Regulatory Adaptation of Staphylococcus aureus during Nasal Colonization of Humans

The nasopharynx is the main ecological niche of the human pathogen Staphylococcus aureus. Although colonization of the nares is asymptomatic, nasal carriage is a known risk factor for endogenous staphylococcal infection. We quantified S. aureus mRNA levels in nose swabs of persistent carriers to gain insight into the regulatory adaptation of the bacterium to the nasal environment. We could elucidate a general response of the pathogen to the surrounding milieu independent of the strain background or the human host. Colonizing bacteria preferentially express molecules necessary for tissue adherence or immune-evasion whereas toxins are down regulated. From the analysis of regulatory loci we found evidence for a predominate role of the essential two-component system WalKR of S. aureus. The results suggest that during persistent colonization the bacteria are metabolically active with a high cell surface turnover. The increased understanding of bacterial factors that maintain the colonization state can open new therapeutic options to control nasal carriage and subsequent infections.


Introduction
Staphylococcus aureus is both a major human pathogen and a ubiquitous commensal and colonizer of the skin and mucous membranes. Although multiple body sites can be colonized, the anterior nares form the main ecological niche of this species. Approximately 20% of the healthy human population is persistently and 80% intermittently colonized with S. aureus in the nose [1]. Nasal carriage has been identified as a major risk factor for the development of subsequent mostly endogenous infections [2,3].
Several bacterial factors were determined to be involved in S. aureus nasal colonization: the wall teichoic acid (WTA) [4], clumping factor B (clfB) [5], the capsular polysaccharide (cap) [6], the iron-regulated surface determinant IsdA (isdA) [7], and the autolysin SceD (sceD) [8]. All factors were so far studied in vitro or in animal models with the exception of the adhesin ClfB, whose causal involvement was shown in humans [5]. Expression of most virulence and adherence factors were shown in vitro to be directly or indirectly influenced by diverse regulators such as the accessory gene regulator (Agr), the alternative sigma factor B (SigB), and the sae locus [9,10,11,12]. The composition of the cell envelope is modified by the regulatory action of graRS (also known as apsXRS) [13,14] or the essential regulatory system walKR [15].
A better understanding of the bacterial factors maintaining the colonization state can be important for controlling nasal carriage and subsequent infections. In the present study transcript analysis using quantitative real-time PCR was performed directly on nose swabs from persistently colonized healthy individuals. We could show that in the human nose most global virulence regulators are not active, with exception of the essential two-component regulatory system WalKR. Furthermore, S. aureus nasal colonization is characterized by the expression of genes mediating adhesion, cell surface dynamics/remodeling, the expression of immune evasion genes and the lack of toxin transcription.

Characteristics of nose specimens from healthy S. aureus nose carriers
For the analysis of S. aureus gene expression during human nose colonization we selected four persistently colonized individuals (A, B, C and D) with a history of repeated S. aureus-positive nose swabs. None of the volunteers received any antibiotics immediately before or between samplings. In general, S. aureus loads in the nose varied considerably over time, as exemplified for the two carriers A and D (Fig. 1). Specimens for transcript analysis were collected at two different time points: at the first sampling nose swabs were obtained from all four individuals, and at the second sampling 12 months later they were collected only from individual A and D. CFUs between 10 4 -10 5 /swab were detected in the noses of the volunteers at these time points. Bacteriological analysis revealed that all volunteers were colonized with a single S. aureus phenotype over time with the exception of individual A who carried a hemolytic and a non-hemolytic strain simultaneously at the second sampling (table 1). All carriers were colonized by distinct clones. Individuals A and D retained the same S. aureus strain over the whole investigation period. Detection of genes encoded on mobile elements revealed that none of the isolates contained a SCCmec cassette or the phage-encoded Panton-Valentine leukocidin gene. All nose isolates were lysogenic for hlbconverting phages encoding either staphylokinase (sak), the staphylococcal complement inhibitor (scn) and/or chemotaxisinhibitor protein (chp) with exception of the ß-hemolysin (HLB) producing variant of individual A.
Most of the regulatory circuits controlling S. aureus gene expression have been described for prototypic strains. However, clinical isolates often exhibit a different regulation pattern. To elucidate the general properties of our four nose isolates we first analyzed the bacteria under in vitro conditions with quantitative RT-PCR (table S2) and for selected genes by Northern Blot analysis in comparison with the three characterized S. aureus strains Newman, HG001 and USA300 (Fig. 2). So far no expression data in S. aureus were available for the essential two-component system walKR. We could show that it is maximally transcribed during the post-exponential phase in all nose isolates and in strain Newman. Next we determined transcription of phenol-soluble modulin (encoded by psm), which until now has only been described for caMRSAs such as USA300 [16]. We could detect an equally high expression in the post-exponential phase in all strains. In prototypic strains protein A is maximally expressed during the exponential phase due to the repressive action of the global regulator agr [17]. This could be confirmed but none of the four nose isolates showed the typical transcription pattern. Similar results were already obtained by us for CF isolates [18]. The immune-modulator molecule scn was reported to be transcribed mainly during exponential growth [19]. We could not reproduce this observation either in the nose isolates or in the three laboratory strains.

