseminates throughout the host and elicits the major clinical symptoms of the disease (1–3). Expression of DT is negatively regulated at the transcriptional level by iron and the diphtheria toxin repressor protein DtxR (3, 4). Although essential for bacterial physiology, iron is not readily available for invading pathogens since much of it is sequestered by host proteins such as transferrin, lactoferrin, and hemoglobin (Hb) (5–7). To thrive in such a stringent environment, bacteria utilize a variety of tightly regulated iron- and heme-scavenging systems (7, 8). C. diphtheriae, in particular, acquires heme iron by utilizing the HmuTUV and HtaA heme transport system (9–11). Upon entering the cytosol, the heme moiety is degraded by the heme oxygenase HmuO, causing the release of the heme-associated iron (12, 13). The transcriptional expression of the hmuO promoter is under dual regulation; in high-iron environments, DtxR represses transcription of hmuO, while the presence of heme or Hb activates transcription (14). Heme-dependent activation of hmuO is mediated by the ChrSA and HrrSA two-component signal transduction systems (TCSs) (14, 15). Two-component signal transduction systems are ubiquitous in bacteria and are critical for enabling bacteria to sense and rapidly adapt to environmental stimuli through modulation of gene expression (16). C. diphtheriae encodes 11 putative TCSs, but the functions for only the ChrSA and HrrSA systems are known.

The prototypical TCS includes a histidine kinase (HK) and a response regulator (RR) (17, 18). In the presence of an inducing signal, the HK undergoes autophosphorylation at a conserved histidine (17, 18). The phosphoryl group is subsequently transferred to an aspartate residue in the RR, which induces a conformational change enabling the RR to function as a DNA binding transcriptional regulator. Although most bacterial TCSs are highly specific, cross talk between similar is observed in rare instances systems (19–22). The bacterial response to the environmental signal is itself tightly regulated. In many cases, HKs have kinase activity and function as phosphatases, where they remove the phosphate group from its cognate RR (23–29) (18). Another well-described mechanism of TCS regulation involves the self-catalyzed dephosphorylation of the RR (18, 30–32). In the ChrSA and HrrSA systems, ChrS and HrrS are closely related HKs, while ChrA and HrrA also share significant sequence similarity and function as RRs that are members of the NarL family (14).

Genetic studies showed that while both ChrSA and HrrSA are essential for maximum expression of the hmuO promoter, ChrSA is the predominant regulatory system, contributing almost 80% of the Hb-dependent transcriptional activation (14, 15). ChrSA also activates transcription of genes encoding the HrtAB ABC-type transporter in an Hb-dependent manner (33). The HrtAB system is proposed to export heme, which protects the bacteria from toxic levels of heme. The HrrSA TCS has no role in transcriptional regulation at the hrtAB promoter (33).

In vitro studies showed that phosphorylated ChrA binds to the hmuO promoter region (15). DNase I footprinting results identified a protected region that contains two highly conserved 10-bp sequences as well as a portion of the putative hmuO promoter (15). Unphosphorylated ChrA displayed minimal binding to the hmuO promoter region. DNA binding studies with HrrA were not previously performed. ChrSA and HrrSA also regulate the promoter for the hemA gene, which encodes a putative glutamyl-tRNA reductase. The hemA gene is the first open reading frame in a four-gene operon that is predicted to encode enzymes involved in the early steps of heme biosynthesis (14). Previous studies showed that both ChrSA and HrrSA repressed transcription at the hemA promoter, and the presence of Hb in the growth medium resulted in a 3-fold repression of hemA transcription in the wild-type (WT) strain (14). Studies with various mutants with mutations in the chrSA and hrrSA genes revealed that repression of the hemA promoter also occurred in the absence of Hb (14). Mutations in the chrSA genes had a minimal effect on expression of the hemA promoter, while deletions of the hrrSA system had various effects on hemA transcription (14). Deletion of the hrrS gene resulted in repression of hemA transcription that was independent of Hb, while deletion of hrrA or the hrrSA double mutation resulted in hemA repression that was primarily Hb dependent. The deletion of both TCSs resulted in full derepression of hemA transcription in the presence and absence of Hb, indicating that both TCSs contribute to the repression of transcription at the hemA promoter (14). Because of the unusual phenotype of the mutants and the evidence for cross talk between the TCSs, the contribution of each TCS to the regulation of transcription at hemA could not be demonstrated with certainty in that earlier study (14).

In this study, we show that both ChrSA and HrrSA are required for transcriptional repression of the hemA promoter and that each TCS affects transcription in unique ways. DNase I footprinting studies indicate that ChrA and HrrA bind to the same regions at the hemA and hmuO promoter regions. Phosphorylation of ChrA significantly enhanced DNA binding in vitro, while the phosphorylation state of HrrA had no effect on its binding activities at either promoter region. Phosphorylation studies performed both in vivo and in vitro suggest that a complex mechanism is involved in the phosphorylation and dephosphorylation of ChrA and HrrA. Collectively, these data provide further insight into the biochemical properties of the RRs and their respective regulatory effects on heme homeostasis in C. diphtheriae.

MATERIALS AND METHODS

Bacterial strains and media.

Escherichia coli DH5α (ThermoFisher Scientific, Carlsbad, CA) was utilized for routine cloning, while E. coli BL21(DE3) (EMD Biosciences, Gibbstown, NJ) was used for protein expression. C. diphtheriae C7(−) is a nontoxinogenic wild-type strain (34). Luria-Bertani (LB) medium was used for culturing E. coli, and heart infusion broth (Difco, Franklin Lakes, NJ) containing 0.2% Tween 80 (HIBTW) was used for routine growth of C. diphtheriae. Bacterial stocks were maintained in 15% glycerol at −80°C. Antibiotics were added to LB medium at 50 μg/ml for kanamycin and at 100 μg/ml for spectinomycin, and for C. diphtheriae cultures in HIBTW medium, antibiotic concentrations were 100 μg/ml for spectinomycin and 50 μg/ml for kanamycin. Antibiotics, Tween 80, and hemoglobin (human) were obtained from Sigma Chemical Co. (St. Louis, MO).

Plasmid construction and DNA manipulation.

The plasmids utilized in this study are listed in Table 1. PCR-derived DNA fragments were initially cloned into the pCR-Blunt II-TOPO vector (ThermoFisher Scientific), and C7(−) genomic DNA was used as a template for PCRs. The promoter-probe vector pSPZ (35), used for the construction of the lacZ promoter fusions, contains a promoterless lacZ gene and replicates at low copy number in C. diphtheriae strains.

TABLE 1.

