Factors impacting the regulation of nos gene expression in Staphylococcus aureus

ABSTRACT Staphylococcus aureus nitric oxide synthase (saNOS) contributes to oxidative stress resistance, antibiotic tolerance, virulence, and modulation of aerobic and nitrate-based cellular respiration. Despite its involvement in these essential processes, the genetic regulation of nos expression has not been well characterized. 5′ rapid amplification of cDNA ends on nos RNA isolated from S. aureus UAMS-1 (USA200 strain) and AH1263 (USA300 strain) revealed that the nos transcriptional start site mapped to an adenine nucleotide in the predicted Shine-Dalgarno site located 11 bp upstream of the nos ATG start codon, suggesting that the nos transcript may have a leaderless organization or may be subject to processing. The SrrAB two-component system (TCS) was previously identified as a positive regulator of nos RNA levels, and experiments using a β-galactosidase reporter plasmid confirmed that SrrAB is a positive regulator of nos promoter activity. In addition, the quorum-sensing system Agr was identified as a negative regulator of low-oxygen nos expression in UAMS-1, with activity epistatic to SrrAB. Involvement of Agr was strain dependent, as nos expression remained unchanged in an AH1263 agr mutant, which has higher Agr activity compared to UAMS-1. Furthermore, nos promoter activity and RNA levels were significantly stronger in AH1263 relative to UAMS-1 during late-exponential low-oxygen growth, when nos expression is maximal. Global regulators Rex and MgrA were also implicated as negative regulators of low-oxygen nos promoter activity in UAMS-1. Collectively, these results provide new insight into factors that control nos expression. IMPORTANCE Bacterial nitric oxide synthase (bNOS) has recently emerged in several species as a key player in resistance to stresses commonly encountered during infection. Although Staphylococcus aureus (sa)NOS has been suggested to be a promising drug target in S. aureus, an obstacle to this in practice is the existence of mammalian NOS, whose oxygenase domain is like bacterial NOS. Increased understanding of the nos regulatory network in S. aureus could allow targeting of saNOS through its regulators, bypassing the issue of also inhibiting mammalian NOS. Furthermore, the observed strain-dependent differences in S. aureus nos regulation presented in this study reinforce the importance of studying bacterial NOS regulation and function at both the strain and species levels.

year, and MRSA is recognized as a leading cause of hospital-acquired infections (1).A primary contributor to the success of S. aureus as a pathogen is its flexible and adaptable metabolism and virulence, as this bacterium is capable of fluctuating between multiple forms of respiration and fermentation and implementing a wide variety of virulence machinery.
S. aureus nitric oxide synthase (saNOS) has recently emerged as a key player in aerobic respiration as well as the switch between aerobic and anaerobic metabolism (2)(3)(4)(5).NOS produces nitric oxide (NO) via the two-step oxidation of L-arginine to L-citrulline and NO (6).While mammalian NOS contains a reductase domain that transfers electrons from NADPH to the oxygenase domain where this reaction occurs (7), bacterial NOS contains only the oxygenase domain and is thought to promiscuously use available reductase partners (8).In mammals, NO is an important cellular signaling molecule involved in nervous system signaling (9,10), regulating blood pressure (11,12), and mitochondrial respiration (13,14), as well as the immune response against pathogens as a cytotoxic agent (15).Bacterial NOS is found in some Gram-positive genera such as Staphylococcus, Bacillus, Streptomyces, and Deinococcus, with certain functions of this enzyme appearing to differ by genus (16).S. aureus NOS has been implicated in oxidative stress resistance (17)(18)(19), antibiotic tolerance (19,20), virulence (18,19,21), heme stress resistance (22), the switch to nitrate-based respiration during low-oxygen growth (5), and aerobic respiration (2)(3)(4)).An S. aureus UAMS-1 nos mutant demonstrated altered aerobic respiration, including increased membrane potential, increased respiratory dehydrogen ase activity, and increased accumulation of reactive oxygen species (ROS) (2).This points to a role for saNOS in the modulation of aerobic respiration, although the exact mechanism has yet to be elucidated.
Despite the emerging appreciation for the importance of saNOS in S. aureus physiology and virulence, surprisingly little is understood about how the expression of nos is regulated.Expression of nos is upregulated under low-oxygen growth relative to aerobic growth (18), and the two-component regulatory system SrrAB has been identified as a positive regulator of nos responsible for increased nos expression during low-oxygen growth (3,23).SrrAB is thought to sense and respond to the reduced state of the respiratory menaquinone pool to regulate genes in response to low oxygen and nitrosative stress (24,25).Under nitrosative stress, SrrAB upregulates certain genes involved in the nitrosative stress response, anaerobic metabolism, and cytochrome biosynthesis (25) which have also been shown by transcriptome studies to be upregula ted in the nos mutant under aerobic respiratory conditions (2).In addition, the interplay between saNOS and SrrAB appears to be more complicated than one-way regulation, as an S. aureus nos srrAB mutant displays distinct changes in physiology and gene expression relative to wild-type and single-mutant strains (3).The global transcriptional regulator MgrA was also recently shown to affect nos expression under anaerobic and nitrosative stress conditions (26).In this study, we further characterize SrrAB and MgrA as regulators of nos, determine the effect of other candidate nos regulators and strain dependency on nos expression, and identify the transcriptional start site (TSS) of the S. aureus nos transcript.

