sRNA-mediated crosstalk between cell wall stress and galactose metabolism in Staphylococcus aureus

Abstract

Staphylococcus aureus is an opportunistic pathogen responsible for a wide range of diseases in humans. During infections, this bacterium is exposed to various stresses that target its cell wall, such as oxidative or acid environments as well as various cell wall-acting antimicrobials. Staphylococcus aureus has effective regulatory systems for responding to environmental stresses, enabling the expression of factors necessary for its survival. Bacterial small RNAs (sRNAs) play a crucial role in this adaptation process. In this study, we show that RsaOI, an S. aureus sRNA, accumulates under acid stress conditions. This response is mediated via the two-component system VraSR, which is associated with the cell wall damage response. As a component of the VraSR regulon, RsaOI contributes to the survival of S. aureus under acid stress and affects its susceptibility to glycopeptide antibiotics. Our findings reveal that RsaOI targets the lacABCDFEG operon, which encodes components of tagatose pathway, a unique mechanism responsible for galactose metabolism in S. aureus. By antisense base pairing near the ribosome binding site of lacD, RsaOI inhibits the expression of this gene, encoding tagatose-6-phosphate aldolase. This regulation disrupts the tagatose pathway, impairing galactose utilization in S. aureus. These findings highlight the role of RsaOI in the mediation between cell wall stress responses and a specific metabolic pathway.

Graphical Abstract

Graphical Abstract.

Introduction

Staphylococcus aureus is a member of the commensal human microbiota that can act as an opportunistic pathogen, responsible for both nosocomial and community-acquired infections [1]. Its remarkable ability to cause a wide range of infections and establish successful colonization is largely due to its adaptability and resilience. Staphylococcus aureus can survive and proliferate in various hostile environments within the human host, including acidic conditions [2]. Such acidic microenvironments are commonly encountered on mucosal surfaces, in infected tissues where the immune response creates localized acidic conditions, and within phagolysosomes in the case of intracellular bacteria. To withstand these stresses, S. aureus employs a variety of adaptive mechanisms, including regulatory networks that allow rapid modulation of cellular processes essential for survival [2]. One such adaptation involves reducing the permeability of the cell membrane and cell wall to protons by altering their charge composition. During acid stress, Gram-positive bacteria increase the expression of genes involved in cell wall modifications, such as the cap gene, linked to capsule biosynthesis, and the dlt operon, which adds positive charges to teichoic acids, enhancing resistance to acidic conditions [3–5]. Another approach is the use of proton pumps such as the F₀F₁-ATPase, which pump excess protons out of the cell [6, 7]. Additionally, bacteria increase the concentration of alkaline compounds within the cell to counteract the acidification of the cytoplasm. For this, S. aureus increases the import of amino acids and osmolytes to maintain intracellular pH through decarboxylation reactions that consume protons [8, 9]. Moreover, ammonia production is also promoted, either from urea through urease activity or via the arginine deaminase system [7, 10]. Finally, energy metabolism and metabolic pathways are significantly reoriented. Respiration is favored over fermentative metabolism to limit additional acidification of the environment [8].

Global transcriptomic analyses have offered insights into how S. aureus adjusts its gene expression profiles to adapt to acidic environments, revealing the complexity of its stress response mechanisms [5, 7, 11, 12]. Despite these advances, the regulators that coordinate acid survival mechanisms remain partially characterized. Studies focused on a small number of transcriptional regulators provide only fragmentary insights into the network, yet they hint at its complexity. For instance, transcription of the operon encoding urease, a crucial enzyme in acid stress responses, was shown to be controlled by the global regulators CcpA, CodY, and Agr [13]. In addition, two-component systems (TCSs), which integrate a signaling sensor kinase and a response regulator, play important roles in adapting to acidic stress. For example, the GraRS TCS has been shown to be involved in resistance to antimicrobial peptides under acidic conditions [14] and to enhance S. aureus survival within the acidified phagolysosomes of macrophages [15]. In addition, other staphylococcal TCSs, such as KdpDE and VraSR, were also proposed to participate in the control of low pH survival [2, 8].

Although transcriptomic studies provide extensive information on acid stress responses, they often overlook post-transcriptional regulation. However, adapting to environmental changes often requires transcription factors to work in conjunction with post-transcriptional mechanisms mediated by small RNAs (sRNAs) [16, 17]. Most sRNAs are RNA molecules between 50 and 500 nucleotides in length and play crucial roles in various cellular processes, especially under specific growth and stress conditions, enabling faster regulatory responses than transcriptional mechanisms alone [18]. Typically, sRNAs act through antisense base pairing with the mRNA of their target genes [18–20]. While numerous studies highlight the involvement of diverse protein regulators, the role of sRNAs in orchestrating these adaptive responses remains less understood, despite emerging evidence suggesting that sRNAs are essential for fine-tuning responses to environmental stressors.

In this study, we analyzed transcriptomic responses of S. aureus under acidic pH, with a focus on sRNA expression. Our analysis showed that the sRNA RsaOI exhibits a marked increase in expression under acidic conditions. RsaOI was initially characterized as a putative regulatory RNA of ∼250 nucleotides in length in S. aureus strain N315 [21]. Transcripts corresponding to RsaOI have been identified through RNA-seq studies conducted across various genetic backgrounds and growth conditions [22–26]. Notably, rsaOI was found to be differentially expressed in S. aureus strains exhibiting vancomycin resistance (VRSA) or intermediate susceptibility (VISA) when exposed to different antibiotics [27]. These findings initially suggested a potential role for RsaOI in glycopeptide resistance. This hypothesis was further supported by evidence showing strong rsaOI induction in response to vancomycin, as well as the identification of atl, which encodes a major cell wall autolysin, as one of its molecular targets [28]. We further investigated the regulatory mechanisms governing RsaOI expression and characterized its targetome. Our data reveal that the TCS VraSR, known for its role in cell wall stress response, is activated under acidic conditions and induces the expression of RsaOI. RsaOI targetome analysis identified that this sRNA inhibits the expression of lacD. The latter is a part of the lactose utilization operon that is responsible for lactose and galactose metabolism. These findings suggest that RsaOI functions as a regulatory mediator, linking cell wall stress responses to metabolic adjustments, particularly in coordinating lactose metabolism under acid stress. This crosstalk highlights a regulatory strategy by which S. aureus fine-tunes its metabolic priorities to optimize growth and survival in challenging environments.

Materials and methods

Bacterial strains, plasmids, and growth conditions

The strains and plasmids utilized in this study are detailed in Supplementary Table S1. Supplementary Table S2 lists all the primers used. Unless otherwise specified, S. aureus strains were grown in brain heart infusion medium (BHI, Oxoid) at 37°C with shaking (160 rpm) or on BHI agar (Oxoid) at 37°C. Escherichia coli DH5-α strains were grown in lysogeny broth (LB) at 37°C with shaking (160 rpm) or on LB agar plates at 37°C. For the maintenance of plasmids or resistance cassettes, antibiotics were used at the following concentrations: for S. aureus, erythromycin and chloramphenicol at 10 μg/ml; and for E. coli, ampicillin at 50 μg/ml. The purified plasmids were used for the transformation in S. aureus RN4220 strain by electroporation. The φ80 phages prepared from RN4220 were then used to transduce plasmids in HG003 strains. pIMAYΔlacR vector is a pIMAY [29] derivative containing the PCR (polymerase chain reaction)-amplified lacR upstream and downstream sequences cloned by Gibson assembly (using primers from Supplementary Table S2) as described [30]. The lacR gene was deleted in S. aureus HG003 strain using pIMAY as described previously [30]. Other S. aureus mutants were constructed by transducing the erythromycin resistance cassette from the Nebraska Tn mutant library USA300 [31] into the HG003 strain using φ80 phage.