Plasmids used in this study

Plasmid	Relevant characteristics and/or usea	Reference or source
pCR-BluntII-TOPO	Cloning PCR fragments, Knr	ThermoFisher Scientific
pTOPO-PhemA	315-bp hemA promoter region in pTOPO, Knr	This study
pAKS293	PhmuO in pKS II Bluescript, Ampr	48
pSPZ	Promoter probe vector, Spcr	35
PhemA-lacZ	pTOPO-PhemA 315-bp insert in pSPZ, Spcr	This study
pGEX-6P-1	Expression vector with GST tag, Ampr	GE Healthcare
pGST-ChrS	chrS kinase domain cloned into pGEX-6P-1, Ampr ChrSGST	33
pGST-HrrSKin1	720-bp hrrS kinase domain cloned into pGEX-6P-1, Ampr	This study
pGST-HrrSKin3	735-bp hrrS kinase domain cloned into pGEX-6P-1, Ampr	This study
pGST-HrrSKin3	768-bp hrrS kinase domain cloned into pGEX-6P-1, Ampr	This study
pET28a	Expression vector, Knr	Agilent Technologies
pET-ChrA	chrA cloned into pET28a vector, Knr ChrAHIS	15
pET-HrrA	hrrA cloned into pET28a vector, Knr HrrAHIS	This study
pECK18mob2	C. diphtheriae shuttle vector, Knr	49
pECK-ChrSA	chrSA cloned into pECK18mob2, Knr	14
pECK-HrrSA	hrrSA cloned into pECK18mob2, Knr	14
pECK-ChrSA/HrrSA	chrSA and hrrSA cloned into pECK18mob2, Knr	This study

^aAmp, ampicillin; Kn, kanamycin; Spc, spectinomycin.

Expression vectors carrying the cloned full-length hrrS gene, which includes the coding region for the N-terminal membrane-anchored sensor domain, were unstable in E. coli, and therefore, only HrrS C-terminal fusion constructs were developed (not shown). Plasmids pGST-HrrS_Kin1, pGST-HrrS_Kin2, and pGST-HrrS_Kin3, contain three different regions of the HrrS C-terminal domain and were constructed by PCR amplification of 720-bp, 735-bp, and 768-bp fragments, respectively. The amplified fragments were ligated into the expression vector pGEX-6P-1 to engineer an N-terminal glutathione S-transferase (GST) fusion with the putative HrrS C-terminal kinase domains. All three of the GST-HrrS constructs were tested for autophosphorylation and for the ability to phosphorylate HrrA_His. GST-HrrS_Kin3 was used in the in vitro phosphatase studies. A 639-bp DNA fragment carrying the hrrA gene with an N-terminal histidine tag was directionally cloned into pET28a to generate pET-HrrA. The four chrSA and hrrSA open reading frames were combined and cloned into pECK18mob2 to generate pECKhrrSA/chrSA (GenScript, Piscataway, NJ). Plasmid pECK18mob2 is an E. coli-C. diphtheriae shuttle vector that replicates at low copy in C. diphtheriae. This plasmid also contained 574 bp upstream of chrSA and 61 bp upstream of hrrSA to ensure that the expression of the TCSs was driven from their native promoters. A hemA-lacZ transcriptional fusion was constructed by ligating a 315-bp DNA fragment carrying the hemA promoter region into the multiple-cloning site of pSPZ.

β-Galactosidase assay.

C. diphtheriae strains containing promoter-lacZ fusion constructs were cultured overnight in HIBTW medium at 37°C and subsequently inoculated at an optical density at 600 nm (OD₆₀₀) of 0.2 into fresh HIBTW medium containing various supplements as indicated. After 20 to 22 h of growth at 37°C, LacZ activity was determined as previously described (36).

Protein purification and antibody production.

ChrS_GST and ChrA_His were purified as previously described (15) with the following modification. Briefly, cell pellets harboring either ChrS_GST or ChrA_His were resuspended in their respective lysis buffers supplemented with 1× Complete EDTA-free protease inhibitor cocktail (Roche, Pleasanton, CA). Cells were transferred to 2.0-ml microtubes partially filled with 0.1-mm-diameter glass beads (BioSpec Products, Bartlesville, OK). A FastPrep 24 sample preparation system (MP Biomedical, Santa Ana, CA) was used to mechanically lyse the cells. The lysate was separated from the glass beads by centrifugation at 15,000 rpm for 10 min at 4°C and subsequently transferred to the appropriate purification resin.

The same purification protocols used for ChrS_GST and ChrA_His were utilized for HrrS_GST and HrrA_His, respectively. The recombinant proteins were stored at −80°C in phosphate-buffered saline (PBS)–20% glycerol. GST C-terminal fusions were used for ChrS and HrrS. Purified Bordetella pertussis response regulator BvgA was a gift from E. S. Stibitz. The purified histidine kinases and response regulators were utilized for production of polyclonal antibodies in guinea pigs using standard methods (Cocalico Biologicals, Inc., Stevens, PA). Monoclonal antibodies for the detection of BvgA were also provided by E. S. Stibitz.

EMSA.

Electrophoretic mobility shift assay (EMSA) reactions were conducted as previously indicated (15). RR proteins used in the EMSA procedure were phosphorylated by ChrS as follows. ChrS was incubated in phosphorylation buffer (with or without 1 mM ATP) for 10 min at 30°C, followed by the addition of ChrA or HrrA in a volume of 15 μl. The reaction mixture was incubated for an additional 10 min at 30°C, and the phosphorylated ChrA or HrrA proteins were then added to the binding reaction mixture (15), which included the DNA fragment (0.1 fmol of biotin-labeled DNA) harboring either the hemA or hmuO promoter region in a 20-μl final volume. After a 20-min incubation, the reactions were terminated with the addition of EMSA loading dye, and the mixtures were loaded onto a 5% nondenaturing polyacrylamide gel and separated at 4°C in 0.5× Tris-borate-EDTA buffer (Bio-Rad, Hercules, CA). The binding products were detected as previously described (15).

DNase I protection experiments.

To generate fragments for DNase I footprint analysis, pAKS293 (37) and pTOPO-P_hemA were digested with HindIII/PstI. The 306-bp P_hmuO and 414-bp P_hemA fragments were radiolabeled with [α-³²P]dCTP at their 5′ ends with Klenow fragment (37). Approximately 60,000 cpm of radiolabeled DNA was added to the binding reaction mixture (8.4 mM NaH₂PO₄, 11.6 mM Na₂HPO₄, 50 mM NaCl, 1 mM MgCl₂, 200 μg bovine serum albumin [BSA], 2 μg salmon sperm DNA, 10% glycerol), which contained the indicated amounts of phosphorylated RR protein in a 50-μl final volume. DNase I footprint analysis was conducted as previously described (15). RR proteins used in the DNase I experiments were phosphorylated by ChrS_GST as described above for the EMSA procedure. A sequencing ladder for both promoters was prepared by the Sanger method utilizing the USB-Sequenase version 2.0 DNA sequencing kit (USB Corporation, Cleveland, OH) as directed by the manufacturer.