Generation of S. aureus mutants
For the generation of UAMS-1 agr and srrAB agr mutants, primer sets agr_5′F + agr_5′R and agr_3′F + agr_3′R (Table S2) were used to PCR amplify two segments from UAMS-1 genomic DNA overlapping the 5′ and 3′ ends of the agr locus.The 5′ segment contained restriction sites EcoRI and XhoI on the 5′ and 3′ ends, respectively; the 3′ segment contained restriction sites XhoI and SalI on the 5′ and 3′ ends, respectively.Each segment was separately cloned into pCR-Blunt, sequenced, cut out by restriction digest, and ligated to each other and into the temperature-sensitive allele replacement vector pJB38 (29), which was digested with EcoRI and SalI.The ligation of the two segments eliminates RNAII (agrBDCA) as well as RNAIII, creating a complete agr deletion allele.Plasmid pJB38-Δagr was then phage transduced from S. aureus RN4220 into UAMS-1 (wild type) or KB6004 (UAMS-1 ΔsrrAB).Once its presence in the target strain was confirmed, a temperature-sensitive allele replacement event was initiated by growth at 43°C (the nonpermissive temperature for plasmid replication) on TSA plus 10 µg/mL Cm to promote chromosomal integration via homologous recombination at the agr locus, as described in (30).A second recombination event was induced by growing a single isolated colony in TSB (no antibiotic) for 5 days at 30°C with subculturing every 24 h.Screening for the loss of the plasmid was completed by picking and patching colonies onto TSA and TSA plus 5 µg/mL Cm.PCR with primers agr_up_F + agr_down_R (Table S2), located upstream and downstream of the agr locus, was used to confirm the presence of the agr deletion allele in strains UAMS-1 and KB6004.
For generation of JLB316 (AH1263 agr mutant), the regions upstream of RNAIII and downstream of agrA were amplified by PCR using primers listed in Table S2 and cloned into pJB38 (29).Briefly, the downstream fragment was amplified from the AH1263 genome using primers NS24 and JBKU123, digested with KpnI and SalI, and ligated into the same sites of pJB38 to generate pMP8.Next, the upstream fragment was amplified using primers JBKU120 and JBKU121, digested with EcoRI and NheI and cloned into the same sites of pMP8 to create pMP9, which was subsequently moved into RN4220 and AH1263 by electroporation and phage transduction, respectively.Allele replacement mutagenesis was then performed as described in references (29,30).
For generation of the rex mutant, the previously created pJF102 plasmid containing Δrex::kan was used (31).This plasmid was phage transduced into wild-type UAMS-1, with growth at 30°C for all steps.Once its presence in the target strain was confirmed, a temperature-sensitive allele replacement event was initiated by growth at 43°C (nonpermissive temperature for plasmid replication) on TSA plus 50 µg/mL Kan to promote the integration of the plasmid into the chromosome via homologous recom bination at the rex locus.A second recombination event was induced by growing a single isolated colony in TSB (no antibiotic) for 5 days at 30°C with subculturing every 24 h.Screening for the loss of the vector was completed by picking and patching colonies onto TSA plus 50 µg/mL Kan and TSA plus 5 µg/mL Cm.PCR was used to confirm the presence of the rex::kan insertion allele.

nos-lacZ reporter design
Plasmid pJB185 (27) containing a codon-optimized lacZ gene was used to create nos-lacZ fusion reporter plasmids pJBnos1 and pJBnos2.The primers used to amplify the nos promoter region segments (approximately 500 bp region upstream of the nos ATG start codon) cloned into each reporter construct (nos_F + nos_R for pJBnos1 and nos2_F + nos2_R for pJBnos2) are listed in Table S2.The regions of the DNA sequence upstream to nos cloned in each reporter plasmid are also indicated in Fig. 1A.The region cloned upstream of lacZ in pJBnos1 includes the nos predicted start codon and a 508 bp region upstream, including the putative Shine-Dalgarno (SD) sequence.Plasmid pJBnos2 contains the 500 bp region upstream to the predicted SD sequence and does not include this or the nos start codon.Instead, a non-native SD sequence and translation enhancer region (28) present upstream of lacZ in pJB185 are used for translation in this construct.Briefly, the nos promoter regions were each PCR amplified from UAMS-1 genomic DNA, cloned into pCR-Blunt, sequenced, cut out by restriction digest, and ligated into pJB185 upstream of the codon-optimized lacZ gene.Restriction enzymes (EcoRI and XbaI for pJBnos1; EcoRI and BamHI for pJBnos2) were chosen to include or eliminate the appropriate portions of the pJB185 sequence.Plasmid constructs were phage transduced from RN4220 into UAMS-1, KB6004, KR1300, KR6300, rex, mgrA, AH1263, and JLB316, as indicated in each experiment.