To induce acidic stress, cells were grown for 2–6 h in an LB medium supplemented with 100 mM HEPES to maintain pH to 7. Cells were then divided, pelleted by centrifugation at 3000 rpm for 5 min, and resuspended in fresh LB medium buffered with HEPES to pH 5 or 7 for 30 min.

For metabolic tests, S. aureus strains, wild type (WT), ΔrsaOI (both carrying pICS3 plasmid), or ΔrsaOI complemented strain (carrying pICS3-PamiA-rsaOI plasmid), were cultured at 37°C in LB or NZM broth and then diluted to 1:100 ratio in fresh media. When necessary, the media were supplemented with glucose or galactose at a concentration of 11 mM and HEPES at 100 mM. Growths were measured by a Biotek microplate reader.

Construction of gfp–reporter transcriptional and translational fusions

pCN33- and pCN38-based vectors containing gfp–reporter fusions were constructed using primers from Supplementary Table S2. gyrB sequence of pCN33-PtufA-gyrB-gfp vector was replaced by 5′ lacD sequence to produce pCN33-PtufA-lacD-gfp. For the pCN38-PrsaOI-gyrB-gfp construction, rsaOI promoter region (−99 nucleotides to +39 after the rsaOI transcriptional start) was amplified from S. aureus HG003 genomic DNA by PCR. The rsaOI promoter fragment was placed instead of PtufA promoter of vector pCN33-PtufA-gyrB-gfp. All cloning experiments were performed with Gibson Assembly Master Mix (New England Biolabs). The reactions were then transformed into E. coli DH5-α by heat shock at 42°C. The constructs were confirmed by DNA sequencing, and then electroporated into RN4220 before being transduced into HG003 or its derivatives using bacteriophage Φ80. When necessary, S. aureus HG003 strain was used to co-transform the lacD−gfp fusion vector with the rsaOI expressing plasmid. Cultures of these co-transformed S. aureus strains were grown at 37°C in LB or BHI supplemented with 10 μg/ml chloramphenicol and erythromycin. Fluorescence and OD₆₀₀ measurements were driven by a Biotek microplate reader as previously described [32]. To induce rsaOI expression, cells were grown in LB medium for 2 h, and then divided, with one sample receiving 10 μl of HCl 37% per 10 ml of culture.

Induction of acidic stress for acid susceptibility tests and proteomic studies

To induce rsaOI expression by acidic pH, cells were grown for 2 h in an LB medium supplemented with 100 mM HEPES to maintain pH at 7. Cells were pelleted by centrifugation at 3000 rpm for 5 min and resuspended in a fresh LB medium buffered with HEPES to pH 5 for an additional 2 h. After stress induction, cells were collected and pelleted by centrifugation for proteomic studies or plated on BHI agar plates for CFU (colony forming unit) counting.

Antibiotic susceptibility tests

Ten-fold serial dilutions of overnight cultures of WT HG003, ΔrsaOI, or complemented strain (ΔrsaOI/p-rsaOI) were plated and incubated for 24 h at 37°C on TSA (Tryptic soy agar) or TSA supplemented with 1 μg/ml vancomycin. Spot population analysis profile (spot PAP) assay allows rapid, sensitive, and reproducible qualitative and quantitative testing of antibiotic resistance. For this, 5 µl of each dilution was dropped on TSA or TSA supplemented with 1 μg/ml vancomycin and then incubated for 24 h at 37°C. Dilutions were deposed from top to the most concentrated (without dilution) to the dilution 10⁻⁶.

RNA sequencing method

Total RNA extraction was performed as previously described [33] on three independent replicates. RNA quality was then evaluated on a High Sensitivity RNA Screen Tape chip (Agilent). Ribosomal RNA depletion, complementary DNA (cDNA) library preparation, and sequencing were performed by Genewiz^® platform using their strand-specific RNA-seq protocol. The resulting FASTQ raw data were imported into the Galaxy web platform, using the public server at https://usegalaxy.eu/ for data analysis. Reads were cleaned using Trimmomatic (version 0.39) and mapped to the S. aureus NCTC8325 reference genome, including the file containing sRNA genes annotated for NCTC8325 (https://srd.genouest.org/browse/NCTC8325), using the BWA (version 0.7.18) algorithm. Counting files were generated with HTseq-count (version 2.0.5) and differentially expressed transcripts were identified using DEseq2 (version 1.40.2). Transcripts and genes with a P-value <.05 and a fold change >2 were considered significantly differentially expressed across biological replicates.

Preparation of samples for proteomic studies

Bacteria were prepared as described in the “Induction of acidic stress for acid susceptibility tests and proteomic studies” section. After pH induction, samples were centrifuged at 4500 rpm for 15 min at 4°C, and washed twice with cold phosphate buffered saline. The pelleted cells were then resuspended in 500 μl of Tris–HCl (pH 8) supplemented with a mini cOmplete (Roche) protease inhibitors cocktail, 2 U of DNAse, and 2 U of RNAse. Bacteria were lysed using a Fast-Prep system and then centrifuged for 10 min at 12 000 rpm at 4°C. Protein concentration was measured using a Qubit assay, and 500 μg was precipitated with 17.5% TCA overnight at 4°C. The samples were centrifuged at 9000 rpm for 15 min at 4°C, and supernatant was discarded. The pellets were rinsed twice with two volumes of cold acetone (100%) and mixed by inversion, followed by the same centrifugation as described before. Samples were dried in a speed vacuum and stored at −80°C.

Protein solubilization and quantification

TCA-precipitated proteins were then solubilized in 200 μl of R2D2 denaturing buffer {7 M urea, 2 M thiourea, 2 mM tri-N-butylphosphine, 20 mM dithiothreitol (DTT), 0.5% (w/v) 3-(4-heptyl)phenyl-3-hydroxypropyl dimethylammoniopropanesulfonate (C7BzO), 2% (w/v) 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate hydrate} and sonicated for 30 s. Protein concentrations were evaluated by Bradford analysis (Bio-Rad). Samples were stored in aliquots (30 μg) at −20°C until further use.

Protein digestion

Twenty-five micrograms of proteins were mixed with SDS (sodium dodecyl sulfate) loading buffer [62 mM Tris–HCl, pH 6.8, 20% glycerol (v/v), 0.04% bromophenol blue (w/v), 0.1 M DTT, SDS 4% (w/v)], heated at 95°C for 5 min, and then loaded onto a large SDS–PAGE (sodium dodecyl sulfate–polyacrylamide gel electrophoresis) stacking gel 7%. An electrophoresis was performed (10 mA, 2 h) to concentrate proteins. After migration, gels were stained with Coomassie Blue G250 and destained [50% ethanol (v/v) and 10% acetic acid (v/v)]. The revealed protein band was excised, completely destained, and washed three times with water. Samples were then reduced with 10 mM DTT for 1 h at room temperature and alkylated with 15 mM iodoacetamide for 45 min in the dark. Gel bands were treated with 50% acetonitrile (ACN)/50% ammonium bicarbonate 10 mM, pH 8 (two times, 5 min), and dried with 100% ACN (three times, 10 min). Then, proteins were digested with trypsin (1 μg per band, ratio of 1:25), overnight at 37°C, in 10 mM ammonium bicarbonate, pH 8. Peptides were extracted with 100% ACN (three times, 10 min) and then dried using a Speedvac concentrator (SPD111V, Thermo Fisher Scientific) and stored at −20°C. For each type of bacteria (WT and mutants), five biological replicates were carried.

