https://orcid.org/0000-0002-8357-1380
Figures
Abstract
In Staphylococcus aureus, the quorum‑sensing accessory gene regulator (agr) system is the major virulence regulator. The four agr‑types (I-IV) have been associated with distinct infection outcomes, but their direct contribution to virulence regulation and metabolism has remained unresolved due to tight linkage between agr‑type and genetic background. To disentangle agr‑type‑specific effects, we used congenic Newman strains in which the native agr-locus has been replaced with each of the four agr-types, alongside a agr mutant. We performed RNA‑sequencing during early exponential (1h30), late exponential (6h), and stationary (12h) growth phases. Despite similar growth kinetics, agr‑types displayed distinct activation profiles based on agrA and RNAIII expressions. Agr-I and agr-IV showed early, strong expressions, agr-II displayed intermediate expressions and agr-III initiated weak expressions only in stationary phase. This agr‑type‑dependent activation timing was the dominant driver of global transcriptional changes. Early activation in agr‑I and agr‑IV induced robust expression of phenol‑soluble modulins, capsule biosynthesis genes, and pore‑forming toxins, whereas agr‑II and agr‑III expressed delayed or alternative virulence pathways, including upregulation of superantigen‑like genes. Among all types, agr‑IV exhibited the broadest transcriptional response, encompassing both virulence and metabolic pathways, including differential regulation of nucleotide and fructose metabolisms. Pairwise differential expression, over‑representation analysis, and gene‑set enrichment consistently revealed agr‑type‑specific virulence and metabolic programs. Agr‑III, which activated latest and weakest, showed limited transcriptional change until stationary phase, whereas agr‑I, agr‑II, and agr‑IV displayed progressively broader virulence and metabolic remodeling. Together, these findings demonstrate that agr‑type determines virulence and metabolic gene expression profiles primarily by dictating the timing and magnitude of agr-activation, even within an identical genetic background and growth environment. This work provides a systematic framework for understanding agr‑type‑specific regulatory strategies and their potential roles in S. aureus pathogenesis.
Citation: Pivard M, Bär J, Demeter T, Mairpady Shambat S, Zinkernagel AS (2026) Staphylococcus aureus agr-type vs genetic background: molecular signatures determining differential metabolism and virulence potential. PLoS One 21(6): e0350108. https://doi.org/10.1371/journal.pone.0350108
Editor: Herminia de Lencastre, The Rockefeller University, UNITED STATES OF AMERICA
Received: March 6, 2026; Accepted: May 9, 2026; Published: June 3, 2026
Copyright: © 2026 Pivard et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The raw reads for this study have been deposited in the European Nucleotide Archive (ENA) at EMBL-EBI under accession number PRJEB107700 (https://www.ebi.ac.uk/ena/browser/view/PRJEB107700).
Funding: A.S.Z. was supported by the Swiss National Science Foundation (SNSF) project grant 310030_204343 and M.P. by the Promedica Stiftung grant 84803, no website. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. There was no additional external funding received for this study.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Staphylococcus aureus is a commensal bacterium colonizing the human skin and nares [1], as well as a human pathogen causing life-threatening infections [2,3]. Its pathogenic potential is largely attributed to virulence factors, tightly regulated by networks, such as the accessory gene regulator (agr) system [4]. The agr is a quorum-sensing, two-component regulatory system classified into four types, based on agr-locus (agrBCDA) polymorphisms, affecting predominantly the autoinducing peptide (AIP) encoded by agrD and its sensor AgrC. The agr-locus is structured around two key promoters: P2, which governs expression of the agrBDCA operon, and P3, which controls production of RNAIII, the main effector of the system [5]. Each strain harbors a single agr-locus, and each clonal complex (CC) or sequence type (ST) typically encompass only one agr-type [6]. Previous studies reported associations between agr-type and infection severity [7–9]. However, assessing the specific contribution of agr-type to virulence is challenging due to the confounding effects of genetic background. To date, only two studies have compared agr-types within an identical genetic background, focusing on a limited set of agr-regulated targets [10,11]. To gain a comprehensive understanding of the impact of the agr-type on the global transcriptome, we performed RNA sequencing at multiple growth phases on congenic S. aureus strains representing each agr-type (agr-I, agr-II, agr-III, agr-IV) and a knock-out mutant (agr) in the same genetic background (Newman) [11]. Despite equal bacterial density, major differences in the kinetics of agr-activation between agr-I/IV and agr-II/III were observed, leading to differential expression of virulence factors and metabolic pathways.