Mapping the nos TSS
A 5′ rapid amplification of cDNA ends (RACE) was performed according to a protocol provided by Life Technologies.Briefly, total RNA was isolated from S. aureus UAMS-1 and AH1263 cultures grown for 6 h (late exponential growth phase) in low-oxygen conditions using the RNeasy Mini Kit (Qiagen) as previously described (32).nos cDNA was then generated using the iScript Select cDNA Synthesis Kit (Bio-Rad) and the primer nos-GSP1 (Table S2).cDNA was purified using the Zymo Clean and Concentrator kit (Zymo Research) and used in a homopolymeric tailing reaction with 2 mM dCTP and 15 units of TdT (Invitrogen).The reaction product was amplified further in several rounds of PCR using nested genespecific primers (nos-GSP2 and nos-GSP3; Table S2), and products were visualized on a 1% agarose gel using gel electrophoresis.Once a suitable amount of DNA was visualized, products were sent for Sanger sequencing (Genewiz) with primers nos-GSP3 and nos-screen (Table S2).The TSS (or the first nucleotide of the mRNA) was then determined from sequencing results by locating the homopolymeric tail and the first nucleotide following it.Similar methods were used with lacZspecific primers (lacZ-GSP1, lacZ-GSP2, lacZ-GSP3, and lacZ-screen; Table S2) to map the TSS in nos-lacZ fusion plasmids pJBnos1 and pJBnos2.Sanger sequencing results are provided in Supplementary file 1.

β-galactosidase assays (modified Miller assays)
All S. aureus wild-type and mutant strains containing pJBnos1, pJBnos2, or pJB185 (promoterless vector control) were grown in aerobic and/or low-oxygen conditions as described above.Cell pellets were harvested at 2-h (early exponential phase) and/or 6-h growth (late exponential phase) in each condition, with OD 600 recorded at each time point.At 2 h, cells were harvested from a 20 mL volume of aerobic cultures and a 40 mL volume of low-oxygen cultures by centrifugation at 4,500 rpm.At 6 h, 5 mL was harvested from aerobic cultures and 20 mL from low-oxygen cultures by centrifugation at 4,500 rpm.Pellets were stored at −80°C until further processing.For isolation of cellular protein, cell pellets were thawed and resuspended in 800 µL 1× phosphatebuffered saline (PBS), then transferred to lysing matrix B tubes (MP Biomedicals) and lysed using a FastPrep (MP Biomedicals).The FastPrep was run at 6 m/sec for 30 s two times, with tubes placed in ice for 2 min between runs.Lysates were centrifuged at 4°C for 10 min at 12,000× g, then transferred and split into two new sterile tubes for use in BCA and β-galactosidase assays.The Pierce BCA protein assay kit (Thermo Fisher Scientific) was used to measure lysate protein concentration.β-galactosidase assays (modified Miller assays) were performed using ONPG substrate as described in reference (28).Activity was reported in modified Miller units (MU) with the formula MU = 1,000 × [A 420 /(t × v × mg/mL)], whereby t = time of reaction in minutes, v = volume (mL) of cell lysate, and mg/mL = protein concentration.

RNA isolation and qPCR
To measure late-exponential-phase (6-h growth) nos gene expression in wild-type UAMS-1 and isogenic srrAB, agr, and srrAB agr mutant strains, as well as in wild-type AH1263 and isogenic agr mutant strain, aerobic and low-oxygen cultures of each strain were grown as described above.At 6-h growth, 5 mL of each aerobic culture and 20 mL of each low-oxygen culture were centrifuged at 4°C, 4,500 rpm for 10 min, and cell pellets were resuspended in RNAlater.Pellets were stored at −80°C until further processing.RNA was extracted using a FastPrep system with Lysing Matrix B tubes as previously described (32).RNA was further treated with Turbo DNase (Ambion Turbo DNA-free kit) to remove any contaminating DNA, and the RNA concentration and purity were quantified using a BioTek Take3 plate.Similar growth conditions and RNA isolation methods were used with strains UAMS-1, AH1263, UAMS-1 pJBnos1, and UAMS-1 pJBnos2 for 5′ RACE.For quantitative real-time PCR (qPCR), cDNA was synthesized from 0.75 µg of purified RNA using an iScript reverse transcriptase kit (Bio-Rad).Expression of nos was measured using iQ SYBR Green Supermix (Bio-Rad) and a CFX Connect real-time system (Bio-Rad).Relative fold expression normalized to that of the reference housekeeping gene sigA was calculated using the Livak method (2− ΔΔCT ), as previously described (18,33).The primers used for qPCR are listed in Table S2.For UAMS-1 and isogenic mutants, qPCR was performed on three biological samples with three technical replicates per sample.For AH1263, qPCR was performed on two biological samples with three technical replicates per sample.

Statistical analysis
Unless otherwise indicated, assays were conducted in at least three independent experiments.Statistical analysis for all data was completed with SigmaPlot software (version 14; Systat Software Inc.) or GraphPad Prism (version 9; GraphPad Software, Dotmatics).Data were tested for normality and equal variance prior to choosing the appropriate parametric or nonparametric test.

