Glucose-induced active lysogeny of prophage ΦSa3XN promotes Staphylococcus aureus virulence

Abstract

Background

Active lysogeny is a newly characterized mechanism that the dynamic integration and excision of prophages serve as molecular switches to coordinately regulate bacterial gene expression without generating progeny virions. The Sa3int family phages, the most prevalent prophages in Staphylococcus aureus, specifically integrate into the β-toxin-coding gene hlb. While infection conditions favor the loss of Sa3int phages and the emergence of Hlb-producing variants, highlighting their potential for active lysogeny, the environmental cues and underlying mechanisms controlling the peculiar life cycle of Sa3int phages remain largely unexplored.

Methods

In this study, we identified a Sa3int phage, designated ΦSa3XN, from the methicillin-resistant S. aureus strain XN108. The active lysogeny feature of ΦSa3XN was analyzed by combinational PCR, plaque assay, transmission electron microscopy, and DNase protection assay. Additionally, glucose-induced active lysogeny of ΦSa3XN and its impact on S. aureus virulence were evaluated via reporter assay, electrophoretic mobility shift assay, hemolytic assay, and mouse infection models.

Results

ΦSa3XN acts as a genuine molecular switch, capable of excision without producing progeny phages. Glucose serves as an environmental cue that triggers ΦSa3XN excision and reinstates hlb expression, wherein the catabolite control protein A (CcpA) directly binds to the promoter region of cI and suppresses the expression of CI repressor, thus switching the phage life cycle. Moreover, glucose-induced active lysogeny of ΦSa3XN significantly enhances bacterial hemolytic activity, exacerbating skin inflammation and subcutaneous abscess formation in hyperglycemic mice.

Conclusion

This study illustrates a novel example of active lysogeny for Sa3int phages and elucidates a glucose-responsive CcpA pathway that regulates ΦSa3XN excision to augment S. aureus virulence, advancing our understanding of the sophisticated interactions between S. aureus and phages.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12985-025-02993-3.

Background

Prophages represent one of the most prevalent and genetically diverse entities within bacterial genomes, typically sustaining dual lifestyles: either stably integrating into and co-propagating with the bacterial chromosome or initiating lytic replication to produce progeny phages under stress conditions. As crucial vehicles of horizontal gene transfer, prophages possess diverse functional elements such as virulence determinants, facilitating symbiotic interactions and promoting the survival, adaptation, and evolution of bacterial hosts in a phenomenon known as lysogenic conversion, which has been extensively investigated [1, 2]. Notably, a mechanism of mutually beneficial interactions termed active lysogeny has newly been summarized, in which prophages site-specifically insert into pivotal but conditional genes and precisely excise without inducing bacterial lysis and progeny production [3]. In these scenarios, the controlled integration and excision of prophages act as molecular regulatory switches to timely modulate gene expression in response to environmental cues [3]. To date, active lysogeny has been primarily documented in a limited number of bacterial species, including Listeria monocytogenes [4, 5], Streptococcus pyogenes [6], Bacillus subtilis [7], Escherichia coli [8], Legionella pneumophila [9], Shewanella oneidensis [10], and Staphylococcus aureus [3].

S. aureus is a notorious opportunistic pathogen causing millions of infections globally each year, manifesting as localized skin and soft tissue infections (SSTIs) to life-threatening conditions [11, 12]. The majority of S. aureus strains are lysogens, containing between one and four prophages that contribute up to 20% of the bacterial genome [13, 14]. Among these, the Sa3int family phages are the most prevalent, found in up to 96% of human nasal S. aureus isolates [15]. Sa3int phages specifically integrate into and interrupt bacterial hlb gene, which encodes β-hemolysin, while also carrying immune evasion genes that encode human-specific virulence factors such as SEA, SAK, CHIPS, SCIN, and TarP [16, 17]. Although the disruption of hlb coupled with the acquisition of additional virulence factors appears counterintuitive, emerging evidence suggests that the excision of Sa3int phages occurs during in vivo S. aureus infections, permitting the expression of both Hlb and the phage-encoded virulence factors at a population scale [15]. Consistently, a recent study revealed that prophage ΦSa3mw, a Sa3int phage present in the methicillin-resistant S. aureus strain MW2, acts as a novel regulatory switch mediating conditional hlb expression under in vitro conditions that mimic bacterial pathogenesis [18]. Nevertheless, the environmental cues and underlying molecular mechanisms controlling the peculiar life cycle of Sa3int phages remain largely unexplored.