In vitro phosphorylation assays.

To determine if ChrS_GST phosphorylates HrrA_His in vitro, 1 μg of ChrS_GST was incubated at 30°C in phosphorylation buffer (38) in the presence or absence of 1 mM ATP (New England BioLabs, Ipswich, MA). Subsequently, 1 μg of HrrA_His was added to the reaction mixture, and the proteins were incubated at 30°C for an additional 10 min. Phosphorylation reaction mixtures containing ChrA_His or BvgA were used as positive and negative controls, respectively. The reactions were terminated with the addition of SDS loading dye (1% SDS, 65 mM Tris [pH 6.8], and 80% glycerol, supplemented with bromophenol blue) and separated on polyacrylamide gels supplemented with Phos-Tag acrylamide (Wako Pure Chemical, Richmond, VA) using a previously described protocol with minor modifications (39). Briefly, Phos-Tag gels were composed of a 10% resolving gel [29:1 acrylamide–N,N′-methylene-bisacrylamide (Bio-Rad, Hercules, CA), 350 mM Tris (pH 6.8), 0.1% SDS, 75 μM Phos-Tag acrylamide, and 150 μM Zn(NO₃)₂] and a 4% stacking gel (4% [wt/vol] 29:1 acrylamide–N,N′-methylene-bisacrylamide, 350 mM Tris-Cl [pH 6.8 at 4°C], and 0.1% SDS). Electrophoresis was performed at 116 V at 4°C for 1 h 30 min in MOPS (morpholinepropanesulfonic acid) running buffer (pH 8.0).

In vivo phosphorylation assays.

For the in vivo detection of ChrA, WT strain C7(−) was initially cultured overnight in HIBTW medium at 37°C and subsequently inoculated to an OD₆₀₀ of 0.2 into fresh HIBTW (Hb⁻) or HIBTW which was supplemented with 200 μg/ml Hemoglobin (Hb⁺). After an overnight incubation at 37°C, cells were harvested by centrifugation at 3,100 rpm for 10 min at 4°C. Subsequently, the cell pellet was resuspended in cold PBS (pH 7.4) supplemented with 1× Complete EDTA-free protease inhibitor cocktail (Roche). Cells were lysed using a FastPrep 24 sample preparation system. The lysate was separated from the glass beads by centrifugation at 15,000 rpm for 10 min at 4°C and subsequently electrophoresed on a Phos-Tag gel. ChrA was visualized by Western blotting using ChrA-specific antibodies.

For the in vivo detection of HrrA, C. diphtheriae C7(−) (WT), C7 ΔchrS, and C7 ΔhrrS were cultured at 37°C in the presence or absence of Hb. As negative controls, C7 ΔchrS ΔhrrS and C7 ΔchrSA ΔhrrSA were cultured in the presence of Hb. All strains were grown to mid-log phase following overnight incubation. Cell lysates were treated identically as described for WT C7(−) above. HrrA was detected by Western blotting using HrrA-specific antibodies.

In vitro analysis of ChrA∼P stability.

The stability of phosphorylated ChrA_His (ChrA_His∼P) was examined using an in vitro procedure as follows. Briefly, ChrS_GST was phosphorylated by incubating 3 μg of ChrS_GST at 30°C for 10 min in phosphorylation buffer containing 1 mM ATP. Subsequently, an equal amount of ChrA_His was added to the phosphorylation reaction mixture and incubated at 30°C for an additional 10 min. The reaction mixture was diluted with 300 μl cold dephosphorylation buffer (50 mM Tris [pH 7.5], 10% glycerol, 100 mM MgCl₂, 2 mM dithiothreitol [DTT], 50 mM KCl) (38) and added to 75 μl of glutathione-Sepharose (GE Healthcare, Pittsburgh, PA). As previously indicated, ChrA contains an N-terminal His fusion, while ChrS (C-terminal region) is fused at its N terminus with GST, which enables ChrS but not ChrA to adhere to the glutathione-Sepharose resin (15). After a 5-min incubation at 4°C, the protein-resin mixture was centrifuged at 15,000 rpm for 30 s at 4°C. The supernatant containing ChrA∼P was quickly separated from the resin and transferred to 30°C; at various time points (0, 1, 5, 10, 30, and 60 min), the reaction was terminated by the addition of SDS loading buffer. The phosphorylation status of ChrA was assessed by Western blotting after electrophoreses on a Phos-Tag gel.

In vitro analysis of HrrA∼P stability.

The effect of the HKs, ChrS and HrrS, on the phosphorylation state of HrrA_His was examined using an in vitro assay as follows. Briefly, 3 μg of HrrA_His was phosphorylated by ChrS_GST in a manner similar to that described above for ChrA_His. Following the separation of the phosphorylated HrrA_His from ChrS_GST, equal amounts of unphosphorylated ChrS_GST or GST-HrrS_Kin3 were added separately to phosphorylated HrrA (HrrA∼P). The protein mixtures were incubated at 30°C, and samples were removed at 0, 5, 10, 30, or 60 min. The reactions were terminated with the addition of SDS loading buffer, and the phosphorylation status of HrrA was assessed by Western blotting after electrophoreses on a Phos-Tag gel.

Western blot analysis with Phos-Tag gels.

Following electrophoresis, the Phos-Tag gel was washed for 10 min with transfer buffer (25 mM Tris base, 0.192 M glycine, 20% methanol) supplemented with 1 mM EDTA to remove Zn⁺ from the gel. The gel was washed again with transfer buffer to remove any chelated metal and transferred to a nitrocellulose filter (ThermoFisher Scientific) using a Mini Protean II system (Bio-Rad) at constant voltage of 100 V for 1 h. The filters were blocked with 5% Blotto (Bio-Rad) for 1 h at room temperature and subsequently incubated with the appropriate antibodies in 5% Blotto. Anti-6×His antibodies (HIS-1 [Abcam, Cambridge, MA]) were used at a dilution of 1:2,000, and BvgA monoclonal antibodies were diluted to 1:5,000. Antibodies for ChrA, HrrA, and HrrS were used at dilutions of 1:3,500, while a dilution of 1:1,500 was used for ChrS. After three 15-min washes with Tris-buffered saline (TBS) plus 0.05% Tween (TBST), the filter was incubated with either rabbit anti-guinea pig IgG–horseradish peroxidase (HRP) (1:50,000) or anti-mouse IgG–alkaline phosphatase (AP) (1:20,000) for 1 h. After three washes with TBST, the filter was developed using 5-bromo-4-chloro-3-indolylphosphate–nitroblue tetrazolium (BCIP/NBT) phosphatase substrate (KPL, Gaithersburg, MD) (for AP-conjugated antibodies) or SuperSignal West Pico chemiluminescent substrate (ThermoFisher Scientific) (for HRP-conjugated antibodies).

