Figures
Abstract
Lactate Dehydrogenase 1 (Ldh1) is a key enzyme involved in Staphylococcus aureus NO·-resistance. Full ldh1-induction requires the presence of glucose, and mutants lacking the Carbon-Catabolite Protein (CcpA) exhibit decreased ldh1 transcription and diminished Ldh1 activity. The redox-regulator Rex represses ldh1 directly by binding to Rex-sites within the ldh1 promoter (Pldh1). In the absence of Rex, neither glucose nor CcpA affect ldh1 expression implying that glucose/CcpA-mediated activation requires Rex activity. Rex-mediated repression of ldh1 depends on cellular redox status and is maximal when NADH levels are low. However, compared to WT cells, the ΔccpA mutant exhibited impaired redox balance with relatively high NADH levels, yet ldh1 was still poorly expressed. Furthermore, CcpA did not drastically alter Rex transcript levels, nor did glucose or CcpA affect the expression of other Rex-regulated genes indicating that the glucose/CcpA effect is specific for Pldh1. A putative catabolite response element (CRE) is located ∼30 bp upstream of the promoter-distal Rex-binding site in Pldh1. However, CcpA had no affinity for Pldh1 in vitro and a genomic mutation of CRE upstream of Pldh1 in S. aureus had no affect on Ldh1 expression in vivo. In contrast to WT, ΔccpA S. aureus preferentially consumes non-glycolytic carbon sources. However when grown in defined medium with glucose as the primary carbon source, ΔccpA mutants express high levels of Ldh1 compared to growth in media devoid of glucose. Thus, the actual consumption of glucose stimulates Ldh1 expression rather than direct CcpA interaction at Pldh1.
Citation: Crooke AK, Fuller JR, Obrist MW, Tomkovich SE, Vitko NP, Richardson AR (2013) CcpA-Independent Glucose Regulation of Lactate Dehydrogenase 1 in Staphylococcus aureus. PLoS ONE 8(1): e54293. https://doi.org/10.1371/journal.pone.0054293
Editor: Tarek Msadek, Institut Pasteur, France
Received: October 30, 2012; Accepted: December 10, 2012; Published: January 14, 2013
Copyright: © 2013 Crooke et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by funding from the National Institutes of Health (NIH) National Institute of Allergy and Infectious Diseases (NIAID) (AI093613 to ARR) and the Pew Charitable Trusts (Pew Scholars Program in the Biomedical Sciences). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Staphylococcus aureus is an important human pathogen causing disease ranging from mild skin and soft tissue infections to severe invasive sepsis, pneumonia, osteomyelitis and endocarditis [1]. The prevalence of multi drug-resistant strains, particularly those resistant to Methicillin (MRSA), complicates effective treatment in many cases [2]. While drug-resistance limits treatment options, the high virulence potential of S. aureus renders even immunocompetent hosts exceedingly susceptible to this pathogen. This is, in part, due to the ability of S. aureus to resist nearly every facet of host immunity. Indeed, S. aureus encodes many factors that limit the efficacy of opsonophagocytosis, cationic peptides, complement and reactive oxygen species [3]. Consistent with this, S. aureus is also remarkably resistant to nitric oxide (NO·), a lipophilic radical that serves as one of the most broad-spectrum antimicrobial effectors of the innate immune system [4], [5]. NO·-resistance distinguishes S. aureus from most other bacterial species, including closely related but less pathogenic members of the Staphylococci [6].
S. aureus NO·-resistance hinges upon the ability of this organism to metabolically adapt to the cytotoxic effects of host NO· [6]. NO· attacks many metabolic enzymes by targeting active sites composed of iron-sulfur clusters, redox-active thiols and heme motifs [7], [8]. Consequently, NO· is a potent antagonist of aerobic respiration in the host as well as in invading microbes [6], [9]. In response to NO·, S. aureus shifts into a metabolic state heavily reliant on fermentative metabolism to combat the negative effects of NO· on respiration [4], [5]. A key enzyme in this response is a lactate dehydrogenase encoded by ldh1, a S. aureus-specific allele not found in other species of Staphylococci [6]. Ldh1 catalyzes the reduction of pyruvate to L-lactate with the concomitant oxidation of NADH to NAD+ for the purpose of restoring redox balance, an essential reaction for cells unable to respire. Consistent with its role in redox-balance, ldh1 is normally repressed by the redox-sensing regulator Rex [10]. Rex binds to the ldh1 promoter (Pldh1) and limits its expression until the level of NADH rises. The binding of NADH to Rex diminishes DNA binding affinity, effectively leading to the derepression of ldh1 as well as many other redox-balancing fermentative enzymes [10], [11]. However, during both NO·-stress and anaerobic fermentation, Ldh1-catalyzed L-lactate production serves as the primary source of redox balance for S. aureus [6].