Materials and methods
Bacterial strains and growth conditions
S. aureus congenic strains agr, agr-I, agr-II, agr-III and agr-IV, were previously constructed in the Newman genetic background by deletion of the native agr-locus and then complemented with any of the four agr-types [11]. Strains were cultured overnight in tryptic soy broth at 37°C, with shaking. Then, strains were subcultured 1:20 for 1h30 three times with double centrifugation-washing with phosphate-buffered saline in-between. The last subculture was incubated for 12h, and samples were collected at 1h30, 6h, and 12h. Bacteria were harvested by centrifugation and stored in RNAlater (Sigma Aldrich) prior to RNA extraction.
RNA extraction and sequencings
Bacterial pellets were resuspended in 10 mM Tris (pH8) with lysostaphin (Biosynth, 100 µg/mL), incubated for 1h at 4°C and 10 min at 37°C. Total RNA was extracted using the RNeasy Plus Mini Kit (Qiagen), followed by double TURBO DNase treatment (Invitrogen). RNA quality and concentration were assessed using the RNA ScreenTape Assay and TapeStation System (Agilent). Library preparation (TruSeq RNA Library Prep, Illumina) and paired-end 150 bp sequencing were performed by the Functional Genomics Center Zurich on an Illumina NovaSeq X Plus, yielding million reads per sample.
RNA-seq and statistical analysis
Reads were processed using a custom pipeline: https://zenodo.org/records/18492655. The steps included quality control and filtering, rRNA removal (SortMeRNA 4.3.7) and alignment to the S. aureus Newman reference genome (RefSeq GCF_000010465.1, bowtie2 2.5.2) and read summarization. DE analysis, ORA and GSEA were conducted in R (v4.5.2) using DESeq2 (v1.50.2, [12]) and clusterProfiler (v4.18.3, [13]). Full analysis scripts and RNAseq count files are available at: https://zenodo.org/records/19554265. Comparisons of agrA and RNAIII relative expression to gyrB were performed using GraphPad Prism (v10.6.1).
Artificial intelligence tools and technologies
The manuscript was written by the authors. Microsoft M365 Copilot (GPT-5.2, last accessed on 14th of April 2026) was employed for grammar correction, text streamlining and improving language clarity, and as a tool to explore alternative phrasings. Github Copilot (Claude Sonnet 4.5, last accessed at 14th of April 2026) was used for code readability and efficiency improvements and streamlining. The authors have thoroughly reviewed, verified, and edited any passages generated by any LLM, taking full responsibility for the manuscript and overall quality and accuracy.
Results
Agr-activation kinetics differ among agr-types
To assess specifically, how the agr-types influence S. aureus transcriptome, we used congenic strains, sequentially subcultured, to reset agr-activation and to reduce AIP accumulation. Samples were collected at 1h30 (early exponential phase), 6h (late exponential phase) and 12h (stationary phase). Since agr-activation in bacterial population is density-dependent, we confirmed that all strains grew similarly and had a comparable density (Fig 1A).