RESULTS

Transcriptional regulation of the hemA promoter by HrrSA and ChrSA.

The hemA gene is the promoter-proximal gene in a putative operon that also contains three additional genes predicted to encode proteins involved in heme biosynthesis (Fig. 1A) (14). Previous studies that examined the effect of deletions in the chrSA and hrrSA genes were unable to clearly define the relative contribution of each of these systems to the Hb-dependent repression at the hemA promoter (P_hemA) (14). To better understand the roles that the ChrSA and HrrSA systems have in P_hemA expression, we examined the effect of the cloned chrSA and hrrSA genes on hemA promoter activity in a C. diphtheriae strain that carried chromosomal deletions for both chrSA and hrrSA (C7 ΔchrSAΔ hrrSA) and harbored a P_hemA-lacZ reporter construct. C7 ΔchrSA ΔhrrSA P_hemA-lacZ carrying various cloned components of the chrSA and hrrSA genes was cultured in HIBTW medium in the presence or absence of Hb, and the transcriptional activity of P_hemA-lacZ was examined. As observed previously, the C7(−) wild-type (WT) strain carrying only P_hemA-lacZ showed approximately 3-fold repression in the presence of Hb (Fig. 1B) (14). The mutant strain, C7 ΔchrSA ΔhrrSA P_hemA-lacZ, carrying the vector control, showed high levels of promoter activity in the presence and absence of Hb (Fig. 1B). When the cloned chrSA and hrrSA genes were both introduced into the double mutant, Hb-dependent repression of hemA promoter activity was restored to wild-type levels (Fig. 1B). To determine how each of the TCSs affects transcription of the hemA promoter, we introduced the cloned hrrSA and chrSA genes separately into the double mutant strain. When hrrSA (phrrSA) was introduced into C7 ΔchrSA ΔhrrSA P_hemA-lacZ, the levels of promoter activity were significantly lower than those observed with the vector control strain under high- and low-Hb conditions. However, a statistically significant decrease in hemA expression was observed in the presence of Hb (Fig. 1B). It should be noted that although the observed levels of promoter activity in C7 ΔchrSA ΔhrrSA P_hemA-lacZ/phrrSA were lower than those in the vector control strain, promoter activity was significantly higher than levels observed in the wild-type strain. When the chrSA genes (pchrSA) were placed into C7 ΔchrSA ΔhrrSA P_hemA-lacZ, hemA promoter activity was reduced in the presence and absence of Hb, although repression of hemA transcription was significantly greater in the presence of Hb (Fig. 1B). These findings indicate that both regulatory systems repress hemA transcription in the presence and absence of Hb. However, the HrrSA system has a greater impact on the repression of P_hemA than ChrSA. No repression of the hemA promoter was observed when the individual cloned genes (chrA, hrrA, chrS, or hrrS) were examined in C7 ΔchrSA ΔhrrSA P_hemA-lacZ (data not shown).

FIG 1.

Regulation of the hemA promoter by the ChrSA and HrrSA two-component systems. (A) Genetic organization of the hemACDB operon. Arrows indicate direction of transcription, and boxes represent putative −10 promoter elements. (B) The hemA promoter was assessed for transcriptional activity in the C7 wild-type strain (wt) and in the C7 ΔchrSA ΔhrrSA double mutant. The hemA promoter fusion P_hemA-lacZ was introduced into the wt and C7 ΔchrSA ΔhrrSA strains, which also carried either the pECK18mob2 vector or the cloned hrrSA or chrSA genes present on pECK18mob2 as indicated (pchrSA, phrrSA, or both sets of genes as indicated by phrrSA/chrSA). Strains were cultured in HIBTW in the absence (−Hb) or presence (+Hb) of 200 μg/ml Hb. LacZ units were determined as described in Materials and Methods. Values are the means from three independent experiments conducted in triplicate. * and **, results for cultures with and without Hb are significantly different (P < 0.01). P values and standard deviations were derived using a statistical software package provided by GraphPad Prism.

ChrA and HrrA bind to the hemA promoter region.

To further elucidate the mechanism by which ChrA and HrrA regulate P_hemA, we assessed the ability of purified ChrA and HrrA to bind a 315-bp DNA fragment that contained the hemA promoter region. EMSAs were performed using biotinylated DNA and various amounts of the RRs that were either phosphorylated or not phosphorylated. The presence of ChrA∼P at levels at or above 1 μg resulted in reduced mobility in the DNA fragment carrying P_hemA (Fig. 2A). However, only a minimal shift in the DNA fragment carrying P_hemA was observed in the absence of ATP, and this was detected only when high levels of ChrA were utilized (Fig. 2B). This finding indicates that ChrA is able to bind to the hemA promoter region and that binding is greatly enhanced when the protein is phosphorylated (ChrA∼P). As shown in Fig. 3A and B, the presence of HrrA resulted in reduced mobility in the DNA fragment carrying P_hemA in the presence and absence of ATP. We next analyzed the binding of HrrA to the hmuO promoter, a promoter that was previously shown to be activated by both ChrA and HrrA in an Hb-dependent manner (14). EMSAs were conducted utilizing a 306-bp DNA fragment that harbored the hmuO promoter and upstream sequences. As with P_hemA, HrrA bound to P_hmuO in the presence and absence of ATP (Fig. 3C and D). Previous studies showed that ChrA bound to P_hmuO only in the phosphorylated state (15). Demonstrating specificity of binding, a negative-control DNA fragment containing the sigA coding region, which was used previously in EMSAs (15), showed no interaction with either HrrA or ChrA in the presence or absence of ATP (data not shown). These EMSA studies provide strong evidence that both ChrA and HrrA directly and specifically regulate the hemA and hmuO promoters, although the requirement for phosphorylation differs between these RRs.

FIG 2.

ChrA binds to the hemA promoter region. EMSAs were conducted to analyze the binding of ChrA to the hemA promoter region. Phosphorylated ChrA (+ATP) (A) and unphosphorylated ChrA (−ATP) (B) were assessed at increasing levels (0, 1.0, 2.0, 4.0, and 6.0 μg [0, 2.35, 4.7, 9.4, and 14.1 μM, respectively]) for their ability to bind a 315-bp biotinylated DNA fragment carrying the hemA promoter region (P_hemA). Experiments were repeated multiple times with similar results. Results from representative experiments are shown.