LC–MS/MS analysis

Peptides were solubilized in 0.1% formic acid (FA) (v/v) and quantified using Pierce quantitative colorimetric peptide assay (Thermo scientific). Peptides (0.2 μg) were subjected to quantitative LC–MS/MS analysis on a high-resolution Orbitrap Eclipse Tribrid mass spectrometer coupled to a Proxeon Easy nLC 1200 (Thermo Scientific). Samples were injected onto an enrichment column (Acclaim PepMap C18, 2 cm × 75 μm, 100 Å, Thermo Scientific). The separation was performed with an analytical column needle (Acclaim PepMap C18, 25 cm × 75 μm, 2 μm, 100 Å, Thermo Scientific). The mobile phase consisted of H₂O/0.1% FA (buffer A) and CH₃CN/0.1% FA (80/20, buffer B). Tryptic peptides were eluted at a flow rate of 300 nl/min using a three-step linear gradient: from 2% to 40% B over 123 min, from 40% to 100% B in 1 min, and 10 min at 100% B. The mass spectrometer was operated in positive ionization mode with capillary voltage and source temperature set at 1.9 kV and 275°C, respectively. The samples were analyzed using higher-energy collision dissociation (HCD) method. The first scan (MS spectra) was recorded in the Orbitrap analyzer (R = 120 000) with the mass range m/z 400–1800. Then, 20 scans were recorded for MS2 experiments. Singly charged species were excluded for MS2 experiments. Dynamic exclusion of already fragmented precursor ions was applied for 30 s, and an exclusion mass width of ±10 ppm. Peptide isolation was achieved in the quadrupole with an isolation window of 1.6 m/z. Fragmentation occurred with an HCD collision energy of 28%. Daughter ions were analyzed in the Orbitrap with a resolution of 15 000. The maximum injection times were 30 and 22 ms for MS and MS2 analyses, respectively. All measurements in the Orbitrap analyzer were performed with on-the-fly internal recalibration (lock mass) at m/z 445.12002 (polydimethylcyclosiloxane).

Protein quantification

Raw data were imported in Progenesis LC–MS/MS software (Waters, ver. 4.1, UK). A two-dimensional map was generated for each sample (retention time versus m/z ratio). The spots present on the 2D maps were then aligned. Monocharged ions and those with a charge >5 were excluded from the analysis. MS/MS spectra from selected peptides were exported for peptide identification with Mascot (Matrix Science, ver. 2.2.04) against the S. aureus NCTC 8325 database available online (https://www.ncbi.nlm.nih.gov/datasets/taxonomy/93061/). Database searches were performed with the following parameters: one missed trypsin cleavage site was allowed; variable modifications: carbamidomethylation of cysteine and oxidation of methionine. Mass tolerances for precursor and fragment ions were both set at 5 ppm. Then, Mascot search results were imported into Progenesis. Peptides with an identification score >13 are retained. Proteins identified with <2 peptides were discarded. For each condition (WT and mutants), the total cumulative abundance of the protein was calculated by summing the peptide abundances. The software compared the intensity of the isotopic mass of all the ions for each condition. The abundances were normalized to perform a relative quantification of each protein between the conditions. The MS proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD060402.

Statistical analyses of proteomic data

Significance analyses of differentially regulated proteins were performed with Perseus software (version 2.0.10.0.). Proteins normalized abundance (from Progenesis) were transformed to log₁₀ space. Significantly regulated proteins between the different strains (WT, mutant, and complemented mutant) were identified by two sample t-test with a P-value ≤.05 and volcano plot was generated.

Functional annotation and functional enrichment analysis

The proteins were functionally annotated via NCBI, UniProt, and KEGG (Kyoto Encyclopedia of Genes and Genomes) functional resources and added to the Perseus software. The significantly up- and downregulated proteins from the three different comparisons were tested for functional enrichment based on KEGG pathways by using the Fisher’s exact test. Annotation terms were considered significant on the basis of an enrichment factor ≥1, an intersection size >2, and a P-value ≤.2.

RNA extractions, northern blots, and RT-qPCR

The cells were collected and pelleted for 10 min at 4°C, 4500 rpm, and then resuspended in RNA lysis buffer consisting of 0.5% SDS, 20 mM sodium acetate, and 1 mM EDTA (pH 5.5). Total RNA extraction was performed as previously described [34]. Northern blot assays were conducted following established protocols [34]. Membranes were hybridized with specific ³²P-labeled probes (Supplementary Table S2) in ExpressHyb solution (Ozyme) and washed according to recommendations from the manufacturer. The membranes were exposed and scanned with a Typhoon FLA 9500 scanner (GE Healthcare). Image quantification was carried out using ImageQuant Tool 7.0. For reverse transcriptase quantitative polymerase chain reaction (RT-qPCR), the total RNA extraction samples were treated with the DNase I Amplification Grade Kit (Invitrogen) to remove any residual DNA. cDNA preparations and RT-qPCR experiments were conducted as previously described [30], with the gyrB gene serving as the normalization control.

In vitro transcription, RNA labeling, and gel retardation assays

Gel-shift assays were performed as described in [30]. RNAs were synthesized from PCR-generated DNA using MEGAscript T7 Kit (Ambion). The transcription template was amplified from HG003 genomic DNA and forward primers included T7 promoter sequences (Supplementary Table S2). Then, RNAs were labeled at 5′-end using [γ-³²P] ATP (Amersham Biosciences) and T4 polynucleotide kinase (Invitrogen). Both labeled and unlabeled RNAs were purified on a 5% acrylamide–urea gel and eluted in elution buffer (20 mM Tris–HCl, pH 7.5, 250 mM NaCl, 1 mM EDTA, 1% SDS) at 37°C. The RNA was then ethanol-precipitated, quantified using a Qubit fluorometer (Thermo Fisher Scientific), and stored at −80°C. Gel-shift assays were conducted as outlined in [30]. The RNAs were denatured in a solution of 50 mM Tris/HEPES (pH 7–7.5) and 50 mM NaCl for 2 min at 80°C, followed by refolding for 10 min at 25°C upon the addition of MgCl₂ to a final concentration of 5 mM. Reactions were carried out in a buffer containing 50 mM Tris–HCl (pH 7.5), 50 mM NaCl, and 5 mM MgCl₂ for 20 min at 25°C. Approximately 0.05 pmol of labeled RsaOI or RsaOI-del was incubated with varying concentrations of lacD mRNA. The samples were supplemented with 10% glycerol and loaded onto a native 4% polyacrylamide gel containing 5% glycerol. The gels were subsequently dried and visualized using a Typhoon FLA 9500 scanner (GE Healthcare).

Results

rsaOI expression is triggered by low pH, improving survival in acidic stress conditions

To identify sRNAs potentially involved in adaptation to a low-pH environment, we performed RNA-seq analysis, comparing the transcriptome profiles of the HG003 strain under neutral (pH 7) and acidic (pH 4) conditions. For this purpose, total RNA was extracted from two HG003 cultures in the exponential growth phase, either maintained at neutral pH or exposed to acidic stress for 30 min, as outlined in the “Materials and methods” section.