A. Bacterial density measured as CFU/mL for each strain and time point. Error bars represent the standard error of the mean (SEM). A minimum of four biological replicates were performed. B and C. Relative expression of agrA (B) and RNAIII (C) to gyrB using RNAseq counts (number of reads of agrA or RNAIII / number of reads of gyrB) for each strain except Newmanagr, as no reads could be detected for neither agrA or RNAIII. Error bars represent the standard error of the mean (SEM). Two-way ANOVA with interaction term of agr-type and timepoint on log10 relative expression and Turkey’s multiple comparisons tests were performed; adjusted p-value < 0.0001 - ****, < 0.05 - *; ns – not significant. Three biological replicates were performed.
We then assessed agrA and RNAIII transcription reflecting the agr-activation state, across strains and time points (Fig 1B-C and Panels A-B S1 Fig). Across the three time points, the four agr‑types showed distinct expression kinetics. At 1h30, all strains had low expression levels for both genes (Fig 1B-C). By 6h, agr‑I and agr‑IV showed high and significant expression of agrA and RNAIII, agr‑II displayed an intermediate induction (significant RNAIII increase compared to 1h30) while agr‑III showed no detectable induction (Panels A-B S1 Fig). At 12h, agr‑I, agr‑II, and agr‑IV displayed high agrA and RNAIII expression, while agr‑III reached intermediate expression levels, comparable with agr‑II at 6h (Fig 1B–C). Both gene expressions were significantly higher in agr-I, agr-II and agr-IV compared to agr-III (Fig 1B-C), and overall, only agrA expression in agr-III increased significantly between 1h30 and 12h (Panel A S1 Fig), consistent with a delayed activation. Overall, despite similar growth kinetics, agr-I and agr-IV activated earlier than agr-II, while agr-III did not show strong agr-activation compared to the other agr-types, even after 12h of growth.
Agr-activation shapes the transcriptomes
To evaluate agr-type-dependent transcriptome variation, we performed a principal component analysis (PCA) on the full dataset (three biological replicates for each time point), omitting the agr-locus to avoid agr-driven bias (Fig 2). Samples clustered primarily by time point, with strong separation between early exponential 1h30 and both 6h and 12h time points (Fig 2A and Panel A S2 Fig).
Dot-plots of the principal component analysis (PCA), with PC1 and PC2 axes, using the entire transcriptome beside the agr-locus, with all time points in A, at time point 1h30 in B, 6h in C and 12h in D. Solid (PC1) and dashed (PC2) lines mark the zero score for each principal component.
Next, we focused on each individual time point for further analysis. We did not find any clustering at 1h30 (Fig 2B and Panel B S2 Fig). For both 6h and 12h, the PC1 axis explained most of the variance, with 56% and 51% respectively (Fig 2C-D and Panels C-D S2 Fig), whereas the PC2 axis only explained 18% and 10% (Fig 2C-B). Hence, three separate clusters of agr-I, agr-IV and all other strains were observed at 6h (Fig 2C and Panel C S2 Fig). At 12h, three different clusters were found: i) agr-I with agr-IV, ii) agr-II samples displaying an intermediate profile, and iii) agr-III clustering with agr (Fig 2D and Panel D S2 Fig). Although variability within the clusters along the PC2 was observed, these findings, in regard with identified agr-activation profiles (Fig 1), suggest that the transcriptome is driven by both time and the agr-activation profile, rather than the agr-type alone, despite identical genetic background and comparable bacterial density of the cultures.