Validation of pJBnos1 as a reporter plasmid for nos expression
Plasmid pJBnos1, a nos-lacZ reporter plasmid containing the nos predicted SD sequence and start codon (Fig. 1A), was used to assess nos promoter activity in UAMS-1 dur ing aerobic and low-oxygen growth (Fig. 1B), with parallel measurements of UAMS-1 containing promoterless pJB185 as a negative control.These experiments revealed that β-galactosidase activity in UAMS-1 pJBnos1 was significantly higher (P < 0.05) than the corresponding promoterless control strain at both time points and growth conditions.Furthermore, peak nos promoter activity in pJBnos1 occurred in low-oxygen culture conditions during late exponential phase (6 h), which correlates with the previously observed pattern of nos RNA levels being maximally expressed at late exponential phase during low-oxygen growth (18) (Fig. 1B).Taken together, these results suggest that nos promoter-driven β-galactosidase activity in pJBnos1 serves as a useful proxy for measuring nos expression.

Annotation of the nos promoter
A 5′ RACE experiment was performed to identify the TSS of nos.In three independ ently performed and sequenced RACE reactions using RNA isolated from UAMS-1 6-h low-oxygen cultures [growth condition previously reported to have maximal nos transcript levels (18)], the TSS consistently mapped to an adenine nucleotide positioned 11 nucleotides upstream of the start codon (TSS-1), in the middle of the predicted SD sequence (Fig. 1A).This experiment was repeated on RNA isolated from strain AH1263 grown under the same conditions and confirmed the same mapped TSS-1 site as in UAMS-1.To further confirm this unusual result, 5′ RACE was also performed on RNA from UAMS-1 harboring the pJBnos1 reporter plasmid using lacZ-specific primers to identify the TSS of lacZ transcription.In two independently performed and sequenced RACE reactions, the TSS of lacZ in this construct mapped to the same adenine nucleo tide (TSS-1) as nos 5′ RACE performed on RNA from UAMS-1 and AH1263 (Fig. 1A).To determine whether TSS-1 is required for transcription, we created another nos-lacZ construct in plasmid pJB185, termed pJBnos2, which contains the same nos promoter region as pJBnos1, but ends just upstream of TSS-1, thereby also eliminating the nos predicted SD sequence and start codon.Instead, pJBnos2 uses a non-native SD sequence engineered upstream of lacZ to drive translation.We predicted that pJBnos2 would have little to no β-galactosidase activity since it lacks TSS-1.Using growth conditions identical to UAMS-1 pJBnos1, modified Miller assays were used to measure β-galac tosidase activity in UAMS-1 pJBnos2 (Fig. 1B).Although β-galactosidase activity was significantly (P < 0.05) decreased in pJBnos2 relative to pJBnos1 at both time points and growth conditions, nos promoter activity at late exponential phase in pJBnos2 was still significantly higher (P < 0.05) than the corresponding promoterless pJB185 control culture (Fig. 1B).Therefore, late-exponential-phase transcription of lacZ can occur to some extent in pJBnos2 under both aerobic and low-oxygen growth conditions, even in the absence of TSS-1.
Promoter activity of pJBnos2 followed a similar pattern to that observed with pJBnos1 (Fig. 1B) and previously quantified nos RNA levels (18), with peak activity observed in late-exponential-phase (6 h) low-oxygen cultures (Fig. 1B).5′ RACE on the pJBnos2 construct using lacZspecific primers, also performed in two independent reactions, identified a thymine nucleotide 30 bp upstream of the nos start codon (TSS-2) and 19 bp upstream of TSS-1 (Fig. 1A).It is unlikely that potential secondary structure in the TSS-1 region blocks readthrough of the reverse transcriptase reaction in the 5′ RACE assay, as repeating this experiment with an extended denaturation (10 min) and higher cDNA synthesis temperature (50°C) identified the same adenine TSS-1 site.Furthermore, the secondary structure in the TSS-1 region predicted by mFold modeling (34) (Fig. S1), combined with the predicted melting temperature of this structure (53.2°C), does not support the idea of a secondary structure interfering with the reverse transcriptase reaction.In addition, if this structure were formed, the reaction would presumably be blocked before the identified adenine nucleotide.Putative −10 promoter elements were estimated upstream of both TSS-1 and TSS-2 (Fig. 1A).It is possible TSS-2 could represent a less preferred TSS that is strong enough to initiate transcription in the absence of TSS-1, which may explain the decreased nos promoter activity of pJBnos2.Given the unusual location of TSS-1 (within the putative SD sequence), it possibly represents a site subject to posttranscriptional mRNA processing, such as by RNase III, RNase Y, RNase J1/J2, or another staphylococcal RNase.RNase III, which contributes to virulence gene regulation through degradation of the mRNA duplex that forms between Agr effector RNAIII and Protein A (spa) (35), is a double-strand specific RNase (36).As mRNA modeling consistently showed the region around TSS-1 as a single-stranded loop (Fig. S1), this RNase would likely only cleave at this position if there were an interaction with another RNA.RNase Y, on the other hand, cleaves single-stranded A-rich or AU-rich sequences (37)(38)(39); the site of TSS-1 does not seem to meet this criterion, as it is surrounded by the CGAGG of the putative SD sequence.RNase J1 and J2 form a heterodimer (RNase J) involved in the maturation and processing of 16S rRNA among many others (40)(41)(42).This complex has both endonuclease and 5′ to 3′ exonuclease activity, which is single-strand specific (39,43,44).While further investigation is required to determine whether mRNA processing indeed occurs at TSS-1, it is noteworthy that the nos transcript has not been specifically identified in published studies as being a candidate for processing and/or regulation in S. aureus by RNase III (45,46) or RNase Y (37).However, a previously published exact mapping of transcriptome ends (EMOTE) assay identified a 5′ monophosphorylated guanine nucleotide (located immediately upstream of the TSS-1 adenine of nos) as being enriched in S. aureus RNase J1 and J1/J2 mutants, suggesting that this RNase complex may regulate the stability of the nos transcript (40).
Alternatively, TSS-1 may be an actual initiation site of transcription, and its close location to the ATG start codon may implicate the nos transcript as potentially having a leaderless organization with no identifiable SD sequence, a phenomenon not well characterized in S. aureus.Leaderless mRNAs permit translation without an identifiable SD sequence (47).It remains unclear exactly how translation is initiated from these leaderless transcripts, but several necessary elements have been implicated, including a local absence of secondary structure and simply the 5′ AUG of the mRNA (48,49).