It has also been suggested that Ldh1 expression requires the presence of glucose in addition to redox imbalance [12]. Glucose-mediated induction of Ldh activity in bacteria is not uncommon. For instance in many bacteria, including members of Lactococci and Streptococci, glucose consumption results in high intracellular levels of fructose-1,6-bisphosphate (FBP) an allosteric activator of Ldh that is essential for activity [13]. The binding of FBP stabilizes tetramer formation and improves substrate binding. Many other forms of bacterial Ldh enzymes act independently of FBP and it has been suggested that these enzymes possess intrasubunit salt bridges that obviate the need for FBP allosteric interations [14]. S. aureus Ldh1 is such an enzyme and consequently does not require FBP for activity [6]. However, in Gram-positive bacteria, high FBP levels resulting from glucose consumption can lead to changes in gene expression in addition to modulating enzyme activity. FBP accumulation drives additional phosphorylation of Histidine-containing phosphocarrier protein (Hpr) on a conserved serine residue (Ser46) through activation of the kinase activity of Hpr-kinase (HprK) [15]. The resulting P∼Ser46-HPr can then directly interact with the staphylococcal catabolite control protein A, CcpA [16]. CcpA-P∼Ser46-HPr then activates the expression of glycolytic genes and represses TCA cycle and gluconeogenic gene expression allowing for maximal use of available glycolytic carbon sources [17]. In addition, CcpA mediates the effects of glucose on the expression of some S. aureus virulence factors including Protein A and FnbB [17]. Here we test whether high intracellular FBP levels activate ldh1 transcription by signaling catabolite control through CcpA in S. aureus, rather than through allosteric activation of Ldh1. Indeed, a putative catabolite response element (CRE) site, the high-affinity CcpA-P∼Ser46-HPr binding site, can be identified upstream of ldh1. However, our results show that utilization of glucose stimulates ldh1 expression but that this induction is not directly mediated by CcpA.
Materials and Methods
Bacterial Strains and Culture Conditions
Mutant strains used in this study are listed in Table 1. S. aureus was cultivated in Brain Heart Infusion medium or in chemically defined medium (PN medium) in which primary carbon sources could be modified. Briefly, PN is a phosphate-buffered medium composed of a primary carbon source, nucleobases (Adenine, 5 mg/L; Guanine, 5 mg/L; Cytosine, 5 mg/L; Uracil 5 mg/L and Thymine 20 mg/L), free amino acids (Ala, 60 mg/L; Arg, 70 mg/L; Asp, 90 mg/L; Cystine, 20 mg/L; Glu 100 mg/L; Gly, 50 mg/L; His, 30 mg/L; Iso, 30 mg/L; Leu, 90 mg/L; Lys, 70 mg/L; Met, 10 mg/L; Phe, 40 mg/L; Pro, 10 mg/L; Ser, 30 mg/L; Thr, 30 mg/L; Trp, 10 mg/L; Tyr, 50 mg/L; and Val, 80 mg/L), vitamins (thiamine, 1 mg/L; niacin, 1.2 mg/L; biotin, 5 µg/L; and pantothenate, 250 µg/L), FeCl3 at 8 mg/L, MgSO4 at 2.5 mg/L and trace elements (ZnCl, 70 µg/L; MnCl, 63 µg/L; Boric Acid, 6 µg/L; CoCl2, 190 µg/L; CuCl2, 2 µg/L; NiCl2, 13 µg/L and Na2MoO4, 31 µg/L). Antibiotic selection in S. aureus (E. coli) was performed using the following concentrations: ampicillin (100 µg/ml), chloramphenicol 20 µg/ml, kanamycin 50 µg/ml (50 µg/ml), spectinomycin 100 µg/ml (500 µg/ml), erythromycin 5 µg/ml (300 µg/ml). Growth was monitored as change in absorbance (660 nm) assessed using a Tecan infinite M200 plate reader in 200 µl cultures within a 96 well plate (100 µl headspace). Bacterial cultures used for total RNA isolation, NAD+/NADH ration determination and Ldh enzyme assays were cultivated in 50 ml of BHI medium shaking (250 rpm) in 250 ml Erlenmeyer flasks to an OD660 ∼ 1.0. For lag phase analyses, 5 ml overnight cultures (BHI) of WT and Δrex S. aureus COL were washed twice in PBS, diluted 1∶1000 in fresh defined medium and growth was monitored in 96 well plates (200 µl cultures, 100 µl headspace). Viable cfu in inocula were enumerated to ensure equal numbers of live bacteria were seeded.
Cloning, Mutant Construction and Reporter Fusions
Promoter::GFP fusions were constructed by directionally cloning PCR fragments into the EcoRI site of pCN52 (promoterless Gfp fusion vector) using Infusion® technology (Clontech). Fusions with pldh1 (fusions 1 through 4) as well as the prpoD fusion were generated by amplifying fragments from S. aureus COL genomic DNA using primers ldh1_fusion1.1A/ldh1_fusions.1B, ldh1_fusion2.1A/ldh1_fusions.1B, ldh1_fusion3.1A/ldh1_fusions.1B, ldh1_fusion4.1A/ldh1_fusions.1B and rpoD_gfp.1A/rpoD_gfp.1B, yielding pJF115, pJF116, pJF117, pJF118 and pJF120, respectively. Promoter fusions for prpoD and pldh1 (full length promoter fragment) were also generated in pJF119 in which the ApaI/XhoI fragment of pCN52 harboring the ermC gene was replaced with a fragment encoding chloramphenicol acetyltransferase (CAT) yielding pJF120 and pJF122, respectively. CRE* was generated by amplifying S. aureus COL genomic DNA with primers hmp_ldh1.5′/CRE_mut.1B and CRE_mut.1A/hmp_ldh.3′. The two resulting amplimers were combined using overlapping PCR using primers hmp_ldh.5′/hmp-ldh.3′. The resulting fragment was cloned into pCR BluntII Topo® (Invitrogen) from which the EcoRI fragment harboring CRE* was moved into pBT2ts. Allelic replacement was performed as described previously [18]. Δrex and ΔccpA were generated by cloning 5′- and 3′-homology regions for each gene on either side of a KmR (rex) or SpR (ccpA) cassette in the pBTK and pBTS yielding pJF102 and pJF103, respectively. Mutants were made using an allelic replacement scheme previously described [18].