Agr-activation triggers specific virulence and metabolic profiles
To investigate how agr‑type shapes the different transcriptomes during growth, we first compared each agr‑type to the agr strain. At 1h30, none of the strains showed differential expression (DE) relative to
agr (S3 Fig and S1 Table), consistent with the PCA clustering (Fig 2B) and the low agr-activity at this time point (Fig 1B-C). At 6h, agr‑I and agr‑IV displayed strong upregulation of all phenol-soluble modulin (psm) genes (Fig 3A and 3D). Agr‑IV showed the most extensive transcriptomic shift, with 233 up‑ and 139 downregulated genes, relative to
agr (S2 Table). Upregulated loci included the capsule operon and several key virulence factors such as hlgCB, lukDE, multiple serine proteases, whereas many superantigen‑like genes (ssl) and amino‑acid metabolism pathways were downregulated (Fig 3D and S2 Table). In contrast, agr‑II and agr‑III showed only minimal changes, primarily low induction of psm genes, reflecting their limited agr-activation at this stage (Fig 3B-C and S2 Table). At 12h, agr‑III remained transcriptionally similar to
agr (Fig 3G), consistent with its slow agr-activation. In comparison, agr‑I, agr‑II, and agr‑IV strongly overexpressed all psm genes (Fig 3E, F, H and S3 Table). Capsule expression remained high in agr‑IV and expression increased in agr‑I as well (Fig 3E, H and S3 Table). Furthermore, agr‑IV displayed the broadest profile of DE, including unique downregulation of purine and pyrimidine metabolism genes (Fig 3H), while hlgCB, lukDE, and protease genes were not significantly induced at 12h anymore.
Volcano plots of the pairwise comparisons of the four agr-types compared to agr, at 6h (A-D) and 12h (E-H). Comparisons to agr-I in A and E, to agr-II in B and F, to agr-III in C and G, and to agr-IV in D and H. Significantly differentially expressed genes with a fold change of |2| (log2 fold change (LFC) of 1) and adjusted p-value < 0.05 are in red. Dashed lines indicate fold change threshold and adjusted p-value threshold (0.05). Gene names of top 15 genes are annotated according to Aureowiki Newman gene annotation; if only pangenome annotation was available “p:” prefix was added to the gene’s name, for genes without gene’s name the four digits following the “NWMN” tag were kept as gene ID. ns: not significant, FC: fold change, Padj: adjusted p-value.
After characterizing variation relative to agr, we analyzed the pairwise relationships within the agr-type strains agr-I, -II, -III and -IV. At 1h30, differences were limited, with only 30 significant DE genes with a fold change higher of |2| (S1 Table). At 6h, 411 significant DE genes (fold change of |2|) were found (S2 Table) and most DE genes (fold change of |4|) fell into three functional categories: virulence factors, capsule genes, and metabolic pathways (riboflavin, fructose, gluconate, S4 Fig). Agr‑IV upregulated capsule and psm genes stronger than other agr-types, while ssl genes were downregulated in agr‑IV but upregulated in agr‑II and agr-III. At 12h, transcriptional differences between agr‑types were weaker overall (149 DE genes), though agr‑IV maintained high capsule and psm expression, matched by agr‑I and, to a lower level, by agr‑II (S3 Table). Agr‑III displayed the lowest variability across time (S4 Fig and S3 Table). Over‑representation analysis (ORA) recapitulated these patterns (Table 1). Capsule biosynthesis was the most enriched pathway in agr‑IV at both 6h and 12h. Virulence‑related pathways, including toxins and secreted proteases, were likewise enriched among upregulated genes in agr‑IV. Notably, when comparing agr‑III and agr‑IV at 6h, the “S. aureus infection” pathway (sae05150) appeared as both up‑ and down‑regulated gene sets with distinct genes in each (Table 1). This indicates that agr‑III is not simply avirulent but expresses an alternative virulence gene panel, shaped by a delayed agr-activation. ORA also identified enrichment of metabolic pathways. At 6h, phosphotransferase system (PTS), fructose and histamine metabolisms were enriched in all agr-types compared to agr-I. While at 12h, nucleotides and arginine metabolisms were differently enriched depending on the agr-type, highlighting that agr‑type influences not only virulence but broader cellular physiology.