FIG 3.

HrrA binds to the hemA and hmuO promoter regions. EMSAs were conducted to study the binding profile of HrrA. Phosphorylated HrrA (+ATP) (A and C) and unphosphorylated HrrA (−ATP) (B and D) were assessed at increasing concentrations for their ability to bind to either the hemA (P_hemA) (A and B) and hmuO (P_hmuO) (C and D) promoter regions. The amounts of HrrA used in the EMSAs were 0, 0.6, 1.0, and 2.5 μg (0, 1.3, 2.2, and 5.6 μM, respectively) for the hemA promoter and 0, 0.1, 0.2, 0.4, and 0.6 μg (0, 0.2, 0.4, 0.8, and 1.3 μM, respectively) for the hmuO promoter. Experiments were repeated multiple times with similar results. Results from representative experiments are shown.

DNase I protection studies at the hemA and hmuO promoters.

Since the EMSA results indicated that both RRs directly bind to the region upstream of hemA, we next utilized DNase I protection studies to determine the specific sequences in the hemA promoter region that are bound by ChrA and HrrA. The footprint experiments showed that both ChrA∼P and HrrA∼P protect an identical 71-bp region from DNase I digestion (Fig. 4A and B). The protected region starts at 16 bp upstream from the GTG translational start codon for hemA and includes two hexameric sequences that likely serve as −10 regions for the putative hemA promoter (Fig. 4C). The protected region remained unchanged when footprinting was conducted with reaction mixtures containing both ChrA∼P and HrrA∼P (data not shown). Sequence analysis of P_hemA identified the presence of two 10-bp sequences within the protected region that may serve as the primary DNA recognition sequences for ChrA and HrrA (Fig. 4C). These 10-bp sites, designated ha1 (AGTTAGTAAG) and ha2 (GGTTGATATT), are divergently oriented and separated by 4 bp (Fig. 4C). The orientations and sequences of these two conserved regions are similar to the ChrA binding sites at the hmuO and hrtAB promoter regions (14). Compared to the consensus sequence derived from the ChrA binding sites at the hmuO and hrtAB promoters, ha1 showed a conservation of 5/10 nucleotides and ha2 contained a match of 8/10 nucleotides (Fig. 4D).

FIG 4.

DNase I protection experiments with ChrA and HrrA at the hemA promoter. (A and B) DNase I protection experiments were conducted to identify ChrA (A) and HrrA (B) binding sites at the hemA promoter. The binding reaction mixtures contained a radiolabeled DNA fragment carrying the hemA promoter region and various amounts of ChrA (0, 0.5, 1, 3, and 6 μg [0, 1.2, 2.4, 7.2, and 14.4 μM, respectively]) and HrrA (0, 1, and 6 μg [0, 2.2, and 13.2 μM, respectively]). (The 1-ìg lane for ChrA∼P is underloaded.) Arrows denote the orientations of ha1 and ha2, while the bar represents the ChrA and HrrA protected region. Hypersensitive sites are indicated (*). Experiments were repeated multiple times with similar results. Results from representative experiments are shown. (C) DNA sequence of the hemA promoter and the adjacent upstream region. The putative −10 hemA promoter elements are in italics and underlined, and the GTG translational start codon for the hemA gene is boxed. Conserved regions, ha1 and ha2, are in bold, and their orientations are represented by arrows. The 71-bp sequence protected from DNase I digestion by ChrA and HrrA is shown by the open bar. (D) Nucleotide alignment of ha1/ha2 to the consensus ChrA binding sequence. Conserved nucleotides in ha1 and ha2 are shown in bold and underlined.

Since ChrA and HrrA had virtually identical binding sites at the hemA promoter, we sought to compare the binding patterns of ChrA∼P and HrrA∼P at the hmuO promoter region. DNase I protection studies at the hmuO promoter region were previously performed with ChrA∼P, where the pattern showed three distinct protected regions separated by hypercleavable sequences (15) (Fig. 5). The hmuO promoter region contains two well-characterized 10-bp binding regions, S2ho and S3ho, which are critical for the binding of ChrA∼P (15) (Fig. 5). DNase I protection experiments at the hmuO promoter showed that HrrA, HrrA∼P, ChrA∼P, and reaction mixtures containing both RRs gave virtually identical protection patterns (Fig. 5). The footprinting results strongly suggest that ChrA and HrrA utilize a common binding site at both the hemA and hmuO promoters.

FIG 5.

ChrA and HrrA bind to similar regions at the hmuO promoter. DNase I protection experiments were conducted to identify ChrA and HrrA binding sites at the hmuO promoter region. ChrA and HrrA were incubated with a radiolabeled 306-bp DNA fragment containing the hmuO promoter region as described in Materials and Methods. Sequences protected from DNase I digestion by ChrA and HrrA are identified by the closed bars. Arrows indicate the locations of the previously characterized ChrA binding sites S2ho and S3ho. Hypersensitive sites are indicated (*). ChrA and HrrA were used at 6 μg in the binding reaction mixture, and both phosphorylated (HrrA∼P) and unphosphorylated HrrA were examined. Experiments were repeated multiple times with similar results. Results from a representative experiment are shown.

In vitro phosphorylation of HrrA.

ChrS is able to phosphorylate its cognate RR ChrA in vitro in the presence of ATP (15). This in vitro function is thought to mimic the activity of these proteins in vivo when an inducing signal such as heme or Hb is present in the growth medium (15, 23). It is not known if the HrrSA system functions in a manner similar to that of ChrSA. Phosphorylation experiments identical to those performed with ChrS and ChrA showed that HrrS was unable to phosphorylate HrrA in vitro (data not shown). Moreover, HrrS was not capable of autophosphorylation in the presence of ATP, a function previously shown for ChrS (15). Phosphorylation activity was not observed in any of the three HrrS-GST fusion proteins (see Materials and Methods). However, ChrS was able to phosphorylate HrrA in vitro under the same conditions that were previously shown for phosphorylation of ChrA (Fig. 6A) (39). As shown in Fig. 6A, both ChrA and HrrA are phosphorylated only in the presence of ChrS and ATP: the slower-migrating band in the Phos-Tag gel is the phosphorylated form of the RR. The negative-control protein BvgA, an RR from B. pertussis, was not phosphorylated by ChrS (Fig. 6A) (39) but is phosphorylated by the nonspecific phosphate donor acetyl phosphate (39) (Fig. 6A). ChrA and HrrA are also phosphorylated by acetyl phosphate (reference 15 and data not shown for HrrA).

FIG 6.