Quantitative Reverse Transcriptase Real Time PCR (Q RT-PCR)
Cells were grown to OD660 = 0.5 and either treated with NO· (2 mM DEA-NO, AG Scientific) for 15 minutes or left untreated. Twenty-five ml of culture was added to an equal volume of ice cold ethanol:acetone (1∶1) and incubated at −80°C until further use. All frozen cell suspensions were thawed at room temperature, pelleted by centrifugation and resuspended in 500 µl TE for mechanical disruption using Lysing Matrix B (MP Biomedicals, Solon, OH) in a standard cell disruptor. One-hundred µl of lysates were used for RNA isolation using an RNAEasy® Mini Kit (Qiagen, Valencia, CA) per manufacturer instructions. RNA was spectrophotometrically quantified and 50 ng of total RNA analyzed per reaction using the Sensimix™ SYBR & Fluorescein One-Step kit (Bioline). Reaction conditions were as specified by Bioline and performed on a MyIQ Single color Real-Time PCR Detection System (BioRad). Primers used for analysis are listed in Table 1. All transcript levels were normalized to those of rpoD, which exhibited little variation across the growth conditions described here. Transcriptional start site for pldh1 was tested using RT-PCR by amplifying either cDNA or genomic DNA with primers ldh1_RT−136.1A/ldh1_UTR_RT.1B and ldh1_RT−176.1A/ldh1_UTR_RT.1B (Table 1). Statistical significance was determined using Student’s t-test (2-tailed).
Electromobility Shift Assays (EMSAs)
CcpA and Rex were purified as N’-terminal and C’-terminal His6 tagged fusions, respectively. CcpA was amplified from S. aureus COL using primers ccpA.1A/.1B (Table 1) and cloned into the BamHI/KpnI sites of pQE-30 (Qiagen) creating plasmid pMWO-100. Rex was amplified from S. aureus COL using primers rex.1A/.1B (Table 1) and cloned into the NdeI/NotI sites of pET-24b (Novagen) to generate pMWO-101. pMWO-100 was maintained in E. coli XL-1 Blue with appropriate selection whereas pMWO-101 was transformed into E. coli BL21 and maintained with appropriate selection. Cultures were inoculated into 500 mL LB medium and cultivated at 37°C until OD600 = 0.4, at such time cultures were induced with 0.5 mM IPTG shaking at 37°C for 5 h. Cells were then pelleted, resuspended in binding buffer (20 mM Tris, pH 8.0, 500 mM NaCL, 10% glycerol, 5 mM imidazole) then lysed via sonication. Cell debris was pelleted via centrifugation and the culture supernatants applied to an equilibrated Ni-NTA column (Qiagen). The column was then washed and eluted in binding buffer with increasing concentrations of imidazole up to 1 M. Eluates were combined and passed over a Superdex 75 gel filtration column (GE Healthcare) for further purification and buffer exchange against 10 mM Tris-HCl, pH 7.5, 100 mM NaCl. Protein was concentrated using a Microcon centrifugal concentrator (Millipore) and glycerol was added to 10% final concentration prior to flash freezing and storage at −80°C. Final protein concentrations were then determined using a Bradford Assay (BioRad). Protein was incubated at room temperature with 250 fmol of both specific and non-specific probe DNA at varying molar ratios for 15 minutes prior to loading. CcpA::DNA mixtures were loaded onto a pre-run (1 h at 75 V) 8.0% Polyacrylamide gel (1X TBE) and run ice-cooled at 125 V. Rex::DNA mixtures were similarly run through 7.0% polyacrylamide gels. Gels were stained 1∶10,000 with GelRed (Biotum) for 5 min, destained with deionized H2O and then imaged via UV-exposure on a GelDoc system (BioRad).
Ldh1 Enzyme Activities
Cultures of S. aureus Δldh2 strains were grown in 1 L volumes in BHI medium to OD660 = 1.0 then aliquoted into 250 ml sealed Sorvall centrifuge tubes and incubated at 37°C for 2 h. Cells were pelleted, disrupted mechanically using Lysing Matrix B (MP Biomedicals, Solon, OH) and then cell debris was removed via centrifugation. Protein concentration for each lysate was determined using a BCA method (Pierce, Rockford, IL). Ldh1 reaction mixtures contained 1 mg total protein from cell-free extracts, 100 mM Tris·Cl pH8.5, and 3 mM NAD+. Reactions were initiated by the addition of 13.9 mM L-lactate. Ldh1 activity was defined as the L-lactate-dependent reduction of NAD+, which was monitored spectrophotometrically at 340 nm (mM ξ340 for NADH = 6.2). Significance was determined using Student’s t-test (2-tailed).
Metabolite Analyses
Extracellular ammonia and glucose levels in culture supernatants were determined enzymatically using commercially available kits (R-Biopharm). Culture samples were heat inactivated at 55°C for 5 min, cells were pelleted and supernatants were used for analyses. NAD+/NADH ratios were also determined using commercially available kits (BioVision). Cells were cultured in PN defined medium with glucose and Casamino Acid carbon sources (0.5% each). Aerated cultures (50 ml in 250 ml Erlenmeyer flasks at 250 rpm) were grown to an OD660 = 1.0 then immediately filtered through a 0.45 µm nylon filter (GE Health). Alternatively, NO·-treated cultures were exposed to 2 mM DETA-NO for 20 minutes prior to filtering. Cells were harvested from the nylon filter and resuspended in 600 µl of cold NAD Extraction Buffer (BioVision), flash frozen in a dry ice/EtOH bath, thawed and then mechanically disrupted via lysing matrix B (MP Biomedicals). Cell debris was removed via centrifugation and supernatants were assayed for NAD+ and NADH levels as per manufacturer instructions (BioVision). Statistical significance was determined using Student’s t-test (2-tailed).