To detect coordinated transcriptional changes missed by single‑gene thresholds, we performed Gene Set Enrichment Analysis (GSEA) (S4 Table). GSEA showed early enrichment of metabolic pathways in agr‑II and agr‑IV at 1h30 and 6h compared to agr-I (S4 Table, sheets 1 and 2), while agr‑III showed broader metabolic enrichment only at 12h (S4 Table, sheet 3), indicating its delayed agr-activation, linked to differential metabolism compared to the other agr-types. Agr‑I and agr‑IV, which exhibited the strongest agr-activation, showed fewer stationary-phase metabolic changes at 12h, consistent with a shift toward virulence‑focused genes expression (S4 Table, sheet 3). Together, DE analysis, ORA and GSEA reveal that agr‑type not only determines the timing and magnitude of core agr‑regulated virulence genes but also drives distinct, temporally resolved metabolic pathways during growth.
Discussion
The agr-system is one of the most studied quorum-sensing systems and the most important virulence regulator in S. aureus [4], yet its strong linkage to genetic background has hindered efforts to investigate agr-type-specific effects. Using congenic strains sharing the same genetic background and exhibiting similar growth dynamics, we demonstrate that different agr-types possess distinct agr-activation kinetics, resulting in clearly separable temporal transcriptomic profiles.
Early agr-activation in agr‑I and agr‑IV drove strong induction of PSMs, capsule biosynthesis, and several pore-forming toxins, whereas agr‑II activation was more gradual and agr‑III activated only in stationary growth-phase, resulting in markedly different virulence signatures in early, late and stationary growth-phase. Importantly, agr‑II and agr‑III did not exhibit an avirulent profile; instead, they expressed alternative virulence factors focusing on superantigen-like genes during early stationary phase. Furthermore, adhesion factors such as spa or fnbA, that are expected to be downregulated in agr-activated bacteria [14] showed a similar profile of downregulation in our experiments for agr-I and agr-IV. This alternative virulence profile of agr-III aligns with clinical observations that both agr‑III or agr-deficient strains frequently cause invasive infections [15,16], indicating reliance on virulence repertoires less dependent on directly agr-regulated toxins. Unexpectedly, the adhesion factor clfA was upregulated at 6h in agr-IV, despite sharing regulatory features with spa and fnbA, highlighting the complexity and partial uncoupling within the agr-regulatory network. The transient reduction of chp expression in agr‑IV despite rapid agr-activation is consistent with previous reports showing limited agr-control over chp and reflects its integration into a broader regulatory network involving SaeRS, SarA‑family regulators, and stress‑responsive pathways [17,18].
Agr‑IV displayed the strongest and fastest agr-activation, as previously observed in another genetic background [19]. This correlated with the broadest transcriptional response, including early, high-level capsule expression and differential modulation of metabolic pathways. These findings suggest that the agr-type not only affects virulence factors but also impacts the metabolic strategies used to adapt to the same environment. The arginine deiminase (ADI) system as well as the kdp and ure operons were upregulated in agr-IV at 12h. All three are involved acid stress response [20]. In addition, the ADI system specifically has been shown to be modulated by available carbon sources [21,22], and our data show that agr-IV upregulated capsule production early, which is known to be strongly glucose dependent [23,24]. Further research is needed to confirm this hypothesis. Our experiments have been performed only in the Newman genetic background in triplicates. Similar studies with different genetic backgrounds to unravel co-evolutionary adaptation of genetic background and the agr-type as well as increased replicates of transcriptomics analysis can further strengthen the robustness of our observations. However, our study suggests that such agr-type‑specific transcriptional architectures may affect the ecological success of specific agr-lineages.
Previous studies found correlations between agr-type or the CC/ST with specific infections [7,8], notably agr-III and the CC30 (agr-III) with the tst carriage and the toxic shock syndrome [19,25]. However, some recent studies show that no specific associations were found between the agr-type and pathogenicity [16] or specific toxin carriage [26], underlying the complexity of S. aureus genetic impact on its virulence potential.
We demonstrated that the agr-type influenced the transcriptomic profile in congenic strains and under controlled growth conditions. However, the underlying molecular mechanisms, whether rooted in promoter architecture, AgrC signaling, or peptide-receptor coevolution, remain to be elucidated.