In vitro phosphorylation of ChrA and HrrA. (A) ChrA and HrrA are specifically phosphorylated by ChrS in the presence of ATP (+). HrrA, ChrA, or BvgA were incubated with ChrS (with [+] or without [−] 1 mM ATP), and the reaction mixtures were electrophoresed on Phos-Tag acrylamide gels. BvgA was phosphorylated alone with 20 mM acetyl phosphate. In the Phos-Tag system, phosphorylated proteins bind to the Phos-Tag acrylamide in the gel matrix, which results in a decrease in the migration of phosphorylated proteins. The rate of migration of phosphorylated proteins can vary greatly among response regulators and does not reflect protein size or level of phosphorylation. Proteins were visualized by Western blotting using anti-His antibodies (for ChrA and HrrA) or BvgA-specific antibodies. (B) HrrA is not autophosphorylated in vitro. Purified HrrA was incubated alone in phosphorylation buffer (with [+] or without [−] 1 mM ATP) at 30°C. Samples were obtained at 0, 10, 20, or 30 min and electrophoresed on Phos-Tag acrylamide gels. As a positive control, HrrA was incubated with ChrS and 1 mM ATP as described for panel A. HrrA was visualized by Western blotting using HrrA-specific antibodies. Experiments were repeated multiple times with similar results. Results from representative experiments are shown.

As shown in Fig. 6B, phosphorylation of HrrA was not observed in the absence of ChrS, which confirms that phosphorylation of HrrA requires ChrS and ATP and that autophosphorylation of HrrA does not occur. These data suggest that ChrS can specifically transfer a phosphate moiety to both ChrA and HrrA and that the phosphorylation of HrrA by ChrS further supports previous evidence for cross talk between these two systems (14).

In vivo phosphorylation of ChrA.

Although previous in vitro work established that ChrA is rapidly phosphorylated by ChrS (15), it has yet to be elucidated whether ChrA phosphorylation can be observed in vivo in the presence of Hb. To determine the in vivo phosphorylation profile of ChrA, the WT strain C7(−) was grown in the presence and absence of Hb, and proteins present in whole-cell lysates were examined in Phos-Tag gels. Surprisingly, only a single band, corresponding to unphosphorylated ChrA, was observed in both the presence and absence of Hb (Fig. 7). This finding indicates that ChrA∼P was not detected in the presence of Hb, a strong inducer of the ChrSA two-component system (14, 40). The inability to detect ChrA∼P in the presence of Hb was unexpected. It is possible that ChrA∼P is unstable and has a short half-life in the absence of Hb or ChrS. Because of instability, the phosphate group on ChrA∼P may be rapidly lost during the processing of the bacteria prior to analyses by electrophoresis and Western blotting. In an attempt to detect ChrA∼P in vivo, we performed numerous modifications to our bacterium- and protein-processing methods, such as using low temperatures, phosphatase inhibitors, and accelerated processing times, among other changes. However, all of these attempts were unsuccessful. ChrA∼P was also not detected in vivo in C7 strains that contain deletions in hrrS or in the hrrSA double mutant. We also utilized a system that we previously developed in E. coli, in which the Hb-dependent activation of the hmuO promoter by ChrSA was reconstituted in E. coli DH5α (40). However, despite high levels of hmuO promoter activity, which was dependent on Hb and ChrSA, we were unable to detect ChrA∼P in this E. coli system; only the unphosphorylated form of ChrA was observed (data not shown). These data suggest that this phenomenon is not unique to C. diphtheriae, since ChrA∼P was also not detected in the reconstituted system in E. coli in the presence of the Hb inducer. These findings suggest either that ChrA∼P levels were below our limit of detection or that ChrA was rapidly dephosphorylated by an unknown mechanism.

FIG 7.

In vivo phosphorylation of ChrA. The C. diphtheriae WT strain C7(−) was grown in HIBTW in the absence (−) or presence (+) of 200 μg/ml Hb, and the phosphorylation status of ChrA was determined. Whole-cell lysates were electrophoresed in a Phos-Tag acrylamide gel, and ChrA was visualized by Western blotting using ChrA-specific antibodies. ChrA phosphorylated in vitro is shown as a control. Experiments were repeated multiple times with similar results. Results from a representative experiment are shown.

To determine if ChrA∼P was undergoing rapid dephosphorylation in vivo, we developed an in vitro assay that enabled us to rapidly analyze ChrA upon phosphorylation by ChrS (Fig. 8A). This method involved the phosphate labeling of the His-tagged ChrA protein (ChrA_His) by the GST-ChrS fusion protein as described above. After phosphorylation of ChrA_His, the reaction mixture was added to GST resin, which bound only to GST-ChrS. The unbound ChrA_His in the supernatant was rapidly separated from both the resin and the GST-ChrS protein and then mixed with SDS-PAGE loading dye. Samples containing ChrA_His were obtained at different time points after separation from GST-ChrS and analyzed by Western blotting on Phos-Tag gels as described above. As shown in Fig. 8B, ChrA∼P was observed only at the earliest time point (T = 0). No contaminating GST-ChrS was present in the ChrA_His samples (data not shown). This finding indicates that the phosphate group on ChrA∼P is rapidly lost within minutes after separation of ChrA∼P from ChrS and ATP. This instability of the phosphate moiety in vitro likely indicates why ChrA∼P was not detected in C. diphtheriae (or E. coli) cultures grown in the presence of Hb. Additionally, the ability to detect ChrA∼P in the in vitro experiment for Fig. 6A is due to the continuous presence of ChrS and ATP. ChrS∼P showed no instability in vitro (Fig. 8C). We were unable to detect ChrS in vivo in either whole-cell lysates or membrane preparations, which we believe is due to the presence of ChrS at levels below the detection with our ChrS-specific antibody.

FIG 8.

Phosphorylated ChrA is unstable in vitro. (A) The flow diagram shows the procedure used to separate ChrA_His∼P from ChrS_GST∼P. ChrA_His was phosphorylated by ChrS_GST in the presence of ATP at 30°C for 10 min. After the incubation, the phosphorylation reaction mixture was mixed with GST resin, which specifically binds to ChrS_GST, and then ChrA_His was rapidly separated from ChrS_GST by centrifugation (cfg) at 15,000 rpm for 30 s. The supernatant containing ChrA_His∼P was placed at 30°C, and samples were removed at various time points and mixed with SDS buffer to terminate the reaction. ChrA_His∼P was analyzed as described for panel B. (B) Phosphorylated ChrA_His (ChrA∼P) was assessed for stability of the phosphate group at various time points after phosphate labeling by ChrS and ATP (0, 1, 5, 10, 30, and 60 min). The phosphorylation status of ChrA was assessed after electrophoreses on a Phos-Tag acrylamide gel as described for Fig. 6. (C) The in vitro stability of ChrS was examined after phosphate labeling of ChrS with (+) or without (−) ATP. After a 30-min incubation, ChrS was electrophoresed on a Phos-Tag acrylamide gel. ChrS was observed by Western blotting using ChrS-specific antibodies. Experiments were repeated multiple times with similar results. Results from representative experiments are shown.