Results
NO·-mediated Induction of Ldh1 Activity Relies on the Presence of Glucose
Exposing S. aureus strain COL to NO· in a chemically defined medium with casamino acids as the primary carbon source did not result in detectible induction of a Pldh1::GFP fusion (Figure 1A). However, supplementation of this defined medium with 0.5% glucose restored robust Pldh1::GFP expression upon NO·-exposure. In contrast, the stimulatory effect of glucose was not observed in the isogenic ΔccpA mutant (Figure 1A). Using quantitative real-time PCR to assess transcript levels from chromosomal ldh1, a >3-log induction in ldh1 transcript levels upon NO·-stimulation in the presence of glucose was observed relative to rpoD (Figure 1B). This induction was diminished by more than 13-fold when S. aureus COL was exposed to NO· in media devoid of glucose. Again, similar to what was observed with the GFP promoter fusions, the positive effect of glucose on ldh1 transcript levels was not apparent in the ΔccpA mutant (Figure 1B). It should be noted, that in aerobic environments in the absence of NO·, ldh1 transcript is barely detectible resulting in large apparent induction ratios upon NO·-exposure. Thus, in the absence of glucose and/or CcpA, ldh1 transcript levels were still effectively induced ∼200-fold by NO· (Figure 1B). However, in comparison to WT cells grown in the presence of glucose, the >1-log reduction in overall ldh1 transcript in the absence of glucose and/or CcpA correlates with the absence of detectible GFP signal on a multi-copy promoter fusion (Figure 1A). Likewise, measuring Ldh1 enzyme activity revealed that without CcpA, Ldh1 activity is barely above background implying that without glucose stimulation, S. aureus effectively lacks Ldh1 activity despite detectible increases in transcript (Figure 1C).
A. WT S. aureus strain Newman and an isogenic ΔccpA mutant harboring Pldh1::GFP promoter fusions were grown in chemically defined medium with 0.5% casamino acids as primary carbon/energy sources. Glucose (0.5%) was added when indicated. Once cultures reached early exponential phase, NO· was administered (1 mM DETA/NO) and fluorescence and optical density were monitored for two hours. B. Quantitative Real-Time Reverse Transcriptase PCR (Q RT-PCR) was used to determine ldh1 transcript levels relative to rpoD in WT S. aureus strain COL and an isogenic isogenic ΔccpA mutant grown in media as described in Figure 1A. NO· was administerd as 2 mM DEA-NO. C. Ldh1 enzyme activity from cell extracts of WT and isogenic ΔccpA S. aureus strain COL lacking ldh2. Cells were cultured in BHI and stimulated with NO·(2 mM DEA-NO) 15 minutes prior to obtaining lysates.
Glucose-dependent Stimulation of ldh1 Requires both CcpA and Rex
In response to NO·, the high level induction of ldh1 transcript in the presence of glucose was identical to the basal ldh1 transcript level in the Δrex mutant (Figure 2A). Furthermore, exposure to NO· did not affect ldh1 transcript levels in the Δrex background. Additionally, inactivation of ccpA in the Δrex background did not attenuate the constitutive expression of ldh1 (Figure 1B). These data imply that CcpA does not activate ldh1 independently of Rex-mediated repression. Thus, glucose-dependent induction of ldh1 relies on the presence of both Rex and CcpA, but CcpA must affect the ability of Rex to repress ldh1 transcription.
A. Q RT-PCR of ldh1 transcript from S. aureus strain COL normalized to rpoD in cells exposed/unexposed to NO· administered as 2 mM DEA-NO 15 minutes prior to RNA isolation. B. ElectroMobility Shift Assay (EMSA) of Pldh1 using purified His-Rex at increasing molar ratios of Rex:DNA (250 fmol promoter DNA in all wells). Internal fragment of hmp was used as a non-specific probe (N.S. band). As predicted by the presence of two Rex sites in Pldh1, two independent shifted bands appear with increasing Rex::DNA molar ratios. C. Schematic representation of Pldh1 and fragments used for GFP:fusions and their relative activity following stimulation with NO· (2 mM DETA-NO). The activity of Fusion 4 was indistinguishable form that of a promoterless control. D. Reverse transcriptase PCR using two different forward primers depicted in Figure 1C to amplify products using both cDNA (cD) and genomic (G) DNA as templates.
Two putative Rex binding sites (TGTGAWWWWWWTCACA) can be identified −160 bp and −218 bp upstream of the ldh1 start-codon. Purified Rex bound specifically to both sites in vitro consistent with the role of Rex in ldh1 repression (Figure 2B). Interestingly, a similar consensus sequence has been proposed for the iron-sulfur cluster containing oxygen sensor ArcR [19]. However, deletion of ArcR had no affect on ldh1 expression under anaerobiosis (data not shown, [10]). Furthermore, the ΔarcR mutant exhibited no alterations in the expression of other Rex-regulated genes including pflA, adhE or ddh implying that the two regulators do not share identical binding sites and that Rex, not ArcR, controls ldh1 expression.