Several genes identified as part of the acid stress response regulon showed increased expression (Supplementary Table S3). For instance, the ure operon encoding urease subunits A, B, and C (ureA, ureB, and ureC), which plays a role in maintaining intracellular pH, was upregulated [13]. Urease activity is known to be the major acid-resistance mechanism [2]. Additionally, the kdpDE operon, which encodes a TCS involved in both acid and osmoprotection, was stimulated [35]. Genes associated with histidine import and metabolism, essential for growth at low pH, were also upregulated [8]. In agreement with previous studies on sudden inorganic stress, we observed a general reduction in the expression of ribosomal protein genes and pyrimidine ribonucleotide biosynthesis pathway, which likely reflects the lowered growth rate after acidification [12]. Among the genes affected by low pH, rsaOI showed one of the most significant responses, with acidic conditions leading to more than a 14-fold accumulation of RsaOI (Fig. 1A and Supplementary Table S4). RsaOI is ∼250 nucleotides long and represents the longest sRNA that we previously identified in strain N315 [21]. Using RACE (rapid amplification of cDNA ends), we mapped the 5′-end of rsaOI from HG003 strain to position 565901 in the NCTC8325 genome, which corresponds to the location in N315 [21]. The rsaOI gene is present in all analyzed S. aureus strains, exhibiting 100% sequence conservation across isolates (Supplementary Fig. S1). Moreover, this sRNA is also found in other staphylococcal species, including Staphylococcus schweitzeri and Staphylococcus argenteus, with a variable sequence conservation (98.4% and 95.7% sequence identity, respectively) (Supplementary Fig. S1). In S. aureus and related staphylococcal species, compensatory mutations confirm that the last 30 nucleotides of RsaOI form a hairpin structure, likely functioning as a rho-independent transcription terminator. This is consistent with results showing that rsaOI transcription is not affected by rho deletion [26]. Moreover, in HG003 strain, rsaOI has its own promoter and transcription terminator, with no open reading frame on the opposite strand. This sRNA can therefore be classified as a bona fide sRNA [36]. rsaOI gene is located between proP gene, encoding for an osmolyte transporter that facilitates bacterial adaptation to osmotic stress [37], and operon vraABC, involved in fatty acid metabolism and shown to be upregulated in staphylococcal strains with increased resistance to vancomycin [38] (Fig. 1B).

Figure 1.

Induced by low pH, rsaOI contributes to survival under acidic stress. (A) Visualization of the most significantly accumulated mRNAs under acidic (pH 4) and neutral (pH 7) conditions using a volcano plot. The genes with the highest differential expression are highlighted and labeled. (B) Genomic localization of rsaOI gene in HG003 strain, situated between proP and vraABC operon. (C) RsaOI expression in response to pH range was analyzed by northern blot analysis using rsaOI-specific probe, with tmRNA used as loading control. (D) rsaOI expression in HG003 WT, ΔrsaOI (both carrying pICS3 plasmid; WT/p and ΔrsaOI/p), or complemented strain (ΔrsaOI, carrying pICS3-rsaOI plasmid) was analyzed by qPCR, normalized to the control gene gyrB, and calculated using 2^−ΔΔCt method for relative quantification. (E) Strains from pannel (D) were cultured in LB medium for 2 h, and then pelleted and resuspended in LB buffered to pH 5. After 2 h at pH 5, samples were collected and plated to determine CFU counts. Statistical analysis was conducted using Student’s t-test. Error bars represent the average of three independent experiments. Statistical significance is indicated by bars and asterisks as follows: *P< .01, **P< .05.

To validate the results obtained from the RNA-seq analysis of the effect of pH on rsaOI expression, northern blot and RT-qPCR analyses were performed. They both showed that the increase in RsaOI levels correlated with a decrease in the pH of medium: the highest accumulation of RsaOI was observed at pH 5 and 4, corresponding to 200- and 400-fold increases, respectively, compared to RsaOI levels at pH 7 (Fig. 1C and Supplementary Fig. S2). This supports the low pH-dependent expression of rsaOI.

We monitored the rsaOI expression profile during growth and found that rsaOI expression increased significantly during the stationary growth phase in nutrient-rich media in HG003. This pattern was conserved across other tested strains (N315, Newman, and USA300; Supplementary Fig. S3A and B), these data align with the highest expression of rsaOI at stationary phase reported for rsaOI from N315 strain grown in nutrient-rich media BHI [21]. We hypothesized that RsaOI increased levels were due to acidification of the medium as a result of glucose consumption during bacterial growth [5]. Consequently, rsaOI expression was analyzed in LB medium with or without glucose, and under buffered pH conditions preventing acidification of the medium. Growth of S. aureus in LB medium with glucose led to a progressive increase in RsaOI levels, which was prevented by buffering the medium to neutral pH (Supplementary Fig. S4A and B). rsaOI expression was also assessed under stress conditions encountered by bacteria during infection. Among the conditions tested, only the acidic stress significantly induced rsaOI expression, either by inorganic (HCl) acids or by organic (acetic acid) acids (Supplementary Fig. S5). These findings support that rsaOI expression is induced by low pH, regardless of the origin of pH lowering.

We further investigated the role of RsaOI in S. aureus acid resistance. To assess this, HG003 WT and rsaOI deleted (ΔrsaOI) strains were exposed to acid stress at pH 5, followed by plating to measure viability. Additionally, the rsaOI deletion was complemented with high-copy plasmid pICS3-rsaOI, expressing a WT copy of the rsaOI gene under the control of its native promoter (ΔrsaOI/p-rsaOI) (Fig. 1D). CFU counts revealed that rsaOI expression significantly improved survival rates compared to the ΔrsaOI strain (Fig. 1E).

rsaOI expression is affected by envelope stress and regulated by the VraSR two-component system

To explore the mechanism underlying RsaOI accumulation, we tested whether the increase in rsaOI expression under acidic conditions was due to its promoter activation. For this, we constructed a plasmid with GFP under control of rsaOI promoter (pCN38-PrsaOI-gfp). As a control, a plasmid with GFP regulated by the tufA promoter was employed (pCN38-PtufA-gfp). Acidic conditions did not affect tufA promoter activity but enhanced the fluorescence associated with the rsaOI promoter (Fig. 2A). To further validate this finding, we created a plasmid with the rsaOI expression driven by the constitutive promoter amiA (pICS3-PamiA-rsaOI). Expectedly, a decrease in pH did not impact RsaOI levels when rsaOI was under the control of the constitutive promoter, whereas there was an induction of rsaOI expression when regulated by its native promoter (pICS3-rsaOI) (Supplementary Fig. S6A). Additionally, we found that the decrease in pH did not affect RsaOI stability (Supplementary Fig. S6B). Taken together, these results indicate that the increase in RsaOI levels under acidic conditions is driven transcriptionally by enhanced rsaOI promoter activity.

Figure 2.

VraR affects rsaOI expression. (A) Fluorescence levels of GFP under control of either tufA or rsaOI promoters. Cells were grown for 2 h in LB medium, followed by addition of HCl. All statistical analyses were performed using Student’s t-test. Error bars represent the average of three independent experiments. Statistical significance is indicated by bars and asterisks as follows: ***P< .001. (B) Analysis of rsaOI levels in HG003 and Sa564 WT strains (WT) and their isogenic mutants. Relative quantification of rsaOI expression levels was measured by qPCR, normalized to the control gene gyrB, and calculated using 2^−ΔΔCt method. (C) rsaOI expression in WT and in ΔvraR mutant according to pH. Cells were cultured in LB medium for 2 h, and then pelleted and resuspended in LB buffered with 100 mM HEPES at pH 5 or 7 for 30 min. (D) Fluorescence levels of GFP under control of rsaOI promoter in both HG003 WT and ΔvraR strains. Cells were grown for 2 h in LB medium, followed by HCl addition. Statistical analysis was performed using Student’s t-test. Error bars represent the average of three independent experiments. Statistical significance is indicated by bars and asterisks as follows: ***P< .001, ****P< .0001. (E) Illustration of the induction mediated by the TCS VraSR on the expression of the sRNA rsaOI during acid stress or in the presence of the glycopeptide antibiotic vancomycin.