Direct transcript analysis in nose swabs from S. aureus carriers
To profile the expression pattern of S. aureus during nasal colonization we performed quantitative transcript analysis directly on nose swabs obtained from persistent carriers. 30 different genes associated with a variety of cellular functions such as virulence regulation, toxin production, adhesion, cell wall dynamics and modification, immune modulation, SOS response and metabolic regulation were selected as targets (table S2). The transcript levels of all the target genes could be reliably quantified in these nose specimens with the exception of four genes: the mRNAs of mprF, dltA, isaA, and ica were below the detection level. In addition, these targets could only be detected from bacteria grown in vitro when 100-fold concentrated template RNA was analyzed, emphasizing the very low levels of these transcripts. The mRNA of the remaining 26 genes was quantified in the ex vivo material from the nose and compared to the transcription of the nose isolates during  growth in vitro. In the first sampling two consecutive swabs from each of the four carriers were analyzed and the ratio of in vivo to minimal or maximal expression in vitro was calculated (Fig. 3A, B). Two of the four original carriers were sampled again 12 months after the first time point and multiple swabs were obtained for statistical analysis. The results are depicted as absolute copy numbers in reference to the housekeeping gene gyrB and in comparison with the expression profile in vitro (Fig. 4, 5). In the following the main results derived from both samplings are summarized according to the functional categorization of target genes:

i.) Global regulators
Characteristically, S. aureus expresses adhesins in the early stages of growth, whereas toxins are produced in the late stages. A complex regulatory network is responsible for this differential gene expression, which is tightly controlled in response to cell density, energy availability and environmental signals. In our study we investigated the activity of five prominent regulators: agr, sae, sigB (detected as the tightly sigB-dependent asp), graRS and walKR. All regulators with the exception of walKR were inactive during nose colonization, as shown by the low expression in vivo compared to maximal expression of the same strains in vitro (Fig. 3, 4). In contrast, the essential two-component system WalKR was highly transcribed in the noses of individual A and D similar to the maximal expression during the post-exponential phase, suggesting a role of this regulator during colonization (Fig. 4).

ii.) Toxins
Invasion of host tissue, bacterial spread, and lysis of host cells is mediated by different types of S. aureus toxins. These molecules are usually expressed in high amounts during post-exponential growth (table S2). Transcript analysis of the alpha-hemolysin (hla), psm and a bi-component leukotoxin homologue termed blhB revealed that in all individuals these toxins were only poorly expressed in vivo (Fig. 3, 4).

iii.) Adhesins
In S. aureus, cell-wall-anchored proteins and other adhesive molecules are implicated in binding to host matrix molecules and may therefore play an essential role in the establishment of colonization. This is consistent with the maximal expression of most adhesins during exponential growth (table S2). A pronounced transcription of the adhesins clfB, fnbA and isdA could be observed in the nose swabs of all investigated carriers (Fig. 3, 4). In contrast, transcript analysis of the post-exponentially expressed adhesins clfA and eap revealed a weak transcription during colonization.
The WTA of S. aureus is known to be an important colonization factor mediating adherence to epithelial and endothelial cells [20]. A multitude of enzymes are involved in WTA biosynthesis, but little is known about their regulation. We characterized the role of the WTA during colonization by analyzing the expression of tagO and tarK, which both contribute to WTA biosynthesis [4,21]. We detected high transcriptional levels of both enzymes in vivo and a constant expression during growth in vitro (Fig. 4).

iv.) Cell-wall modification enzymes
Enzymes involved in cell wall remodeling contribute especially to resistance against antimicrobial peptides. For instance, resistance to lysozyme is mediated by the O-acetyltransferase (oatA) [22] which was shown to be highly expressed during colonization (Fig. 3, 4). A pronounced in vivo expression was also detected for the autolysins atlA and sceD.