In vivo phosphorylation of HrrA.

We next examined the phosphorylation profile of HrrA in the C. diphtheriae WT strain C7(−) and in various mutant strains of C7 that were grown in the presence and absence of Hb. In the wild-type strain, both HrrA∼P and HrrA were detected in the presence and absence of Hb, although HrrA∼P was at slightly higher levels in the presence of Hb (Fig. 9). Only trace levels of HrrA∼P were seen in the chrS mutant (ΔchrS), whereas high levels of HrrA∼P were observed in the hrrS deletion strain (ΔhrrS). HrrA∼P was absent in the double mutant (ΔchrS ΔhrrS). These data indicate that HrrA is phosphorylated in C. diphtheriae regardless of the presence of the Hb inducer and that both ChrS and HrrS have a strong effect on the phosphorylation status of HrrA. ChrS is required for much of the phosphorylation of HrrA, as shown by the trace amounts of HrrA∼P in the chrS mutant. Conversely, HrrS appears to function primarily as a phosphatase, since the levels of HrrA∼P increase significantly in the hrrS mutant (Fig. 9). HrrA was not phosphorylated in the strain mutated for both kinases (ΔchrS ΔhrrS), indicating that both kinases are required for the phosphorylation of HrrA and that no other HKs are involved in the phosphorylation of HrrA.

FIG 9.

In vivo phosphorylation of HrrA. The C. diphtheriae WT strain C7(−) and the C7 ΔchrS and ΔhrrS mutants were cultured in the absence (−) and presence (+) of 200 μg/ml Hb in HIBTW medium. The C7 ΔchrS ΔhrrS and ΔchrSA ΔhrrSA double mutants were cultured under with Hb. Strains were grown to mid-log phase and lysed as described in Materials and Methods. Samples were electrophoresed on a Phos-Tag acrylamide gel, and HrrA was analyzed by Western blotting using HrrA-specific antibodies. The faint signal from the ΔchrSA ΔhrrSA mutant is from the adjacent lane. Results from a representative experiment are shown.

Analysis of phosphatase activity with ChrS and HrrS.

In some bacterial two-component systems, it is not unusual for HKs to also have phosphatase activity. This enables the bacterium to have tighter control over genes that are modulated by these regulatory entities. Since the results in Fig. 9 suggest that HrrS may have phosphatase activity, an in vitro assay was developed to determine whether HrrS and ChrS are able to dephosphorylate HrrA∼P. HrrA was phosphorylated by ChrS∼P and subsequently separated from ChrS∼P using GST resin as described above (Fig. 8A). The supernatant containing only HrrA∼P was incubated with either ChrS or HrrS (HrrS-GST_Kin3) for up to 60 min, and HrrA∼P was then analyzed on a Phos-Tag gel. As shown in Fig. 10A, ChrS has no effect on the levels of HrrA∼P. However, HrrA∼P is rapidly dephosphorylated in the presence of HrrS, with no detection of HrrA∼P after 5 min (Fig. 10B). HrrA∼P showed no decrease in phosphorylation after 60 min in the absence of either ChrS or HrrS, indicating that autodephosphorylation of HrrA does not occur in vitro (not shown). Experiments to assess the phosphatase activity of HrrS and ChrS with ChrA∼P could not be performed because of the instability of the phosphate group on ChrA∼P (Fig. 8B). It appears unlikely that ChrA∼P requires phosphatase activity by either ChrS or HrrS to control phosphorylation, since the phosphate group is rapidly lost by autodephosphorylation (Fig. 8B).

FIG 10.

Phosphatase properties of ChrS and HrrS. HrrA was phosphorylated following the procedure described for Fig. 8A. HrrA∼P was mixed with equal amounts of unphosphorylated ChrS (A) or HrrS (B). The mixture was incubated at 30°C and samples were obtained at 0, 5, 10, 30, and 60 min. The phosphorylation status of HrrA was determined following electrophoresis on Phos-Tag acrylamide gels and Western blotting using anti-HrrA antibodies. Experiments were repeated multiple times with similar results. Results from representative experiments are shown.

DISCUSSION

In this study, we assessed the relative contribution of each of the TCSs to the transcriptional regulation of the hemA promoter by analyzing the effect of the cloned chrSA and hrrSA genes in a C. diphtheriae deletion mutant (C7 ΔchrSA ΔhrrSA) that harbored a hemA-lacZ reporter fusion. The findings demonstrate that both TCS systems contribute in a unique manner to the transcriptional repression of the hemA promoter. The HrrAS system strongly repressed transcription in both the presence and absence of Hb, while ChrAS repressed transcription primarily in response to Hb. This finding expands on previous genetic studies and helps clarify the function for each of these TCSs. We also showed that each of the individual cloned genes from both TCSs showed no repression of the hemA promoter, which suggests that hemA repression by ChrA and HrrA requires either HrrS or ChrS and that other unrelated HKs (or RRs) encoded by C. diphtheriae have no role in the transcriptional repression at the hemA promoter. Unfortunately, plasmid constructs carrying the cloned noncognate HK-RR pairs (ChrA/HrrS and HrrA/ChrS) were unstable when examined in C. diphtheriae, and therefore, an assessment of potential cross talk between the TCSs was not performed using cloned genes in the experiments described in Fig. 1.

EMSAs showed that both ChrA and HrrA are able to bind to the hemA promoter region, suggesting that both RRs are involved in direct regulation of hemA transcription. At the hmuO and hrtAB promoters, the binding regions for ChrA and HrrA (hmuO promoter only) are located upstream of the putative −10 promoter element, a location that is consistent for proteins acting as activators. However, at the hemA promoter, the binding region overlaps the putative −10 element, suggesting that the binding of ChrA and HrrA may limit access of the RNA polymerase to the promoter, resulting in repression of transcription. A surprising finding from this study was that both RRs bind to virtually identical sequences at both the hmuO and hemA promoter regions, which suggests that ChrA and HrrA may be in competition for binding to DNA. However, the transcriptional data suggest that the RRs have an additive effect on activation of transcription at hmuO (14) and on repression of transcription at hemA (Fig. 1), which suggests the ChrA and HrrA are not in competition but rather complement each other to optimize transcriptional regulation at these promoters. An amino sequence comparison of HrrA and ChrA shows they have almost 80% similarity in their C-terminal DNA binding regions, which may account for the shared DNA binding site. The RRs show much less amino acid similarity in their receiver domains, which may reflect the different manner in which these proteins respond to phosphorylation by ChrS and HrrS.