Confirmation of SrrAB and MgrA as nos regulators in UAMS-1
Understanding the nos regulatory network is an important piece of a more complete appreciation of S. aureus physiology, stress resistance, and virulence.The SrrAB TCS responds to both oxidative and nitrosative stress via sensing of the respiratory mena quinone pool and has both positive and negative effects on the expression of genes associated with metabolism (23), nitrosative stress (25), virulence (50), and biofilm formation (51,52).In addition, SrrAB represses the expression of the Accessory Gene Regulator (Agr) TCS (50,53), an important global quorum-sensing regulator [reviewed in references (54,55)].SrrAB has previously been identified as a positive regulator of nos mRNA levels, with nos RNA levels displaying a fourfold decrease by qPCR in a UAMS-1 srrAB mutant (3), and a sevenfold decrease by RNA microarray in a JE2 (LAC-derived strain) srrA mutant (23).The pattern of nos promoter-driven β-galactosidase activity in late-exponential-phase srrAB mutant pJBnos1 cultures (Fig. 2A) correlated with these previously published RNA results, whereby pJBnos1 promoter activity (containing TSS-1) in the srrAB mutant low-oxygen culture displayed a 3.3-fold significant decrease (P < 0.01) compared to UAMS-1.When these experiments were repeated with wild-type and srrAB mutant strains harboring pJBnos2 (lacking TSS-1) (Fig. 2B), similar expression trends were observed, albeit at overall lower activity levels relative to pJBnos1.
Global transcriptional regulator MgrA, a member of the MarR/SarA family that regulates a variety of genes related to cell aggregation and virulence (56)(57)(58)(59)(60)(61), has also previously been implicated in nos regulation, with nos expression patterns in an mgrA mutant suggesting a repressive function of MgrA on nos expression during anaerobic growth, and a possible activating function under nitrosative stress conditions (26).β-galactosidase activity in late-exponential-phase mgrA pJBnos1 cultures confirmed that pJBnos1 promoter activity was significantly (P < 0.01) increased in the mgrA mutant during low-oxygen growth compared to UAMS-1 (Fig. 2A).However, different expression patterns were observed when these experiments were repeated with wild-type and mgrA mutant strains harboring pJBnos2 (Fig. 2B), whereby nos promoter activity was significantly decreased (P < 0.01) in the mgrA mutant during late-exponential-phase low-oxygen growth relative to wild type.This differential regulation of TSS-1 and TSS-2 driven nos promoter activity could be related to the observation that Cys12 (located in the dimerization domain) of MgrA is a target of both oxidation (62) and nitrosylation (63,64).Oxidation of MgrA Cys12 was shown to weaken target promoter binding by MgrA (62), while nitrosylation of this amino acid strengthened promoter binding by MgrA (63).Interestingly, saNOS-derived NO was shown to be required for endogenous nitrosylation of MgrA (63), and mgrA expression was increased in a nos mutant (2), further reinforcing the regulatory interplay between saNOS and MgrA.