The fact that the putative Rex binding sites were between 160 and 218 bp upstream of the ldh1 coding sequence prompted us to map the transcriptional start site of ldh1 using 5′-RACE. However, given the apparent secondary structure predicted to occur within the 5′UTR of the ldh1 transcript, 5′RACE was unsuccessful at identifying the start of ldh1 transcription (Figure 2C). By using a combination of in silico prediction models combined with truncated GFP-promoter fusions and Reverse-Transcriptase PCR we established that the start of ldh1 transcription was likely within the proximal Rex binding site (Figure 2C). That is, a reverse oriented primer within ldh1 was able to amplify a product (later verified by sequencing) with a forward-oriented primer positioned at −136 from ldh1 using both genomic and cDNA templates (Figure 2D). In contrast, using a forward-oriented primer positioned at −176 from the ldh1 ATG was unable to amplify a product specifically from cDNA. Furthermore, Pldh1::GFP promoter fusions were highly active until the cloned fragment was truncated past −150 relative to ldh1 ATG (Figure 2C). Thus, the 26 bp fragment from −150 to −176 likely contains the ldh1 transcriptional start site. A putative −10 and −35 sequence could be identified upstream of this region putting the ldh1 transcriptional start site at the guanine residue −167 bp upstream of the ldh1 start codon. Consistent with Rex-mediated ldh1 repression, this start site lies within the TGTGA inverted repeat of the ldh1-proximal Rex binding site.
The CcpA-dependent Glucose Effect on ldh1 Expression does not Involve Altered Rex Activity
The fact that the presence of glucose stimulates ldh1 transcription in a manner dependent on both Rex and CcpA suggests that CcpA might affect Rex activity thereby modulating ldh1 expression. Indeed, rex transcript levels were a modest 50% increased in the ΔccpA background (Figure 3A). Higher Rex levels could be predictive of more ldh1-repression as seen in Figure 1B. However, this modest increase in rex transcript did not result in reduced expression of other Rex-regulated genes including ddh, ald1 and adhE. Alternatively, the ΔccpA mutant may exhibit less redox-imbalance under NO·-stress thereby increasing the repressive effects of Rex. That is, Rex activity should be more pronounced in conditions of redox balance with relatively low NADH levels (e.g. high NAD+/NADH ratios). However, upon exposure to NO· the ΔccpA mutant exhibited similar redox imbalance compared to WT (same NAD+/NADH ratios) (Figure 3C). In unexposed cells, ΔccpA mutants actually exhibited enhanced redox imbalance (lower NAD+/NADH ratios). This should act to further relieve Rex-mediated repression, however Ldh1 activity is diminished in a Rex-dependent fashion in this strain (Figures 1C and 2A). Thus, CcpA must affect the ability of Rex to repress Ldh1 in a manner that does not alter Rex protein levels or activity. One hypothesis is that CcpA binding to the CRE site upstream of the distal Rex binding site reduces the ability of Rex to bind DNA and therefore repress ldh1 transcription.
A. Q RT-PCR analyses of rex transcript levels normalized to those of rpoD in WT versus ΔccpA S. aureus strain COL following NO· stimulation. B. Q RT-PCR analyses of other Rex-regulated genes normalized to rpoD upon NO·-stimulation. Only ldh1 exhibits CcpA-dependent activation. C. Redox status depicted as NAD+/NADH ratios of WT versus ΔccpA S. aureus COL prior to and after stimulation with NO·.
CcpA does not Interact Directly with the ldh1 Promoter but rather Acts Indirectly by Promoting Glucose Utilization
We sought to show that CcpA binding to the CRE site near Pldh1 could occlude Rex from the promoter and relieve ldh1 repression. However, purified CcpA showed no affinity for Pldh1 over that of non-specific DNA probe (Figure 4A). In contrast, purified CcpA was able to shift the rocD2 promoter, a DNA fragment know to be directly bound by CcpA [16]. The lack of CcpA binding to Pldh1 was surprising given that the putative CRE site upstream of ldh1 was completely consistent with the proposed Rex-consensus sequence from B. subtilis (Figure 4B). CcpA normally binds with higher affinity to CRE sites when complexed with P∼Ser46-HPr, so an attempt to show in vivo CcpA interactions with the CRE of Pldh1 was undertaken. Despite altering seven key base pairs of the CRE consensus (CRE*) in the genomic Pldh1, there was no measurable effect of CRE* on ldh1 transcription (Figure 4C). Similarly, the chromosomal CRE* mutation had no affect on Ldh1 enzyme activity in S. aureus COL (data not shown). Thus, CcpA does not directly bind to the predicted CRE site of Pldh1 in vitro and the CRE site upstream of ldh1 exerts no measurable effects on Ldh1 expression in vivo.
A. EMSA with His-tagged CcpA using Pldh1 (LEFT) or ProcD2 (RIGHT) as probes and an internal hmp fragment as a non-specific probe (N.S. band). Only at highest ratios of CcpA::DNA did non-specific shifted bands become evident using Pldh1 as a probe (white arrows). 250 fmol of DNA probes were used in all wells. B. Alignment of CRE from Pldh1, ProcD2, the B. subtilis consensus sequence and the mutated CRE*. C. Q RT-PCR analyses of ldh1 transcript levels normalized to those of rpoD in WT, ΔccpA and CRE* derivatives of S. aureus strain COL following 15 min. NO· exposure (2 mM DEA-NO).