Supporting information
S1 Fig. Differential agr-activation kinetics of the four agr-types.
A and B. Relative expression of agrA (A) and RNAIII (B) to gyrB using RNAseq counts (number of reads of agrA or RNAIII / number of reads of gyrB) for each strain except Newman agr, as no reads could be detected for neither agrA nor RNAIII (same data as in Fig.1). Error bars represent the standard error of the mean (SEM). Two-way ANOVA tests on log10 transformed relative expressions and Turkey’s multiple comparisons tests were performed; adjusted p-value < 0.0001 - ****, < 0.001 - ***, < 0.01 - **, < 0.05 - *; ns – not significant. Three biological replicates were performed.
https://doi.org/10.1371/journal.pone.0350108.s001
(TIF)
S2 Fig. Cluster identification from PCA analysis.
Dot-plots of the principal component analysis (PCA), using PC1 axis, depending on the agr-type, using the entire transcriptome beside the agr-locus, with all time points in A, at time point 1h30 in B, 6h in C and 12h in D. Standard error of the mean (SEM) for each time point in A, and per strain in B, C and D, are represented in black.
https://doi.org/10.1371/journal.pone.0350108.s002
(TIF)
S3 Fig. No expression differences in the absence of agr-activation.
Volcano plots of pairwise comparisons of the four agr-types to agr at 1h30. Comparisons to agr-I in A, to agr-II in B, to agr-III in C and to agr-IV in D. Dashed lines indicate fold change threshold (fold change of |2| (LFC of 1)) and adjusted p-value threshold (0.05). NS: not significant, FC: fold change.
https://doi.org/10.1371/journal.pone.0350108.s003
(TIF)
S4 Fig. Genes differentially expressed in pairwise agr-type comparisons.
Heatmaps of regularized log counts (normalized raw counts and input for DESeq2 analysis) of the genes differentially expressed for at least one pairwise comparison between the four agr-types (S2 and S3 Tables), with a fold change of |4| (LFC of 2)). 94 genes out of the 411 at 6h (A), and 43 out of the 149 genes at 12h (B) were included (no genes with a fold change of |4| (LFC of 2)) were differentially expressed at 1h30 out of the 30 genes). Biological replicates for each strain are shown. Biological functions from KEGG database and detected in over‑representation analysis (ORA, •) or described in the literature to belong to these pathways (∘) are indicated next to the gene names. Gene names are annotated according to Aureowiki Newman gene annotation; if only pangenome annotation was available “p:” prefix was added to the gene’s name, for genes without gene’s name the four digits following the “NWMN” locus-tag were kept as gene ID.
https://doi.org/10.1371/journal.pone.0350108.s004
(TIF)
S1 Table. All genes significantly differentially expressed for at least one pairwise comparison at 1h30.
Blue: genes with a fold change (FC) <−2; red: genes with a FC > 2; green: adjusted p-value (p-adj) <0.05. Gene names are annotated according to Aureowiki Newman gene annotation; if only pangenome annotation was available “p:” prefix was added to the gene’s name, for genes without gene’s name the four digits following the “NWMN_” tag were kept as gene ID. Negative Binomial GLM with Wald test for pairwise comparisons were performed with Benjamini–Hochberg correction.
https://doi.org/10.1371/journal.pone.0350108.s005
(XLSX)
S2 Table. All genes significantly differentially expressed for at least one pairwise comparison at 6h.
Blue: genes with a fold change (FC) <−2; red: genes with a FC > 2; green: adjusted p-value (p-adj) <0.05. Gene names are annotated according to Aureowiki Newman gene annotation; if only pangenome annotation was available “p:” prefix was added to the gene’s name, if only the symbol annotation was available “s:” prefix was added to the gene’s name, for genes without gene’s name the four digits following the “NWMN_” tag were kept as gene ID. Negative Binomial GLM with Wald test for pairwise comparisons were performed with Benjamini–Hochberg correction.
https://doi.org/10.1371/journal.pone.0350108.s006
(XLSX)
S3 Table. All genes significantly differentially expressed for at least one pairwise comparison at 12h.