Global regulators Rex and Agr affect nos expression
The redox-sensing transcriptional regulator Rex, a major player in S. aureus anaerobic gene regulation, is a known direct repressor of srrAB expression (65).Although inspec tion of the nos promoter region did not yield a convincing candidate Rex binding site, we pursued Rex as potentially involved in nos regulation due to its known connec tion to SrrAB and regulation of metabolic genes (65,66).As mentioned above, SrrAB directly represses the expression of the Agr quorum-sensing system (50).It has not been confirmed whether SrrAB regulation of nos is directly mediated by SrrA promoter binding; therefore, SrrAB could also regulate nos transcript levels indirectly through another regulator, such as Agr.Evidence for a regulatory relationship between SrrAB, Agr, and saNOS can be found in previously published transcriptomic data of a UAMS-1 nos mutant (2), which revealed that expression of srrAB in the nos mutant was downre gulated approximately 1.6-fold and agr operon genes were upregulated 1.7-2.3-fold.Similar Miller assays were performed using UAMS-1 and isogenic rex and agr mutants all harboring either the pJBnos1 or pJBnos2 reporter plasmid (Fig. 2).In the rex mutant, a 1.5-2-fold increase in β-galactosidase activity of pJBnos1 in late-exponential low-oxygen growth was observed (Fig. 2A).Rex is traditionally a repressor whose target genes are de-repressed in response to altered redox as oxygen levels decrease, so Rex-depend ent nos repression could be a result of Rex binding directly to the nos promoter or indirectly acting through its role as a repressor of SrrAB.A recently published com putational identification of S. aureus genome-wide Rex-binding sites using the DNA consensus sequence TTGTGAW6TCACAA (66) did not identify nos as a Rex-regulated gene, which corroborated the lack of a strong Rex-binding consensus sequence in our manual inspection of the nos promoter region.However, an indirect role for Rex may be involved in nos regulation, as transcript levels of several known Rex-regulated genes (pflAB, ldh1, nar, and nir operons) were also altered in a UAMS-1 nos mutant (2, 3), and decreased NADH levels were observed in late-exponential-phase nos mutant cells (2).β-galactosidase activity in late-exponential-phase UAMS-1 agr pJBnos1 cultures was also increased almost twofold during low-oxygen growth (Fig. 2A), suggesting a potential role for Agr as a negative regulator of nos expression under this growth condition.Interestingly, promoter activity in pJBnos2 was significantly (P < 0.01) decreased in both UAMS-1 rex (2-fold) and agr (2.4-fold) mutants under low oxygen growth (Fig. 2B), an opposite pattern to that observed in pJBnos1.Rex, Agr, and MgrA did not have a significant effect on aerobic regulation of nos expression in pJBnos1 (Fig. 2A), while pJBnos2 aerobic expression was modestly increased (P < 0.05) in mgrA and agr mutants (Fig. 2B).In addition, aerobic late-exponential-phase nos expression in pJBnos1 was modestly decreased (P < 0.05) in the srrAB mutant (Fig. 2A).Collectively, these data suggest that TSS-1 (in pJBnos1) and TSS-2 (in pJBnos2) may be subject to differential regulation by MgrA, Rex, and Agr in UAMS-1 when grown under low-oxygen conditions.It is unclear whether this could be related to direct regulation (e.g., altered affinity for the nos promoter by binding of one or more regulators which, in turn, drives expression from TSS-1 or TSS-2), or possibly indirect regulation of other DNA-binding proteins or RNase(s) acting on TSS-1.

Agr and low-oxygen effects on nos expression are strain-dependent
Agr activity in strain UAMS-1 (Agr class III) is relatively low compared to many other clinically relevant S. aureus strains (67,68).To determine whether basal Agr activity impacts nos regulation, the pJBnos1 construct was moved into the USA300 LAC strain AH1263, which has much stronger Agr activity (Agr class I) (69-71), and JLB316 (isogenic AH1263 agr mutant).Surprisingly, nos promoter activity measured from pJBnos1 (containing the UAMS-1 nos promoter sequence) was significantly more active (~3.5-foldhigher) in late-exponential-phase low-oxygen AH1263 cultures compared to UAMS-1 (Fig. 3A).This pattern of increased nos expression in AH1263 was confirmed by compar ing late-exponential-phase low-oxygen nos RNA levels in both wild-type strains by qPCR (Fig. 3B).This striking difference in nos expression between UAMS-1 and AH1263 is not due to differences in nos promoter sequences, as BLAST comparison of the 522 bp upstream sequences to the nos ATG start codon from each strain revealed over 99% identity, with no nucleotide insertions, deletions, or substitutions in the 185 bp sequence immediately upstream to the ATG codon (data not shown).β-galactosidase activities measured during aerobic growth and early-exponential-phase low-oxygen growth in AH1263 also appeared to trend higher compared to UAMS-1, but these differences were not statistically significant (Fig. 3A).No significant differences in nos expression were observed between wild-type AH1263 and its isogenic agr mutant at any of the growth conditions and time points tested, indicating the involvement of Agr in nos regulation may be strain dependent and/or Agr class-dependent (Fig. 3A).
The striking difference in late-exponential-phase low-oxygen nos expression between UAMS-1 and AH1263 suggests that as-yet-unknown strainspecific differences in genetic regulators and/or physiology contribute to the control of nos expression.Although characterization of S. aureus nos mutants has revealed many conserved phenotypes across different strain backgrounds, including UAMS-1 and USA300 strains [increased pigmentation, increased oxygen consumption, increased tolerance to aminoglycoside antibiotics, decreased virulence (2-5, 18, 19)], there also appear to be some nos mutant metabolic phenotypes that are strain dependent.Aerobically grown UAMS-1 nos mutant cells display increased O 2 consumption and increased respiratory dehydrogenase activity (2,3,5), supporting a role for saNOS in slowing aerobic respiration and minimiz ing endogenous ROS, possibly by NO competition with O 2 at the heme active site of cytochromes (72)(73)(74)(75).However, nitrite from saNOS-derived endogenous NO was postulated to promote aerobic respiration in S. aureus JE2 (a USA300 strain) by stimu lating quinol oxidase activity, as the growth defect of this nos mutant could be rescued by nitrate or nitrite (but not urea or ammonia) and this rescue was abolished in a nos qox double mutant (4).In addition, nos mutants in the USA300 background appear to have a more pronounced aerobic growth defect (2, 4) compared to UAMS-1 (2,3), which could relate to the increased basal expression of nos observed between these two strain backgrounds (Fig. 3).