Together, these data suggest that the effect of CcpA on ldh1 expression is indirect. We therefore wanted to determine whether the glucose-mediated activation of Ldh1 was dependent on CcpA at all (Figure 1). Since ΔccpA mutants grow poorly in chemically defined medium with glucose alone as the primary carbon source, throughout this study WT and ΔccpA strains were grown in media with casamino acids as carbon sources supplemented with or without 0.5% glucose (Figure S2). In media supplemented with both casamino acids and glucose, WT cells preferentially consume glucose with little ammonia production indicating limited consumption of amino acids (Figure 5A). In contrast, ΔccpA mutants actively produced ammonia and consumed less glucose compared to WT. Thus, a ΔccpA mutant primarily consumes amino acids in this defined medium whereas WT preferentially performs glycolysis. When “forcing” the ΔccpA mutant to utilize glucose by growing in defined medium with no casamino acids, ldh1 was expressed eight-fold higher compared to growth on casamino acids alone (Figure 5B). This increased expression was evident despite poor growth rates of ΔccpA on glucose (Figure S2). These results indicate that utilization of glucose promotes Ldh1 expression independently of CcpA. Furthermore, the deficiency in glucose utilization observed in the ΔccpA mutant can explain the apparent dependency of Ldh1 expression on CcpA.
A. Metabolite analyses of WT S. aureus COL and an isogenic ΔccpA mutant grown in defined medium devoid of ammonia with both glucose (0.5%) and casamino acids (0.5%) as carbon sources. Glucose utilization is delayed and ammonia production is enhanced in the ΔccpA mutant implying altered carbon source preference. B. Q RT-PCR analyses of ldh1 transcript levels normalized to those of rpoD in ΔccpA S. aureus COL following 15 min NO·-exposure (2 mM DEA-NO) cultivated in defined medium with indicated carbon source. C. Representative growth curve demonstrating the defect of a Δrex mutant compared to isogenic WT S. aureus COL when utilizing gluconeogenic carbon sources. Inset: average time to ½ maximum optical density as a metric for length of lag phase from four independent curves.
The Glucose-requirement for Maximal Ldh1 Activation Prevents Carbon Loss during Gluconeogenesis
Lactate production via Ldh1 provides redox balance by re-oxidizing NADH produced during glycolytic conversion of carbohydrates to pyruvate. However, when carbohydrates are scarce, S. aureus can utilize amino acids, lactate, pyruvate and other gluconeogenic substrates for carbon and/or energy. Under these conditions, high level Ldh activity would be detrimental since pyruvate pools would need to be converted to phosphoenopyruvate (PEP) via oxaloacetate for gluconeogenesis. High Ldh activity would effectively compete for available pyruvate, diminishing flux through gluconeogenesis. The Δrex mutant expresses high Ldh1 activity irrespective of the presence/absence of glucose (Figure 2A). Likewise, Δrex mutants have difficulty growing in media with pyruvate or casamino acids as a sole carbon/energy sources (Figure 5C). That is, Δrex mutants exhibit an extended lag phase (∼2.5 h) only on gluconeogenic carbon sources such as amino acids and pyruvate. Thus, the conserved dependence on glucose catabolism in bacteria for maximal Ldh activity, either transcriptionally as in S. aureus or posttranscriptionally as in many other bacteria, represents a fundamental control mechanism ensuring efficient carbon utilization in the absence of abundant carbohydrates.
Discussion
The ability to resist NO·-mediated toxicity separates S. aureus from most other bacteria including coagulase-negative staphylococcal species [6]. A major difference in the S. aureus NO·-response compared to other bacteria is the metabolic adaptations mounted by this pathogen against nitrosative stress [4], [5]. Glucose has been shown to be essential for S. aureus NO·-resistance [18]. There are several possible explanations as to why glycolysis is the primary central metabolic pathway used by S. aureus under NO·-stress. For instance, NO· may target key gluconeogenic enzymes limiting use of non-glycolytic carbon/energy sources. Alternatively, even with a full repertoire of active non-glycolytic pathways, gluconeogenic metabolism may simply be incompatible with the effects of NO· (e.g. elicits excessive redox imbalance). On the other hand, glycolysis may be required for the induction of key metabolic enzymes in S. aureus during growth in the presence of NO·. Here we demonstrate that, at least in the case of ldh1, glucose stimulates its expression and is required for full Ldh1 activity. Given that Ldh1 is the primary redox-balancing enzyme for S. aureus during periods of diminished respiratory activity, reduced ldh1 transcription in the absence of glucose may partially account for inability of S. aureus to thrive under NO·-stress on gluconeogenic carbon/energy sources.
S. aureus acquired ldh1 after emerging evolutionarily from other staphylococci given that the allele can only be found in S. aureus genomes where it seems to be universally present. Transcript levels of ldh1 are kept virtually undetectable in respiring cells by the direct binding of the Rex repressor to two Rex-sites within Pldh1 (Figure 2). Thus, Ldh1 activity is only detected in redox-stressed cells (increased NADH levels). Here we show, as others have suggested [12], that maximal Ldh1 activity also relies on the presence of glucose. It is interesting that of all the tested Rex-regulated genes, only ldh1 appears to be under glucose/CcpA control (Figure 1B). This creates a scenario whereby in the presence of glucose, S. aureus preferentially balances redox by the production of L-lactate. This is not the case in other staphylococcal species, which produce commensurate levels of D- and L-lactate anaerobically [6]. The “enantiomer preference” of S. aureus may stem from the fact that L-lactate alone can be utilized by lactate-quinone oxidoreductase (Lqo) [18]. Thus, L-lactate may represent more of a metabolic intermediate than D-lactate, which is primarily a metabolic endproduct. However, the true selective advantage of L-lactate production over D-lactate in S. aureus is still unknown.