Interestingly, previous studies have reported that rsaOI expression was also increased in the presence of vancomycin [27, 28], but the regulatory network involved was not identified, although involvement of VraSR system was supposed [28]. Therefore, to identify the factor responsible for RsaOI accumulation under acidic conditions, we tested a vraR mutant together with several mutants lacking transcription regulators and TCSs for rsaOI expression. All tested strains except for the vraR mutant showed elevated rsaOI expression under acidic conditions (Fig. 2B). The absence of VraR prevented the accumulation of RsaOI at both pH 5 and 7, indicating that VraR is required for rsaOI expression (Fig. 2C). Similarly, a vraS mutant yielded the same results, further supporting the role of this TCS in rsaOI regulation (Supplementary Fig. S7A). VraS functions as a histidine kinase, while VraR is the response regulator of the VraSR (VraR/VraS) TCS that governs the staphylococcal response to perturbation in cell wall synthesis [39]. This TCS was initially identified for its role in vancomycin resistance [40].

To test whether acidic stress activates the VraSR system, we evaluated the expression of vraX and cwrA genes, which are directly regulated by VraSR [41, 42]. In accordance with transcriptomic data (Fig. 1A), the expression of both vraX and cwrA was upregulated under acidic stress in WT HG003 strain (Supplementary Fig. S7B and C). However, the expression of both genes was strongly downregulated in the vraR mutant (Supplementary Fig. S7B and C). Those data indicate that cwrA and vraX are part of an acid stress regulon directly regulated by VraSR. Additionally, we observed that acidic conditions led to higher vraSR mRNA levels, consistent with reported findings [12] (Supplementary Fig. S7D).

To confirm that the rsaOI regulation by VraSR occurs via its promoter, HG003 Wt strain and ΔvraR mutant were transformed by plasmid containing GFP reporter driven by the rsaOI promoter (pCN38-PrsaOI-gfp). No induction of fluorescence under acidic condition was observed in ΔvraR mutant in contrast to the WT strain promoter (Fig. 2D). Altogether, our results demonstrate that in addition to responding to cell wall stress caused by previously reported cell wall-damaging agents [43], VraSR TCS mediates response to low acidic growth conditions (Fig. 2E) and induces the transcription of rsaOI in response to a decrease in external pH.

RsaOI impacts S. aureus susceptibility to glycopeptide antibiotics

Since rsaOI was identified as part of the regulon of VraSR, a TCS involved in response to cell wall stress and in resistance to cell wall active antibiotics, we investigated whether RsaOI affected S. aureus susceptibility to glycopeptide antibiotics. For this, we compared the rsaOI mutant strain and its parental WT strain HG003 for susceptibility to vancomycin. Additionally, rsaOI deletion was complemented with a plasmid pICS3-rsaOI. While no differences in growth were observed between the strains in the absence of vancomycin, the loss of rsaOI led to improved growth in the presence of the vancomycin, compared to the WT strain (Fig. 3A and B). Complementation of the rsaOI deletion restored the reduced growth in the presence of vancomycin, confirming that RsaOI contributes to S. aureus susceptibility to this antibiotic. Similar results were obtained using a spot PAP assay (Supplementary Fig. S8) [44].

Figure 3.

RsaOI modulates S. aureus resistance to vancomycin. Ten-fold serial dilutions of overnight cultures of WT HG003, ΔrsaOI, or complemented strain (ΔrsaOI/p-rsaOI) were plated and incubated for 24 h at 37°C. The 10⁻⁶ dilution are shown for TSA condition (A), and the 10⁻⁴ dilution for TSA supplemented with 1 μg/ml vancomycin (B). Diagrams show colony growth quantification. Statistical analysis was performed using Student’s t-test. Error bars represent the average of three independent experiments. Statistical significance is indicated by bars and asterisks as follows: ***P< .001.

Identification of proteins with altered abundance in response to RsaOI

Altogether, these RsaOI features prompted us to investigate its targets and regulatory mechanism. To identify RsaOI-dependent changes in the abundance of individual proteins, we performed quantitative proteomics analysis of proteins whose levels were affected by RsaOI expression. For this purpose, we compared the proteome of HG003 WT, ΔrsaOI containing an empty vector (WT/p versus ΔrsaOI/p, respectively), and ΔrsaOI complemented with the pICS3-rsaOI vector (ΔrsaOI/p-rsaOI). To ensure rsaOI expression from its native promoter in the HG003 WT and complemented strains, bacteria were grown under low-pH conditions (Supplementary Fig. S9A). From 2573 annotated putative proteins in S. aureus NCTC 8325 strain [45], 1112 proteins were identified in HG003 WT, 1012 in ΔrsaOI, and 996 in rsaOI complemented strain. Tryptic peptides were analyzed by nLC–MS/MS (n = 4 per condition). We performed a functional enrichment analysis based on the significantly up- or downregulated proteins in comparison to ΔrsaOI/p versus WT/p and ΔrsaOI/p versus ΔrsaOI/p-rsaOI. Overall, carbon metabolism proteins were upregulated in the ΔrsaOI strain compared to the WT or rsaOI complemented strains (Supplementary Fig. S9B and C). For example, among these proteins, we found Fda (fructose-1,6-bisphosphate aldolase) involved in glycolysis and SucA (2-oxoglutarate dehydrogenase E1 component) in the TCA cycle, as well as LacD (tagatose-1,6-diphosphate aldolase) as part of galactose metabolism. This finding could be relevant in light of the impact of metabolism on medium acidification.

Focusing on proteins with differential abundance between deletion strain ΔrsaOI/p and complemented strain ΔrsaOI/p-rsaOI using t-test (P-value ≤.05), we highlight that a total of 129 proteins were significantly differentially regulated, including 30 proteins with a fold change >2 (21 up- and 9 downregulated in ΔrsaOI strain; Fig. 4A and Supplementary Table S5). Interestingly, the amount of Atl, which was recently shown to be repressed by RsaOI [28], was found to be increased in ΔrsaOI strain (Supplementary Tables S5 and S6). Notably, the transcriptional regulator of RsaOI, VraR, was found to be less abundant in the strain expressing rsaOI. The inhibition of VraR protein levels is particularly noteworthy given the significantly reduced survival of a strain expressing rsaOI in the presence of glycopeptides. Altogether, while the proteomic analysis does not elucidate the underlying regulatory mechanisms, it underscored a substantial impact of RsaOI on the levels of numerous proteins under acidic conditions.

Figure 4.

Investigation of the potential direct targets of RsaOI based on in vivo proteomic data and in silico computational approaches. (A) Volcano plot showing the protein relative abundances between the strains ΔrsaOI/p (ΔrsaOI containing a pICS3 vector) versus ΔrsaOI/p-ΔrsaOI (complemented strain) in relation to the t-test P-value. VraR and LacD are significantly more abundant in ΔrsaOI strain. (B) Venn diagram illustrating the overlap between experimentally identified and in silico predicted RsaOI target candidates. Experimentally identified candidates, representing proteins with significantly different abundances between rsaOI mutant and complemented strain, the top 100 predicted targets from CopraRNA [46] and the top 100 predicted by RNA predator [47] are shown. Shared targets are represented in the overlapping regions. (C) Organization of the lac operon in the HG003 strain and the predicted base-pairing interaction between RsaOI and lac operon mRNA, generated using IntraRNA software [49]. The start codon of lacD is highlighted in red, and position +1 corresponds to the transcription start site for RsaOI.