Blue: genes with a fold change (FC) <−2; red: genes with a FC > 2; green: adjusted p-value (p-adj) <0.05. Gene names are annotated according to Aureowiki Newman gene annotation; if only pangenome annotation was available “p:” prefix was added to the gene’s name, if only the symbol annotation was available “s:” prefix was added to the gene’s name, for genes without gene’s name the four digits following the “NWMN_” tag were kept as gene ID. Negative Binomial GLM with Wald test for pairwise comparisons were performed with Benjamini–Hochberg correction.
https://doi.org/10.1371/journal.pone.0350108.s007
(XLSX)
S4 Table. GSEA results for all agr-type pairwises comparisons.
Ranking metric prioritizes the Wald statistic (res$stat) of DEseq2, pvalueCutoff of 1 was used (no gene selection from DE), Benjamini–Hochberg FDR correction applied, and pathways with adjusted p-value <0.05 retained. minGSSize of 10 and maxGSSize of 500 were used as default parameters. Positively- or negatively-enriched refers to the first strain mentioned in the comparison. Enriched terms are reported with normalized enrichment scores (NES), p-values, and FDR q-values.
https://doi.org/10.1371/journal.pone.0350108.s008
(XLSX)
Acknowledgments
We thank Bo Shopsin for providing the agr-congenic Newman strains. RNA sequencing was performed at the Functional Genomics Center Zurich (FGCZ) of University of Zurich and ETH Zurich. We thank Alejandro Gómez Mejia, Federica Andreoni and Andrea Tarnutzer for their help on scientific discussions and manuscript revision.
References
- 1. Piewngam P, Otto M. Staphylococcus aureus colonisation and strategies for decolonisation. Lancet Microbe. 2024;5(6):e606–18. pmid:38518792
- 2. Howden BP, Giulieri SG, Wong Fok Lung T, Baines SL, Sharkey LK, Lee JYH, et al. Staphylococcus aureus host interactions and adaptation. Nat Rev Microbiol. 2023;21(6):380–95. pmid:36707725
- 3. Cheung GYC, Bae JS, Otto M. Pathogenicity and virulence of Staphylococcus aureus. Virulence. 2021;12(1):547–69.
- 4. Jenul C, Horswill AR. Regulation of Staphylococcus aureus Virulence. Microbiol Spectr. 2019;7(2). pmid:30953424
- 5. Novick RP, Geisinger E. Quorum Sensing in Staphylococci. Annu Rev Genet. 2008;42(1):541–64.
- 6. Holtfreter S, Grumann D, Balau V, Barwich A, Kolata J, Goehler A, et al. Molecular Epidemiology of Staphylococcus aureus in the General Population in Northeast Germany: Results of the Study of Health in Pomerania (SHIP-TREND-0). J Clin Microbiol. 2016;54(11):2774–85. pmid:27605711
- 7. Cafiso V, Bertuccio T, Santagati M, Demelio V, Spina D, Nicoletti G, et al. agr-Genotyping and transcriptional analysis of biofilm-producing Staphylococcus aureus. FEMS Immunol Med Microbiol. 2007;51(1):220–7. pmid:17854479
- 8. Wächter H, Yörük E, Becker K, Görlich D, Kahl BC. Correlations of Host and Bacterial Characteristics with Clinical Parameters and Survival in Staphylococcus aureus Bacteremia. J Clin Med. 2021;10(7):1371. pmid:33800644
- 9. Jarraud S, Mougel C, Thioulouse J, Lina G, Meugnier H, Forey F, et al. Relationships between Staphylococcus aureus genetic background, virulence factors, agr groups (alleles), and human disease. Infect Immun. 2002;70(2):631–41. pmid:11796592