The fact that glucose-stimulation of Ldh1 expression required the presence of Rex implies that glucose somehow modulates that ability of Rex to repress ldh1 transcription. While the ΔccpA mutant produced marginally more rex transcript (∼50% increase over WT), the impaired redox balance of the ΔccpA mutant (high NADH levels) would predict that the excess Rex would be less active in these cells (Figure 3). However, Ldh1 activity is barely detectible in non-glucose grown cells or ΔccpA mutants grown even in the presence of glucose. Furthermore, given that the entire Rex regulon did not exhibit glucose-dependent induction implies that the glucose affect on Rex is specific for pldh1. This does not exclude the possibility that the Rex has higher affinity for sites at the promoters for ddh, adhE, and ald1 and therefore the glucose effect is masked. However, the higher NADH levels in cells cultured on amino acid carbon sources predict that all Rex regulated genes should be derepressed on amino acids. Indeed, ald1 did exhibit a significant >2-fold induction on amino acid media compared to glucose consistent with NADH-mediated Rex inactivation (Figure 3B). The fact that ldh1 expression behaves in an opposing trend (lower expression on amino acids leading to higher NADH-levels) implies a separate form of regulation. The requirement for Rex to observe glucose-stimulated induction also eliminates the role for a “traditional” activator driving ldh1 transcription since that would predict a drop in Ldh1 expression in the Δrex mutant grown in the absence of glucose. Rather, the data suggest that a glucose-responsive regulator is acting as an anti-repressor limiting the ability of Rex to block ldh1 transcription (Figure S1). A CRE consensus site ∼30 bp upstream of the promoter distal Rex site seemed to be in a prime position to allow CcpA to serve as an effective anti-repressor against Rex. However, despite the complete conservation of the Pldh1 CRE with the published consensus from B. subtilis [20], CcpA has no affinity for Pldh1 either in vitro or in vivo (Figure 4). This may result from the presence of a T-A basepair at position 7 of the pldh1 CRE (Figure 4). This position is never occupied by a T in high-affinity CcpA binding sites and only is present in only 3% of low-affinity sites in B. subtilis [20]. The pldh1 CRE is completely conserved among all available sequences (data not shown). Alternatively, the CRE consensus in S. aureus may be significantly divergent from that of Bacillus spp. More investigation into the sequences of the S. aureus CRE required for CcpA binding will explain these curious results.
The fact that the ΔccpA mutant exhibits altered carbon source utilization preferentially oxidizing amino acids over glucose (Figure 5), and the indirect reduction of ldh1 transcription in the ΔccpA background suggest that Pldh1 actually responds to the performance of glycolysis (Figure S1). This glycolysis-stimulated Ldh1 expression theory is further supported by the enhanced transcription of ldh1 in the ΔccpA mutant grown in glucose relative to amino acids despite the poor growth of ΔccpA on glucose alone (Figures 5 and S2). Carbohydrate utilization would result in reduced pH whereas peptide catabolism would raise local pH given the excessive ammonia production. Perhaps Pldh1 responds to a drop in intracellular pH thereby being indirectly affected by glycolysis. Alternatively, organic acid production by glycolytic fermentation may trigger ldh1 transcription. CidR is a regulator known to respond to the presence of very short-chain organic acids (e.g. acetate and lactate), however the ΔcidR mutant exhibited no reported alterations in ldh1 transcription [21]. Thus the mechanism behind enhanced Ldh1 expression in the presence of glucose is still unknown. However, whatever the mechanism, glucose stimulated ldh1 transcription in S. aureus will mimic species that express fructose bisphosphate activated Ldh enzymes in that maximal activity will only be present in cells that are actively performing glycolysis. This level of regulation is necessary to conserve carbon during growth in the absence of glucose (Figure 5). Thus it would seem evolutionarily advantageous to limit high Ldh activity when grow under low carbohydrate conditions. Many bacteria have evolved allosteric control of Ldh to achieve this regulation [13]. S. aureus appears to have evolved to transcriptionally modulate Ldh activity to achieve the same level of control.
Supporting Information
Figure S1.
Model of S. aureus glucose-dependent ldh1 regulation. NO· blocks respiration leading to a buildup of NADH (low NAD+/NADH ratios), which diminishes Rex DNA binding activity leading to derepression of ldh1. The presence of glucose also diminishes the repressive activity of Rex by an unidentified mechanism. CcpA acts to direct S. aureus to preferentially utilize glycolytic carbon sources therefor maximizing the glycolytic effect on Rex-repression. Utilization of glycolytic carbon sources leads to increased steady-state levels of fructose 1,6,-bisphosphate (FBP), which signals the phosphorylation of HPr on a conserved Ser residue. HPr-PO4 acts as a co-activator with CcpA.
https://doi.org/10.1371/journal.pone.0054293.s001
(TIF)
Figure S2.
Growth defect of ΔccpA S. aureus COL when “forced” to use glucose as a primary carbon/energy source. Bacteria were cultivated in chemically defined medium with either 0.5% glucose (Glc), 0.5% casamino acid (a.a.) or the combination (both at 0.5%, Glc/a.a.) as primary carbon/energy sources.
https://doi.org/10.1371/journal.pone.0054293.s002
(TIF)
Author Contributions
Conceived and designed the experiments: ARR AKC JRF MWO. Performed the experiments: AKC JRF SET NPV MWO. Analyzed the data: ARR JRF AKC MWO. Wrote the paper: ARR.