Integration of experimental data and computational prediction highlights potential direct targets of RsaOI

The effect of sRNAs on the protein abundance can occur through sRNA pairings with their corresponding mRNA or indirectly through the activity of sRNAs on regulatory proteins. To identify potential direct targets of RsaOI among candidates identified by proteomic approach, an in silico analysis was conducted using CopraRNA and RNApredator, which predict sRNA–mRNA interactions [46, 47]. These software tools generated a list of putative sRNA targets based on both RNA accessibility and evolutionary conservation of the interaction. Among the top 100 predicted targets for each computational tool, 32 potential interactions were predicted simultaneously by both methods probably because of using different prediction algorithms (Fig. 4B) [48]. For further investigation, we focused on predicted targets that overlapped between the in silico predictions and our proteomic data. Notably, we identified six mRNA targets: five predicted by CopraRNA and two by RNApredator (Fig. 4B and Supplementary Fig. S10A). Interestingly, among these potential targets, vraSR mRNA, encoding the transcriptional regulator of rsaOI, was predicted to interact with RsaOI by CopraRNA (Fig. 4B). The predicted interaction is located in vraS coding sequence, 87 nucleotides upstream of translational initiation site of vraR (Supplementary Fig. S10B).

Regarding the potential regulation of VraSR by RsaOI, we hypothesized that differences in glycopeptide survival could result from RsaOI-mediated inhibition of VraR level. To test this, a vancomycin survival assay using vraR deletion strains was conducted, with or without RsaOI expression driven by the constitutive amiA promoter (Supplementary Fig. S11). The deletion of vraR significantly increased bacterial susceptibility to vancomycin, as previously described. Notably, rsaOI expression in a vraR deletion mutant further affected survival, indicating that RsaOI influences vancomycin tolerance through the regulation of an additional target.

The Venn diagram analysis revealed only a single hit between the two bioinformatics tools and the proteomics data: lacD (SAOUHSC_02 452) (Fig. 4B and C). The lacD gene is within the seven-gene lac operon (lacABCDFEG) involved in galactose metabolism through the tagatose pathway. Using the IntraRNA software [49] RsaOI, is predicted to pair with 25 nucleotides of the lacD mRNA, which includes its ribosome binding site (RBS) and start codon (Fig. 4C). It should be noted that this region is conserved in other staphylococcal species, such as S. argentus and S. schweitzeri, where rsaOI is also present (Supplementary Fig. S12). Altogether, the compilation of in vivo data showing that RsaOI reduces the LacD protein levels and in silico predictions of RsaOI binding at the lacD RBS suggested a post-transcriptional mechanism of lacD regulation involving RsaOI.

RsaOI controls expression of lacD at translational level

To confirm the in silico prediction of RsaOI binding the lac mRNA in the RBS region of lacD, the formation of the RsaOI–lacD mRNA duplex was evaluated by electrophoretic mobility shift assay. A 250-nt-long mRNA fragment of lac mRNA containing the predicted RsaOI binding site was produced. RsaOI was able to form a complex with the lacD mRNA fragment (Fig. 5). The specificity of this interaction was confirmed by the lack of binding interference from an excess of yeast transfer RNA (Fig. 5A). We constructed a mutant version of RsaOI (RsaOI-mut), which lacks 30 nucleotides predicted to be involved in the interaction site with lacD mRNA. RsaOI-mut was unable to form a duplex with lacD mRNA, indicating that these nucleotides are required for direct interaction (Fig. 5A and B).

Figure 5.

RsaOI inhibits lacD expression. (A) Complex formation between RsaOI and lacD mRNA was analyzed by native gel retardation assays. Gel shift assays of purified, labeled RsaOI and RsaOI-mut (0.1 pmol) were performed with increasing concentrations of lacD mRNA. (B) Quantification of complex formation from panel (A) was performed by ImageQuant Tools 7.0. (C) The effect of RsaOI on lacD expression was examined using gfp gene reporter assay. Staphylococcus aureus WT and ΔrsaOI strains containing the pCN33-PtufA-lacD-gfp fusion plasmid were co-transformed with pICS3, pICS3 expressing rsaOI or rsaOI-mut under control of its endogenous promoter, or rsaOI under constitutive promoter amiA. The fluorescent intensity was measured after 12 h of growth in LB medium adjusted to pH 6. Statistical analysis was conducted using a Kruskal–Wallis non-parametric test across all conditions, followed by Student’s t-test to determine significant differences among conditions. Error bars represent the average of four independent experiments. Statistical significance is indicated by bars and asterisks as follows: **P< .01, ***P< .001. (D) In parallel with the panel (C) experiment, strains were collected and pelleted, and relative quantification of lacD–gfp fusion mRNA expression levels was measured by qPCR, normalized to the control gene gyrB, and calculated using 2^−ΔΔCt method.

Next, the effect of RsaOI binding to the lacD RBS on lacD regulation was evaluated in vivo. For this, the sequence of lacD mRNA used for in vitro study of interaction was fused in-frame with GFP reporter and positioned under the control of the constitutive tufA promoter to bypass the regulation on transcriptional level. Plasmid expressing the lacD–GFP reporter fusion was introduced in both the HG003 strain and its isogenic derivative ΔrsaOI. rsaOI was expressed under control of its native promoter (pICS3-rsaOI) or a constitutive promoter (pICS3-PamiA-rsaOI). The effects of RsaOI on lacD expression were evaluated by measuring fluorescence intensity through quantitative microplate assays. To permit rsaOI expression under control of its native promoter, we cultured the cells in LB medium adjusted to pH 6, which allows the induction of rsaOI expression (Supplementary Fig. S2). Expression of rsaOI reduced the fluorescence produced by S. aureus cells containing lacD–GFP translational fusion (Fig. 5C). Notably, for rsaOI under control of endogenous promotor, this regulatory effect was also detectable, though less pronounced, under gradual acidification of BHI medium resulting from glucose metabolism (Supplementary Fig. S13A). The decrease of fluorescence was correlated with rsaOI expression level, since the high level of rsaOI expressed from a strong amiA promoter (Supplementary Fig. S13B) resulted in more pronounced fluorescent inhibition (Fig. 5C). Moreover, the expression of rsaOI-mut allele (Supplementary Fig. S13C) unable for lacD mRNA binding in vitro (p-rsaOI-mut) had no effect over the fluorescence, thereby confirming the specificity of the regulation. To note, that for unknown reasons, we were unable to introduce the vector expressing the rsaOI-mut under the control of the constitutive amiA promoter in S. aureus. To determine whether RsaOI affected lacD by altering its mRNA levels, we measured the amount of lacD–gfp mRNA depending on RsaOI expression (Fig. 5D). No significant effect of RsaOI on mRNA levels was noticed suggesting that RsaOI acts through an antisense mechanism, leading to translational repression of lacD.

RsaOI impairs galactose utilization in S. aureus

Galactose is usually metabolized via the well-known Leloir pathway [50], where d-galactose is converted through d-galactose-l-phosphate and d-glucose-l-phosphate to d-glucose-6-phosphate. However, S. aureus lacks the enzymes of the Leloir pathway and in this bacterium, as opposed to other staphylococcal species such as S. intermedius, S. saprophyticus, and S. xylosus, galactose is converted to d-galactose-6-phosphate, which is further metabolized through tagatose derivatives (d-galactose-6-phosphate → d-tagatose-6-phosphate → d-tagatose-1,6-diphosphate → d-glyceraldehyde-3-phosphate + dihydroxyacetone-phosphate) [51] (Fig. 6A). Thus, S. aureus metabolize galactose exclusively via tagatose pathway [52], which is mediated by the lac operon (Figs 4 and 6A).