) Stress response and metabolic regulators
Bacterial exposure to stress conditions triggers the SOS response in which the RecA protein is the key enzyme. To test whether exogenous pressure is exerted on S. aureus in the nose we measured recA transcription. Expression was generally similar to or even below the minimal expression in vitro (Fig. 3), indicating the absence of SOS stimuli in the nasal environment.
Next we analyzed the activity of two metabolic regulators to gain insight into the physiological state of S. aureus during colonization. Bacteria adapt to amino acid insufficiency by a complex series of regulatory events known as the ''stringent response'' in which the bifunctional enzyme RelA is a key factor [23]. The repressor CodY has been described as inhibiting genes mainly involved in nitrogen metabolism [24]. Measurement of relA codY transcription during colonization revealed a basal expression of both metabolic regulators in the human nose (Fig. 3).

vi.) Immune evasion and immune modulatory factors
Protein A is a cell-wall-anchored protein involved in multiple immune-modulatory processes [25,26]. In all nose specimens with the exception of individual C spa transcription could be detected suggesting a positive role of this molecule (Fig. 3, 4). S. aureus produces extracellular capsular polysaccharide, which is thought to protect the pathogen against opsonophagocytic killing by polymorphonuclear leukocytes [27]. Expression analysis of the cap operon revealed low-level production during nose colonization in most of the individuals.
In S. aureus prophages which integrate specifically into the ßhemolysin (Hlb) gene are widely distributed, especially in nose isolates [28]. These prophages typically encode immune evasion molecules like SAK, SCIN, and CHIPS, which may contribute to the colonization capability of the bacterium by interacting with the host's innate immune system [29]. Transcript analysis of sak and chp revealed that these genes play an important role during persistent colonization since expression of both was clearly pronounced compared to the expression during growth in vitro (Fig. 3, 4, 5). In contrast, transcript analysis of the other phageencoded immune-evasion gene scn revealed only a weak expression in vivo.
Differences in S. aureus gene expression depending on host and time S. aureus gene expression during colonization may be influenced by individual factors and therefore vary in the different human hosts. However, most of the analyzed transcripts were expressed in similar amounts in all carriers, suggesting a general response of the pathogen to the nasal environment. This was most obvious when analyzing transcripts which showed a highly variable expression during growth in culture, e.g., spa and cap. Individual A harbored a strain with high in vitro spa expression level (up to 70 copies spa/ gyr) whereas the strain from individual D was a weak producer (0.4 copies spa/gyr). For both strains in vivo transcript levels of 4-8 copies spa/gyr could be detected. Similarly cap transcription leveled out during nose colonization (Fig. 3). clfB and chp, on the other hand, were not uniformly expressed in the noses of all individuals (Fig. 5).
Nasal colonization may necessitate a constant adaptation of the pathogen to its environment, which is probably reflected in a change in gene expression over time. However, when we reanalyzed S. aureus transcription in the noses of two of the original four carriers after 12 months, expression levels of most of the analyzed genes had remained unchanged in comparison to the first sampling. The only differently expressed genes were the lukhomologue blhB and the adhesin eap, which were more highly expressed at the second sampling (Fig. 4).