Agr is epistatic to SrrAB in regulating nos expression in S. aureus UAMS-1
A srrAB agr double mutant was made in strain UAMS-1 to investigate possible regulatory interaction between these two systems.β-galactosidase assays were performed in the double mutant with pJBnos1 (TSS-1) and pJBnos2 (TSS-2), and activity was compared to that in wild type and the respective single mutants (Fig. 2A and B).Interestingly, the activity of pJBnos1 in the srrAB agr double mutant phenocopied the agr single mutant.As previously observed, nos promoter activity was decreased in the srrAB mutant and increased in the agr mutant in late-exponential-phase low-oxygen cultures (Fig. 2A).The srrAB agr double mutant mimicked the agr single mutant, with the activity of pJBnos1 increased approximately twofold in low oxygen (Fig. 2A).Activity of pJBnos2 in the agr nos mutant also phenocopied the agr mutant, displaying reduced expression under low-oxygen growth and modestly increased expression under aerobic growth (Fig. 2B).
qPCR was also used to measure nos RNA levels in wild-type UAMS-1 and its isogenic srrAB, agr, and srrAB agr mutants.RNA was isolated from 6-h aerobic and low-oxygen cultures, the same growth conditions used in the β-galactosidase assays, and nos transcript levels were expressed as a fold change relative to the wild-type sample in each condition (Fig. 3C).Decreased nos RNA levels were observed in the srrAB mutant, which has been consistently observed here (Fig. 2 and 3C) and in previously published studies (3,23).In the agr and srrAB agr mutants, significant differences in nos transcript levels compared to the wild type (Fig. 3C) were not observed.While this differs from the increase seen with pJBnos1 in β-galactosidase assays (Fig. 2A), the fact that the addition of an agr mutation reverses the decreased nos promoter activity and transcript levels of the srrAB mutant holds true.
Agr is likely an indirect and/or posttranscriptional negative regulator of nos expres sion, since (i) no putative AgrA-binding site was identified in the nos promoter region and (ii) most Agr regulation is posttranscriptional through the activity of RNAIII.Given that SrrAB is known to repress Agr (50), decreased nos expression in the srrAB mutant could be a consequence of de-repressed Agr activity, which, in turn, could repress nos expression.This scenario is consistent with increased low-oxygen pJBnos1 expression observed in the UAMS-1 agr mutant (Fig. 2A).However, data from our laboratory and others have shown what seems to be an extensive interplay between saNOS and SrrAB (3,4).SrrAB upregulates several respiratory genes in response to altered respiration caused by a nos mutation (3,4).In addition, saNOS and SrrAB were observed to be epistatic to each other in regard to different phenotypes; SrrAB is epistatic to saNOS in the expression of nitrosative stress genes, NAD + /NADH ratio, and in vivo virulence, while saNOS is epistatic to SrrAB in aminoglycoside tolerance (3).This, along with the decreased expression of nos in a srrAB mutant, would seem to support SrrAB as a direct positive regulator of nos.Although it does remain possible that saNOS and SrrAB are independently involved in similar respiration-related processes and that decreased nos expression in the srrAB mutant is attributed to a stronger repressive effect of Agr, another possible explanation is that SrrAB and Agr independently regulate nos expression, with SrrAB as a positive regulator and Agr as a negative regulator.Decreased nos expression in the srrAB mutant could be an additive effect from both regulators, with SrrAB not present to activate nos and increased agr expression caused by the srrAB mutation leading to further nos repression.In this scenario, the observation that nos expression in the UAMS-1 srrAB agr double mutant mimics increased nos expression in the agr single mutant could mean that mutation of agr induces a SrrAB-independent mechanism of nos activation, or that there is an alternative unknown regulator that is regulated or repressed by Agr, that is responsible for activating nos expression.

Conclusions
Deciphering the nos regulatory network in S. aureus is key to further understanding how saNOS is involved in aerobic and anaerobic metabolism, stress resistance, and virulence.Overall, this study has identified two potential nos TSSs that may be differ entially regulated by growth conditions and/or specific global regulatory systems, as summarized in Fig. 4.This study has also further confirmed the roles of SrrAB and MgrA in nos regulation and has identified Agr and Rex as putative negative regulators of nos expression in low-oxygen UAMS-1 cultures.Although the effects of potential interactions between these regulators, TSS-1, and TSS-2 on nos expression needs further study, it is possible that Agr acts through MgrA, as RNAIII has been shown to increase MgrA levels through stabilization of the mRNA (76).In addition, the increased low-oxygen nos expression observed in a rex mutant could be attributed to SrrAB, as Rex is a known repressor of srrAB expression (65).It is not likely a coincidence that all identified putative nos regulators are deeply intertwined with each other with respect to their roles in regulating metabolic genes; therefore, it is not surprising that mutation of multiple regulators could be causing similar effects on nos expression.We have also identified strain-dependent differences in nos expression and regulation, highlighting the importance of studying saNOS in different strain backgrounds.In addition, we have mapped two potential transcriptional start sites within the nos promoter, with TSS-1 appearing to be the stronger of the two.The location of TSS-1 within the putative SD sequence suggests that the nos transcript may be subject to unusual regulations such as RNA processing or leaderless translation.While further studies are needed to confirm the exact mechanisms behind this regulatory network, this current work contributes to our understanding of how nos expression is controlled, particularly in low-oxygen conditions, to function in the switch from aerobic to anaerobic metabolism (5), helping to build a more complete picture of how nos integrates into the transcriptional network that controls and responds to S. aureus metabolism.