Figure 6.

rsaOI overexpression impairs using of galactose as an external source of carbons. (A) Illustration of the tagatose pathway, a metabolic pathway that catabolizes galactose or lactose through the lac operon. (B) Growth curves of S. aureus HG003 WT, ΔrsaOI, or complemented strain expressing rsaOI under constitutive promoter amiA were cultured in NZM media supplemented with glucose or galactose at 11 mM. Growth was followed during 16 h using a Biotek microplate reader. Error bars represent the average of three independent experiments.

Since the tagatose pathway is the sole route for galactose utilization in S. aureus and regarding the effect of RsaOI on lacD expression, we evaluated an impact of RsaOI on bacterial growth in medium supplemented with galactose. We used NZM medium, which lacks external sugar, and NZM supplemented with either glucose or galactose (Fig. 6B). In these media, we compared the growth of the S. aureus HG003 WT, ΔrsaOI, and a strain that constitutively expresses rsaOI in a condition-independent manner (ΔrsaOI/p-PamiA-rsaOI). All strains showed similar growth in NZM alone. Additionally, the supplementation with glucose significantly enhanced growth of all strains. However, when galactose was added, we observed differences in growth depending on rsaOI. Indeed, galactose improved the growth of both WT and ΔrsaOI strains. In contrast, the strain constitutively expressing RsaOI exhibited no growth improvement with galactose compared to NZM alone (Fig. 6B). This suggests that expression of RsaOI inhibits galactose utilization likely due to the repression of the lac operon.

Discussion

Acid stress is a common challenge encountered by S. aureus throughout its life cycle. This stress can be either endogenous, triggered by sugar metabolism, or exogenous, such as exposure to host defenses or antimicrobials. To counteract the deleterious effect of pH lowering, S. aureus utilizes specific mechanisms and alters its transcriptome by upregulating the acid resistance regulon including genes involved in urea degradation, amino acid uptake, or factors that modify the membrane charge [2].

In this study, we examined gene expression changes in response to acid stress. In addition to identifying genes belonging to the characteristic acid stress response regulon [2], transcriptomic analysis revealed an sRNA, RsaOI, that accumulates significantly under acidic pH conditions. Further analysis showed that the expression of rsaOI is induced by low pH, regardless of whether inorganic or organic acid is used and whether the acidification is caused by external acid or results from glucose metabolisms. We identified VraR, the response regulator of the VraSR TCS, as a key activator of rsaOI expression under acidic conditions (Fig. 7). Specifically, VraR activates the rsaOI promoter in response to pH decrease. VraSR is homologous to the LiaSR TCSs found in other Gram-positive bacteria, such as Bacillus subtilis, Streptococcus mutans, and Enterococcus faecalis [53, 54]. In accordance with TCS from B. subtilis, Lactococcus lactis, and other Gram-positive bacteria, the staphylococcal system also contains a third component (VraT) corresponding to LiaF [55]. In response to an unknown signal, likely induced by exposure to cell wall-targeting agents, VraS undergoes autophosphorylation, leading to its activation and subsequent phosphorylation of VraR [43, 56]. The phosphorylated VraR dimerizes and binds to the promoter of its own operon (vraUTSR) as well as the promoters of numerous genes that constitute the so-called cell wall stress stimulon (CWSS) [57] (Fig. 7). Genes induced by this mechanism are involved in diverse cellular processes, including DNA replication and repair, carbohydrate metabolism, and, most notably, a substantial number of genes essential for cell wall biosynthesis and remodeling. This regulation enables an adaptive response to maintain cell envelope integrity under stress conditions [56]. However, both the stimuli that activate VraSR and the specific configuration of VraSR-regulated genes can vary significantly, depending on the strains and experimental procedures used [56, 58] (Fig. 7). While, in other Gram-positive bacteria, the liaSR homologs have been shown to be involved in acid stress response [55], the VraSR system of S. aureus was initially reported to be induced by cell wall-targeting antibiotics, including vancomycin, teicoplanin, and β-lactams, but not by general stresses, such as acid pH [43, 56]. However, our transcriptome and follow-up analyses revealed that two well-characterized direct VraSR targets, cwrA and vraX [59, 60], are significantly upregulated under acidic conditions in a VraR-mediated manner, demonstrating the activation of VraSR-mediated response by low pH. In line with this data, highlighting the role of VraSR in acid stress response, recent study has demonstrated that VraSR is essential for the growth of S. aureus at low pH [8]. Furthermore, in addition to vraS and vraR, numerous genes encoding proteins involved in cell wall assembly and maintenance were also found to be essential for growth at low pH, thus confirming the importance of cell wall in the acid stress response of S. aureus [8].

Figure 7.

RsaOI acts as a mediator between cell wall stress and the tagatose metabolic pathway. In response to stress generated by exposure to glycopeptides such as vancomycin or to acid stress, the VraSR system, involved in resistance to cell wall-targeting antibiotics, is activated. The sensor unit VraS phosphorylates VraR, thereby activating it, while it was initially kept inactive through phosphorylation by the Stk1 kinase [62]. Once activated, VraR induces the transcription of genes belonging to the CWSS. This stimulon can be divided into two subsets: the first includes genes induced exclusively upon glycopeptide exposure, primarily involved in cell wall synthesis and dynamics (e.g. pbp2, fmtA, sgtB,andmurZ); the second subset comprises genes induced both by glycopeptides and acidic conditions (e.g. cwrA,vraX, and rsaOI). Under these stress conditions, RsaOI represses the expression of its target genes: atl,ptsH (Hpr), and lacD. LacD encodes the tagatose pathway aldolase, which is essential for galactose metabolism. RsaOI represses lacD expression through base pairing at the RBS, thereby reducing tagatose pathway activity and fine-tuning sugar metabolism in response to environmental stressors.

Interestingly, the panel of genes stimulated by VraR varies depending on the perceived stimulus. Notably, acidic stress results in only partial induction of the VraSR-controlled regulon (Fig. 7). Our analysis, as well as previously published data on the transcriptional response to acid stress, did not reveal the complete set of VraSR targets that are typically induced in response to cell wall-inhibitory antibiotics [43, 54, 61]. The underlying mechanism driving the selective activation of specific genes within the regulon remains unclear, but it could involve differences in VraR phosphorylation dynamics [62], promoter architecture, or interactions with additional regulatory factors that modulate gene expression in a stimulus-dependent manner. Further studies are needed to elucidate how VraSR coordinates this nuanced regulatory response to diverse environmental challenges.

The identification of rsaOI as part of VraR regulon induced by pH lowering aligns with studies showing that expression of rsaOI is stimulated by vancomycin treatment [27, 28]. Consistent with these findings and the inclusion of rsaOI within the cell wall stress regulon, we demonstrated that RsaOI influences bacterial susceptibility to glycopeptide antibiotics. Moreover, the recent study demonstrated that RsaOI controls the expression of the autolysin atl involved in S. aureus cell division and cell wall turnover, emphasizing the role of RsaOI in cell wall stress response [28]. The same study used the CLASH approach to map the sRNA–mRNA interaction network in the vancomycin-intermediate S. aureus (VISA) strain JKD6008 and identified several RsaOI–mRNA interactions associated with the endoribonuclease RNase III [28]. In addition to atl, RsaOI was found to repress the expression of ptsH, encoding protein Hpr, involved in catabolite repression by interacting with CcpA [28] (Fig. 7). These findings support the role of RsaOI in metabolic regulation and suggest that RsaOI coordinates carbon metabolism and cell wall turnover during vancomycin treatment.