References
- 1. Lowy FD (1998) Staphylococcus aureus infections. N Engl J Med 339: 520–532 .
- 2. Lambert M (2011) IDSA guidelines on the treatment of MRSA infections in adults and children. Am Fam Physician 84: 455–463.
- 3. Foster TJ (2005) Immune evasion by staphylococci. Nat Rev Microbiol 3: 948–958 .
- 4. Richardson AR, Dunman PM, Fang FC (2006) The nitrosative stress response of Staphylococcus aureus is required for resistance to innate immunity. Mol Microbiol 61: 927–939 .
- 5. Hochgräfe F, Wolf C, Fuchs S, Liebeke M, Lalk M, et al. (2008) Nitric oxide stress induces different responses but mediates comparable protein thiol protection in Bacillus subtilis and Staphylococcus aureus. J Bacteriol 190: 4997–5008 .
- 6. Richardson AR, Libby SJ, Fang FC (2008) A nitric oxide-inducible lactate dehydrogenase enables Staphylococcus aureus to resist innate immunity. Science 319: 1672–1676 .
- 7. Richardson AR, Payne EC, Younger N, Karlinsey JE, Thomas VC, et al. (2011) Multiple targets of nitric oxide in the tricarboxylic acid cycle of Salmonella enterica serovar typhimurium. Cell Host Microbe 10: 33–43 .
- 8. Toledo JC, Augusto O (2012) Connecting the chemical and biological properties of nitric oxide. Chem Res Toxicol 25: 975–989 .
- 9. Brown GC, McBride AG, Fox EJ, McNaught KS, Borutaite V (1997) Nitric oxide and oxygen metabolism. Biochem Soc Trans 25: 901–904.
- 10. Pagels M, Fuchs S, Pané-Farré J, Kohler C, Menschner L, et al. (2010) Redox sensing by a Rex-family repressor is involved in the regulation of anaerobic gene expression in Staphylococcus aureus. Mol Microbiol 76: 1142–1161 .
- 11. Wang E, Bauer MC, Rogstam A, Linse S, Logan DT, et al. (2008) Structure and functional properties of the Bacillus subtilis transcriptional repressor Rex. Mol Microbiol 69: 466–478 .
- 12. Seidl K, Stucki M, Ruegg M, Goerke C, Wolz C, et al. (2006) Staphylococcus aureus CcpA affects virulence determinant production and antibiotic resistance. Antimicrob Agents Chemother 50: 1183–1194 .
- 13. Garvie EI (1980) Bacterial lactate dehydrogenases. Microbiol Rev 44: 106–139.
- 14. Arai K, Ichikawa J, Nonaka S, Miyanaga A, Uchikoba H, et al. (2011) A molecular design that stabilizes active state in bacterial allosteric L-lactate dehydrogenases. J Biochem 150: 579–591 .
- 15. Titgemeyer F, Hillen W (2002) Global control of sugar metabolism: a gram-positive solution. Antonie Van Leeuwenhoek 82: 59–71.
- 16. Li C, Sun F, Cho H, Yelavarthi V, Sohn C, et al. (2010) CcpA mediates proline auxotrophy and is required for Staphylococcus aureus pathogenesis. J Bacteriol 192: 3883–3892 .
- 17. Seidl K, Müller S, Francois P, Kriebitzsch C, Schrenzel J, et al. (2009) Effect of a glucose impulse on the CcpA regulon in Staphylococcus aureus. BMC Microbiol 9: 95 .
- 18. Fuller JR, Vitko NP, Perkowski EF, Scott E, Khatri D, et al. (2011) Identification of a Lactate-Quinone Oxidoreductase in Staphylococcus aureus that is Essential for Virulence. Front Cell Infect Microbiol 1: 19 .
- 19. Makhlin J, Kofman T, Borovok I, Kohler C, Engelmann S, et al. (2007) Staphylococcus aureus ArcR controls expression of the arginine deiminase operon. J Bacteriol 189: 5976–5986 .
- 20. Marciniak BC, Pabijaniak M, de Jong A, Dűhring R, Seidel G, et al. (2012) High- and low-affinity cre boxes for CcpA binding in Bacillus subtilis revealed by genome-wide analysis. BMC Genomics 13: 401 .
- 21. Yang S-J, Dunman PM, Projan SJ, Bayles KW (2006) Characterization of the Staphylococcus aureus CidR regulon: elucidation of a novel role for acetoin metabolism in cell death and lysis. Mol Microbiol 60: 458–468 .
- 22. Kreiswirth BN, Löfdahl S, Betley MJ, O'Reilly M, Schlievert PM, et al. (1983) The toxic shock syndrome exotoxin structural gene is not detectably transmitted by a prophage. Nature 305: 709–712.
- 23. Duthie ES, Lorenz LL (1952) Staphylococcal coagulase; mode of action and antigenicity. J Gen Microbiol 6: 95–107.
- 24. Dyke KG (1969) Penicillinase production and intrinsic resistance to penicillins in methicillin-resistant cultures of Staphylococcus aureus. J Med Microbiol 2: 261–278.
- 25. Sabath LD, Wallace SJ, Gerstein DA (1972) Suppression of intrinsic resistance to methicillin and other penicillins in Staphylococcus aureus. Antimicrob Agents Chemother 2: 350–355.
- 26. Brückner R (1997) Gene replacement in Staphylococcus carnosus and Staphylococcus xylosus. FEMS Microbiol Lett 151: 1–8.