- 10. Tan L, Huang Y, Shang W, Yang Y, Peng H, Hu Z, et al. Accessory Gene Regulator (agr) Allelic Variants in Cognate Staphylococcus aureus Strain Display Similar Phenotypes. Front Microbiol. 2022;13:700894. pmid:35295312
- 11. Geisinger E, Chen J, Novick RP. Allele-dependent differences in quorum-sensing dynamics result in variant expression of virulence genes in Staphylococcus aureus. J Bacteriol. 2012;194(11):2854–64. pmid:22467783
- 12. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. pmid:25516281
- 13. Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7. pmid:22455463
- 14. Williams P, Hill P, Bonev B, Chan WC. Quorum-sensing, intra- and inter-species competition in the staphylococci. Microbiology (Reading). 2023;169(8):001381. pmid:37578829
- 15. Pivard M, Bastien S, Macavei I, Mouton N, Rasigade J-P, Couzon F, et al. Targeted proteomics links virulence factor expression with clinical severity in staphylococcal pneumonia. Front Cell Infect Microbiol. 2023;13:1162617. pmid:37077532
- 16. Bastien S, Meyers S, Salgado-Pabón W, Giulieri SG, Rasigade J-P, Liesenborghs L, et al. All Staphylococcus aureus bacteraemia-inducing strains can cause infective endocarditis: Results of GWAS and experimental animal studies. J Infect. 2023;86(2):123–33. pmid:36603774
- 17. Rooijakkers SHM, Ruyken M, van Roon J, van Kessel KPM, van Strijp JAG, van Wamel WJB. Early expression of SCIN and CHIPS drives instant immune evasion by Staphylococcus aureus. Cell Microbiol. 2006;8(8):1282–93. pmid:16882032
- 18. Liu Q, Yeo WS, Bae T. The SaeRS Two-Component System of Staphylococcus aureus. Genes. 2016 Oct;7(10):1–20.
- 19. Jarraud S, Lyon GJ, Figueiredo AM, Lina G, Vandenesch F, Etienne J, et al. Exfoliatin-producing strains define a fourth agr specificity group in Staphylococcus aureus. J Bacteriol. 2000;182(22):6517–22. pmid:11053400
- 20. Zhou C, Fey PD. The acid response network of Staphylococcus aureus. Curr Opin Microbiol. 2020;55:67–73. pmid:32361405
- 21. Reslane I, Watson GF, Handke LD, Fey PD. Regulatory dynamics of arginine metabolism in Staphylococcus aureus. Biochem Soc Trans. 2024;52(6):2513–23. pmid:39656074
- 22. Reslane I, Halsey CR, Stastny A, Cabrera BJ, Ahn J, Shinde D, et al. Catabolic ornithine carbamoyltransferase activity facilitates growth of Staphylococcus aureus in defined medium lacking glucose and arginine. mBio. 2022;13(3):e0039522.
- 23. Visansirikul S, Kolodziej SA, Demchenko AV. Staphylococcus aureus capsular polysaccharides: a structural and synthetic perspective. Org Biomol Chem. 2020;18(5):783–98. pmid:31922180
- 24. O’Riordan K, Lee JC. Staphylococcus aureus capsular polysaccharides. Clin Microbiol Rev. 2004;17(1):218–34. pmid:14726462
- 25. Contou D, Colin G, Travert B, Jochmans S, Conrad M, Lascarrou J-B, et al. Menstrual Toxic Shock Syndrome: A French Nationwide Multicenter Retrospective Study. Clin Infect Dis. 2022;74(2):246–53. pmid:33906228
- 26. Wang M, Zheng Y, Mediavilla JR, Chen L, Kreiswirth BN, Song Y, et al. Hospital dissemination of tst-1-positive clonal complex 5 (CC5) methicillin-resistant Staphylococcus aureus. Front Cell Infect Microbiol. 2017;7:101.