In our study, we analyzed the impact of rsaOI expression on the protein levels in acidic conditions. Consistent with previous findings [28], we identified that the deletion of rsaOI results in an increase in Atl protein levels (Supplementary Tables S5 and S6), thereby confirming repression of atl by RsaOI. However, proteomic analysis of strain in acidic conditions did not reveal changes in Hpr protein levels depending on rsaOI expression (Supplementary Tables S5 and S6). It is known that the genetic background of the strains and experimental conditions can significantly influence regulatory network in S. aureus and the outcome of sRNA regulation [63, 64]. Thus, the differences in RsaOI regulation observed between the two studies could be attributed to distinct stress conditions employed and variations in the genetic backgrounds of used strains as well as their differing levels of glycopeptide susceptibility. In our study, among the proteins whose levels depend on RsaOI, many are involved in carbon metabolism. Proteins such as Fda, SucA, and LacD exhibited increased levels in the ΔrsaOI strain, suggesting that RsaOI represses their expression. Notably, under our experimental conditions, RsaOI does not affect the levels of Hpr, a protein involved in catabolite repression, indicating that the observed changes in the levels of metabolism-related proteins are independent of Hpr. Altogether, the identification of numerous metabolism-related proteins regulated by RsaOI highlights the sRNA’s pivotal role in metabolic regulation.

It was shown that some sRNA–mRNA interactions recovered by in vivo proximity-dependent ligation do not affect mRNA transcript or protein abundance [65]. Nevertheless, some common and associated targets were revealed using both RNase III-CLASH and proteomic approaches (Supplementary Table S6 and Supplementary Fig. S14). In addition to previously mentioned atl, another mRNAs found to bind RsaOI by RNase III-CLASH analysis was also identified in our study: rocF mRNA, which encodes arginase, an enzyme that converts arginine to ornithine [66] (Supplementary Table S6 and Supplementary Fig. S14). Interestingly, although also identified by CLASH approach PxpB (SAOUHSC_01712), part of the multicomponent ATP-dependent 5-oxoprolinase complex PxpABC, involved in the degradation of 5-oxoproline, a spontaneous by-product of glutamine metabolism and other cellular processes [67], was not part of proteins detected in our proteomic analysis; however, two proteins encoded by the same operons (SAOUHSC_01710 and SAOUHSC_01711) were found to be affected by RsaOI expression, supporting the possibility of RsaOI-mediated regulation of the pxp operon. Notably, this operon has also been shown to be regulated by CcpA [68], suggesting a possible link with carbon metabolism. Additionally, proteomic analysis revealed that RsaOI reduces HemA protein levels. This finding is consistent with the RNase III-CLASH results, which showed RsaOI binding to hemX mRNA, a regulator of hemA expression [69]. Altogether, RNase III-CLASH capture of RNA–RNA complexes and comparative proteomic analysis under low pH uncovered a broad set of potential RsaOI targets, many of which are linked to cell wall turnover, stress response, and carbon metabolism. Further studies are needed to elucidate the mechanism of RsaOI action on these targets with connection to cell wall stress.

In this study, the juxtaposition of proteomic data with in silico analysis identified lacD as a direct target of RsaOI. RsaOI represses the expression of lacD by antisense pairing to its RBS, resulting in reduced LacD protein production. By downregulating lacD, which encodes for tagatose-1,6-diphosphate aldolase, an enzyme of the tagatose pathway, RsaOI modulates galactose metabolism, thereby impacting the bacterium’s capacity to assimilate galactose (Fig. 7). Notably, overexpression of rsaOI leads to a complete loss of ability to metabolize this sugar (Fig. 6B).

Bacteria frequently encounter diverse carbon sources in their environments and have developed mechanisms to prioritize the uptake and metabolism of substrates that support rapid growth and competitive advantage. Glucose is the preferred carbon source for most bacteria, which usually suppresses the utilization of other sugars in its presence. Furthermore, sugar uptake and metabolism must be carefully regulated to prevent harmful imbalances in metabolite levels and to avoid deleterious accumulation or depletion of metabolites [70, 71]. In S. aureus, transcription of the lac operon is reported to be regulated by multiple mechanisms, including control by the LacR repressor, catabolite repression [72], by glucose in a CcpA-independent manner [73], and it is shown to be induced by intracellular galactose-6-phosphate [74]. Here, we demonstrate that the expression of lac operon is also controlled at post-transcriptional level by RsaOI. Similarly, in Gram-negative bacteria, galactose metabolism is controlled not only at the transcriptional level but also post-transcriptionally by sRNA. Spot 42 sRNA binds to the galETKM operon mRNA, selectively inhibiting the translation of galK, and thereby causing discoordinate expression of the gal operon [75]. This underscores the critical role of post-transcriptional regulation in fine-tuning genes expression for efficient galactose metabolism.

By regulating lacD, RsaOI is responsible for the connection between cell wall stress response induced by VraSR and galactose metabolism (Fig. 7). In Gram-positive bacteria possessing the Leloir pathway, galactose is metabolized through this pathway leading to the formation of intermediates, UDP-glucose and UDP-galactose, which serve as precursors for cell wall biosynthesis and sugar substitution of lipoteichoic acid [76]. In S. aureus, which does not possess the Leloir pathway, galactose can be converted by the tagatose pathway into tagatose-6-phosphate, which is further processed to produce glyceraldehyde-3-phosphate and dihydroxyacetone phosphate. These intermediates enter central metabolic pathways (like glycolysis), contributing to the pool of building blocks needed for synthesizing essential biomolecules, including the peptidoglycan layers of the cell wall. Disruptions in central metabolism can lead to changes in peptidoglycan synthesis, impacting cell wall thickness and stability [77]. Moreover, the enzyme tagatose-1,6-bisphosphate aldolase, encoded by lacD, although exhibits its highest affinity for d-tagatose-1,6-bisphosphate but can also use other d-hexose bisphosphate stereoisomers as substrates, including sorbose bisphosphate, psicose bisphosphate, and fructose bisphosphate [78]. Fructose monophosphate participates in the formation of UDP-GlcNAc, which is also a precursor for the cell wall [79]. Additionally, LacD may possess moonlighting function beyond its role in tagatose pathway, as proteomics studies of E. coli identified d-tagatose-1,6-bisphosphate aldolase 2 on the cell surface [80]. Interestingly, several studies have reported links between the lac operon and cell wall-acting antibiotics, suggesting its potential role in peptidoglycan modification; however, the exact mechanisms remain to be elucidated [81–83].

In conclusion, in S. aureus, sRNAs were shown to play a role in the response to variety of environmental stresses, including pH changes, osmotic stress, and nutrient limitations [84]. Some of them, such as SprX and 6S RNA, have been found to influence the bacterial susceptibility to antibiotics [34, 85]. Here, we report RsaOI as an environment-sensitive sRNA, induced through VraSR TCS under cell wall stress. RsaOI modulates bacterial metabolism by regulating the tagatose pathway. Therefore, this study highlights the important role of sRNAs in maintaining cellular integrity and the overlapping responses to multiple stresses and carbon metabolism.

Supplementary data

Supplementary data is available at NAR online.

Conflict of interest

None declared.

Funding

S.M. thanks the Région Normandie for the engineer financial support. Europe gets involved in Normandy with European Regional Development Fund (ERDF). This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement no. 101034329. This project has received funding from the Normandy Region under the WINNING Normandy program. M.G., H.R., A.R., and S.C. thank University of Rennes for financial support (from “Défis émergents” program). Funding to pay the Open Access publication charges for this article was provided by INSERM U1230.

Data availability

The RNA-seq analysis data have been deposited to https://www.ncbi.nlm.nih.gov/sra/PRJNA1182076. The MS proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the date set identifier PXD060402.