Discussion
The human nose is the primary reservoir of S. aureus, and evolution of the species was probably driven by adaptation to this milieu. In general, bacteria react to environmental stimuli by differential gene expression which is controlled by various interacting regulatory networks. Our results indicate that in the nose important adhesive molecules (clfB, isdA, fnbA, atlA, eap, WTA), genes involved in cell surface dynamics/remodeling (sceD, oatA, atlA), and immune-modulatory factors (sak, chp, spa, eap) are prominently expressed, whereas major toxins (hla, psm) are not transcribed. Adherence to human desquamated nasal epithelial cells is an important factor in successful nasal colonization and was shown to be mediated by clfB, isdA and WTA in vitro [7,30], in the  . Direct transcript analysis of S. aureus genes in the nose of of two persistent carriers (individuals A and D) twelve months after the first sampling. Transcripts were quantified in reference to the transcription of gyrB directly in the nose swabs (in vivo, black columns) and after growth to the exponential phase (exp, white columns) and the post-exponential phase (pxp, gray columns). Values from three separate samplings (individual A) and four separate samplings (individual D), respectively were used to calculate the mean expression. Statistically significant differences between the in vivo and in vitro results are indicated: ns, not significant, P.0.05; * P,0.01 to ,0.05; ** P,0.001 to 0.01; *** P,0.001. Gene name abbreviations see nares of rodents in in vivo models [4,31,32] and in humans [5]. Thus, known colonization factors could indeed be proven to be activated on the transcriptional level in the nose by S. aureus. This supports our assumption that the results obtained from expression profiling can unravel essential factors for nose adaptation.
Nasal secretions form the first line of defence of the innate immune system against inhaled bacteria [33,34]. The pathogen is able to counteract the most important mechanisms, namely antimicrobial molecules such as defensins and lysozyme, the complement system, immunoglobulins and phagocytes. S. aureus is lysozyme-resistant due to the combined action of OatA and WTA [22], both of which are produced by the pathogen during nose colonization. The highly transcribed staphylokinase is able to bind alpha-defensins, thus inhibiting their bactericidal effects [35]. In addition, SAK is able to activate surface-bound plasminogen into plasmin, which than has the ability to cleave the complement compound C3b and IgG, thereby efficiently preventing opsonization and subsequent phagocytosis by neutrophils [36]. Complement activation and opsonization are also blocked by the expression of protein A [25]. Surprisingly, we could not detect expression of the important complement inhibitor SCIN in the nose although it was shown that nasal secretions contain complement proteins [37].
To gain insight into the underlying regulatory network we analyzed five prominent regulators. The virulence regulators agr, sigB and sae were not activated during nose colonization. Accordingly, the typical target genes were only weakly expressed, toxins which are regulated by agr and sae, and clfA which is regulated by sigB [10,38]. Surprisingly, the graRS system which controls specific resistance mechanisms against antimicrobial peptides [13,22] was not expressed either, despite various cationic antimicrobial peptides in the nasal fluid. This low graRS expression is consistent with the high expression of oatA (generally downregulated by graRS) and the low expression of the two genes dltA and mprF (both strongly activated) [22]. However, another positively influenced factor, the autolysin atlA, was strongly expressed in the nose, indicating additional regulatory activators for this gene. Indeed, we could show that the essential, twocomponent system WalKR is active during nose colonization. Although only some of the target genes are known, it is interesting to note that atlA, sceD and sak are all positively regulated by walKR [39]. These factors were all expressed in the human nose. Thus, the WalKR regulatory system may play an important role in adaptation to the nasal environment. This is supported by our recent finding that the WalKR system is also the master regulator of adaptive gene expression in a cotton-rat model of nose colonization [40]. Furthermore, the high expression of cell wall metabolism genes (atlA, sceD) as well as the activity of WalKR strongly indicate that cell wall dynamics are critical for hostpathogen interaction and prolonged colonization of the human nose. However, little is known about the signaling cascade leading to activation of WalKR in S. aureus.
The physiological milieu of the human nose has an important influence on the adaptation of S. aureus to this ecological niche, and the host signals which are sensed by the pathogen are largely unknown. The expression of isdA is indicative of iron-limited conditions in vivo, since this molecule was shown to be highly ironregulated [31]. The pathogen seems to be actively dividing in the nose environment, as suggested by the expression of many of the enzymes involved in cell-wall biosynthesis (tagO, tarK, atlA, sceD). This notion is additionally supported by the fact that most of the genes expressed in the human nose are those which are normally expressed in vitro during the exponential growth phase (clfB, fnbA, isdA, sceD, chp), indicating active S. aureus cells. On the other hand, we found little indication that the typical stress response factors (recA, relA, codY, sigB) are needed in the colonization process. They may be required in a more hostile environment, e.g. during infection.
In summary, we could elucidate here for the first time the expression pattern of S. aureus during asymptomatic colonization of the human nose. So far, direct transcript analysis has been performed during different kinds of infection [11,18,41,42,43,44,45]. Taken together the results indicate that S. aureus is able to specifically adapt to different niches in the human host. These findings significantly enhance our understanding of the complex host-pathogen interplay.