FIG 1
FIG 1 Diagram of the nos promoter region (A) and validation of growth-phase-and oxygen-dependent pJBnos1 and pJBnos2 promoter activities (B).(A) The genome-annotated start codon (black bolded) and upstream putative Shine-Dalgarno sequence (SD; blue bolded) are indicated.The transcriptional start site (TSS-1) identified by 5′ rapid amplification of cDNA ends (RACE) using nos cDNA template from UAMS-1 and AH1263, and lacZ cDNA template from UAMS-1 pJBnos1 was mapped to an adenine nucleotide (red bold) located in the predicted SD sequence.5′ RACE using lacZ cDNA from UAMS-1 pJBnos2 mapped a second TSS to a thymine nucleotide (red bold; TSS-2) located 30 bp upstream to the nos ATG start codon.Putative −10 elements (green bold) are also indicated upstream of each TSS.The region cloned upstream of lacZ in pJBnos1 is indicated by both the solid underline and dotted underline, while the region cloned upstream of lacZ in pJBnos2 contains only the solid underlined sequence.(B) β-galactosidase activity of pJB185 (promoterless), pJBnos1, and pJBnos2 in wild-type S. aureus UAMS-1 was assessed in cell pellets harvested from aerobic or low-oxygen TSB + G cultures at 2-and 6-h growth, as described in Materials and Methods.Data are reported in modified Miller units.All data represent the average from three independent experiments; error bars indicate the standard error of the mean (SEM).*P < 0.05 (Holm-Sidak method) compared to pJB185 at each time point and growth condition; **P < 0.05 (Holm-Sidak method) compared to pJBnos1 at each time point and growth condition; # P < 0.05 (Tukey's test) compared to pJB185 at each time point and growth condition.

FIG 2
FIG 2 SrrAB, MgrA, Rex, and Agr affect nos expression in S. aureus UAMS-1.Cell pellets were harvested from aerobic or low-oxygen TSB + G cultures at 6 h for β-galactosidase assays, as described in Materials and Methods.The activity of pJBnos1 (A) and pJBnos2 (B) in wild-type UAMS-1 and isogenic srrAB, mgrA, rex, agr, and srrAB agr mutants is reported in modified Miller units.All data represent the average from three independent experiments; error bars indicate the standard error of the mean (SEM).*P < 0.05 (t-test) compared to UAMS-1 under the same growth condition, **P < 0.01 (t-test) compared to UAMS-1 under the same growth condition.

FIG 3
FIG 3 Low-oxygen and Agr-dependent nos expression is strain-dependent.(A) The activity of pJBnos1 in wild-type UAMS-1, AH1263 (LAC), and JLB316 (LAC agr mutant) reported in modified Miller units.Pellets were harvested from aerobic or low-oxygen TSB + G cultures at indicated time points for β-galactosidase assays, as described in Materials and Methods.All data represent the average from three independent experiments; error bars indicate the standard error of the mean (SEM).*P < 0.05 (two-tailed t-test) compared to UAMS-1 6-h low-oxygen time point.(B) nos RNA levels in UAMS-1 and AH1263(LAC) were measured from low-oxygen TSB + G cultures of UAMS-1 and AH1263 at 6-h growth, as described in Materials and Methods.Quantitative real-time PCR was performed on reverse-transcribed cDNA from each sample using nosspecific primers.The Livak (2− ΔΔCT ) method was used to determine the relative fold change of nos expression, using sigA expression as the reference gene and UAMS-1 as the calibrator.(C) nos RNA levels were measured in 6-h low-oxygen cultures of wild-type UAMS-1 and isogenic srrAB, agr, and srrAB agr mutants.Quantitative real-time PCR was performed on reverse-transcribed cDNA from each sample using nosspecific primers.The Livak (2− ΔΔCT ) method was used to determine the relative fold change of nos expression, using sigA expression as the reference gene and the UAMS-1 sample as the calibrator for each growth condition.All data represent the average from two to three biological replicates; error bars indicate the SEM.*P < 0.05 (two-tailed t-test) compared to UAMS-1.

FIG 4
FIG 4 Summary of nos regulation.(A) Based on pJBnos1 data in this paper, SrrAB and unknown strain-dependent factors are positive regulators (green) of TSS-1-dependent nos expression during low-oxygen growth, whereas Agr, Rex, and MgrA appear to have repressive functions (red) on TSS-1-dependent nos expression in strain UAMS-1.TSS-1 may represent a site for RNA processing or possibly represents leaderless translation.(B) Based on pJBnos2 data, TSS-2-dependent nos expression is subject to positive regulation by SrrAB under both low-oxygen and aerobic growth conditions, whereas MgrA, Rex, and Agr-dependent regulation are dependent on the growth condition (all three regulators positively regulate TSS-2-dependent nos expression during low-oxygen growth, whereas Agr and MgrA repress expression under aerobic growth conditions).