ChrA binding to DNA was dependent on phosphorylation in vitro, whereas HrrA was able to bind DNA in both the phosphorylated and unphosphorylated states. RRs in other systems are known to bind DNA in the absence of phosphorylation (41–43). Phosphorylation of ChrA was previously shown to be required for binding to the hmuO and hrtAB promoters, which is consistent with the predicted mechanism for activation of ChrA by the Hb inducer (15). It is proposed that in the presence of Hb, ChrS is autophosphorylated, and then it subsequently phosphorylates ChrA, enabling the RR to bind DNA. The activity of the ChrAS system at the hemA promoter is largely consistent with this proposed mechanism, although a low level of repression was observed in the absence of Hb; the reason for this activity is not known, but it may be due to a low level of phosphorylation in the absence of Hb or weak DNA binding by ChrA in the unphosphorylated state (Fig. 1). The ability of HrrSA to represses transcription of hemA in the absence of Hb (Fig. 1) suggests that HrrA binding to DNA may not require phosphorylation, which is consistent with our findings in the in vitro DNA binding studies. However, the in vivo phosphorylation studies with HrrA described in Fig. 9 indicate that HrrA is phosphorylated in the presence and absence of Hb in the wild-type strain and in the chrS and hrrS mutants but not in the hrrS chrS double mutant. This observation suggests that HrrA is phosphorylated under all conditions tested in vivo and that HrrA is phosphorylated in vivo by both HrrS and ChrS in both the presence and absence of the Hb inducer. This finding is quite surprising, since it indicates that ChrS is able to phosphorylate HrrA in the absence of Hb (as observed in the ΔhrrS strain [Fig. 9]). This function was not predicted for ChrS, based on its previously observed activity at the hmuO and hrtAB promoters, where activation of transcription by ChrA was strictly Hb responsive and dependent on ChrS. We were unable to determine whether ChrS phosphorylates ChrA in vivo, due to the inability to detect any phosphorylated form of ChrA in either the presence or absence of Hb (Fig. 7). Because we were unable to determine the phosphorylation status of ChrA in vivo, we do not know if phosphorylation of ChrA is Hb dependent and whether ChrS and/or HrrS is involved in the in vivo phosphorylation of ChrA. We believe that the inability to detect ChrA∼P in vivo is likely due to the rapid autodephosphorylation of ChrA observed in the in vitro experiments described in Fig. 8A and B. Response regulators that undergo rapid autodephosphorylation similar to that observed for ChrA have been described previously, and it is thought to be a mechanism to allow for tight control of gene expression (18, 44).

Autophosphorylation of HrrS in vitro was not observed despite testing multiple GSK-HrrS constructs that contained various sequences at the 5′ end (see Materials and Methods). While autophosphorylation of HrrS was not observed either in vitro or in vivo, the in vivo phosphorylation study in Fig. 9 suggests that HrrS possesses minimal kinase activity, since weak phosphorylation of HrrA is observed in the chrS mutant (ΔchrS). It is unclear why HrrA requires phosphorylation in vivo, since in vitro results indicate that HrrA binds DNA regardless of its phosphorylation status. It is possible that under in vivo conditions, phosphorylation of HrrA enhances binding or alters the manner in which HrrA interacts with DNA or with RNA polymerase, such that HrrA∼P is more efficient in affecting gene expression.

The information in Table 2 and Table 3 provides an overview of the findings from this study and previous studies with the ChrSA and HrrSA TCSs. Table 2 includes a summary of the transcriptional regulation data for the hmuO, hemA, and hrtAB promoters, while Table 3 provides a summary of the phosphorylation and DNA binding studies performed in the current study.

TABLE 2.

Transcriptional regulation of the hmuO, hemA, and hrtAB promoters by the HrrSA and ChrSA TCSs

TCS	Effect on transcription ata:
	hmuO promoter (Hb)	hemA promoter (Hbb)	hrtAB promoter (Hb)
ChrSA	Activation	Enhanced repression	Activation
HrrSA	Activation	Enhanced repression	No regulation

^aHb, requirement for Hb to regulate transcription. Results for hmuO are from reference 14, and hrtAB data are from reference 33.

^bHb significantly enhanced repression of hemA transcription for both TCSs.

TABLE 3.

Factors effecting phosphorylation of the components of the HrrSA and ChrSA systems

TCS protein	Phosphorylation activity
	In vitro					In vivod
	ChrSa	HrrSa	Auto-Pb	Auto-de-Pb	DNA bindingc
ChrA	Yes	No	No	Yes	Yes	Not detected
HrrA	Yes	No	No	No	No	Yes
ChrS			Yes	No		Not determined
HrrS			No	Not determined		Not detected

^aChrS and HrrS were assessed for their ability to transphosphorylate ChrA and HrrA.

^bProteins were examined for ability to autophosphorylate (auto-P) and autodephosphorylate (auto-de-P) in vitro.

^cRequirement for phosphorylation of the RR to bind DNA (determined by EMSA). ChrA and HrrA bind to identical sequences upstream of the hmuO and hemA genes.

^dIn vivo phosphorylation of ChrA was not detected, which is likely due to rapid dephosphorylation. In vivo phosphorylation of HrrA was associated with both ChrS and HrrS and occurred in the presence and absence of Hb. Phosphorylation for ChrS was not determined, due to the inability to observe ChrS in vivo because of the limitations of the Western blot system. Phosphorylation of HrrS in vivo was not detected.

Proteins homologous to the HrrSA and ChrSA systems were recently described in Corynebacterium glutamicum (45–47). In C. glutamicum, ChrSA primarily controls heme-dependent transcription of hrtAB, which affects heme tolerance, while the HrrAS system regulates transcription of hmuO, heme biosynthetic enzymes, and heme-containing proteins (45, 47). Both RRs repressed transcription of the paralogous TCS, and, as with C. diphtheriae, cross talk between these systems was also demonstrated (46). The HK components in the C. glutamicum TCSs were reported to possess both phosphatase and kinase activities (46), where the phosphatase activity in vitro was restricted to their cognate RRs, while each HK was able to phosphorylate both RRs in vitro. While there are similarities between the C. glutamicum and C. diphtheriae TCSs, the phosphorylation and phosphatase activities appear to be distinctly different and suggest a different mode of signal transduction in these related systems.

The current study expands on the previous analyses of the ChrSA and HrrSA TCSs in C. diphtheriae (14, 15). The findings here underscore the complex interactions between the ChrSA and HrrSA systems in the regulation of the hemA promoter and demonstrate how the two systems complement one another to refine and control transcription in the presence and absence of Hb.

Funding Statement

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.