Ethics statement
The ethic committee of the University Hospital of Tübingen approved the study design. Written consent was obtained from all participants involved in the study according to their guidelines.

RNA isolation
For in vitro transcript analysis, nose isolates were grown overnight in CYPG (10 g/l casamino acid, 10g/l yeast extract, 5 g/l NaCl, 0.5% glucose and 0.06 M phosphoglycerate) [48], diluted to an initial OD 600 of 0.05 in fresh medium and grown to the exponential (OD 600 = 0.5) or post-exponential (OD 600 = 0.5+4h) phase. Bacteria were harvested by centrifugation and dissolved in 1 ml Trizol reagent (Invitrogen). For in vivo transcript analysis, a cotton wool swab was moistened in 250 ml nuclease-free water and the left and right anterior nares of the human volunteers were swabbed. The swab was vigorously vortexed and 10 ml of the suspension were put aside for bacteriological analysis. The cotton wool was removed from the swab using sterile tweezers and both the suspension and the cotton wool were directly treated with 1 ml of Trizol LS reagent (Invitrogen).
Bacteria were lysed and RNA isolation was performed as described [41] with modifications to increase the amount of isolated RNA. First, precipitation was performed for 60 min at 220uC with the addition of 0.5 M ammonium acetate and 50 mg/ ml GlycoBlue (Ambion). Second, after precipitation a prolonged centrifugation step of 1 h at 12,000 g was carried out. The dried RNA pellet was dissolved in 20 ml nuclease-free water.
To eliminate contaminating DNA each RNA sample was digested with 8 U of RNase-free DNaseI (Roche), 5 mM MgCl 2 , and 16 U of RNasin (Promega) for 30 min at 25uC. DNaseI treatment was stopped using DNase inactivation reagent (Ambion).

Quantitative RT-PCR
Sequence-specific RNA standards for quantification were prepared as described previously [41]. Briefly, gene-specific primers with a 59-extension including the T7 promoter sequence (table S1) were used in a standard PCR. T7-driven in vitro transcription was performed in a standard transcription assay (T7-MEGAshortscript, Ambion). After subsequent DNase I treatment, RNA was recovered using the MEGAclear Kit (Ambion). RNA quantification was performed spectrophotometrically. RNA standards were diluted to 1610 6 , 3.16610 5 , 1610 5 , 3.16610 4 , 1610 4 and 3.16610 3 copies/ ml. 1 ml of RNA standards, 1 ml (of a 1:100 dilution) of in vitro and 3 ml of in vivo RNA were transcribed into cDNA using Super-ScriptIII Reverse Transcriptase (Invitrogen) and 200 ng of random hexamer primers (Fermentas). cDNAs were frozen at 220uC using Eppendorf LoBind tubes (Eppendorf) for prolonged storage.
Quantitative real-time PCR was carried out using the Light-Cycler instrument and the LightCycler DNA amplification kit for hybridization probes or for SYBR Green (Roche) with cDNA diluted 1:5 (in vivo) or 1:10 (in vitro). Reaction mixtures were prepared using primers listed in table S1 or published primers [18,41]. All primers were designed to account for possible gene polymorphisms (if necessary, wobbles were introduced, as in the case of eap and blhB) and were evaluated for optimal binding on the genome-sequenced S. aureus strains in silico. In the case of extensive polymorphisms of targets (e.g. eap) PCR efficiencies were additionally determined experimentally. Possible cross-reactions with S. epidermidis, a bacterial species often found alongside S. aureus in nose specimens, were excluded. The number of copies of each sample transcript was then determined with the aid of the LightCycler software. The specificity of the PCR reaction was verified on 3% agarose gels.
The absence of contaminating DNA was proven for each sample by quantitative real-time PCR using gyrB-specific primers. No amplification product could be detected in any of the samples.

Statistical analysis
Statistical analysis was performed with the Prism 4.0 package (GraphPad Software) using the one-way ANOVA test and Bonferroni's multiple comparison post-test. P,0.05 was considered to be statistically significant.