In this study, we identified a Sa3int family phage, designated ΦSa3XN, within the methicillin-resistant S. aureus strain XN108. We demonstrated that ΦSa3XN is capable of excising from the XN108 chromosome but is unable to produce infectious phage particles, indicating the feature of active lysogeny. Furthermore, we found that glucose induces ΦSa3XN excision through the action of the catabolite control protein A (CcpA), which directly binds to the promoter region and represses the expression of cI, the gene encoding the master phage repressor controlling prophage activation and the lysis-lysogeny conversion. Additionally, the glucose-induced active lysogeny of ΦSa3XN significantly enhanced bacterial hemolytic activity and promoted S. aureus virulence in hyperglycemic mice with SSTIs. Collectively, this study presents a novel example of active lysogeny for Sa3int phages and elucidates a CcpA-mediated molecular mechanism that regulates the lysis-lysogeny switch of ΦSa3XN, advancing our understanding of the complex symbiotic interactions and coevolutionary dynamics between S. aureus and its prophages.

Methods

Bacterial strains and culture conditions

The bacterial strains and plasmids used in this study are detailed in Table S1. Unless specified otherwise, S. aureus strains were cultivated in tryptic soy broth (TSB; Oxoid, UK) at 37 °C with agitation at 200 rpm or grown on TSA (TSB supplemented with 1.5% agar). For the maintenance of plasmid pGFP and its derivatives (Table S1), cultures were supplemented with 100 µg/mL of ampicillin for E. coli and 10 µg/mL of chloramphenicol for S. aureus.

Construction of mutant Δhlb^C

The deletion of the 3’-terminal region of hlb, caused by ΦSa3XN insertion, was constructed in the S. aureus strain XN108. Briefly, the upstream and downstream regions immediately flanking the hlb 3’-terminal were amplified from XN108 genomic DNA and fused through overlapping PCR using primers listed in Table S2. The fused fragment was ligated into the BamH I and Sac I-digested vector pBT2 using Gibson assembly (NEB, USA). The resultant plasmid pBT2Δhlb^C was initially transformed into the S. aureus strain RN4220 for modification and then electroporated into XN108. The mutant Δhlb^C was screened via homologous recombination, detected by PCR, and verified through sequencing as previously described [19]. Note that the native stop codon of hlb is retained in Δhlb^C to ensure the correct translational termination of hlb 5’-terminal after ΦSa3XN excision.

Quantitation of ΦSa3XN excision frequency

XN108 and derivatives were grown under indicated conditions, including treatment with 1 µg/mL mitomycin C (MMC) or 2% glucose, and overnight genomic DNA was isolated for quantitative PCR (qPCR) analysis. The excised and integrated copies of ΦSa3XN were amplified using the primer pair attB-qF/qR targeting the attB site and attR-qF/qR targeting the attR site, respectively (Table S2). The excision rate of ΦSa3XN was quantified via qPCR and calculated using the 2^−ΔΔCT method as described previously [18]. The housekeeping gene gyrA was utilized for normalization.

DNase protection assay

The DNase protection assay was performed as previously described with minor modifications [20]. Briefly, overnight cultures of XN108 or ΔΦSa3XN were diluted 1:100 into fresh TSB and incubated to early log-phase (2.5 h) at 37 °C. Then, MMC was added at a concentration of 1 µg/mL to induce potential prophage activation until the cultures became clear. The culture supernatants were centrifuged at 10,000×g for 5 min, followed by filtration through a 0.45-µm filter. The filtrate samples were treated with DNase I (5 µg/mL) at 37 °C for 30 min, and the enzyme was inactivated by incubation at 65 °C for 10 min. The presence of intact phage DNA was detected via PCR amplification using prophage-specific primers (Table S2). Supernatants without DNase treatment were simultaneously amplified as PCR-positive controls.

Plaque assay

Log-phase cultures (200 µL) of indicator strains (RN4220, XN108, and ΔΦSa3XN) were combined with 5 mL of phage top agar (TSB supplemented with 0.75% agar) and subsequently poured onto a TSA plate, allowing them to air dry at room temperature to form an even layer. Filtered supernatants from XN108 and ΔΦSa3XN, with or without induction by 1 µg/mL MMC, were prepared as described above. These supernatants were 10-fold serially diluted, and 5 µL aliquots were spotted onto the lawns of the indicator strains. The plates were incubated at 37 °C and checked for plaque formation following overnight incubation.

Transmission electron microscopy (TEM)

Transmission electron microscopy was performed as described previously [20]. Briefly, filtered culture supernatants of XN108 and ΔΦSa3XN that induced with 1 µg/mL MMC were placed on a carbon-coated copper grid, negatively stained with 2% phosphotungstic acid (pH 4.5), and visualized using a TECNAI 10 electron microscope (Philips, The Netherlands).

Assessment of hemolytic activity

The hemolytic activity of culture supernatants was assessed as previously described with minor modifications [21]. Briefly, stationary-phase cultures of S. aureus XN108 and derivatives under indicated growth conditions were centrifuged at 17,000×g for 5 min, and 400 µL of the supernatant was collected into a clean microcentrifuge tube. An equal volume of 2% defibrinated sheep blood in PBS was then mixed and incubated statically at 37 °C for 1 h, followed by overnight incubation at 4 °C. The samples were pelleted at 500×g for 5 min, and 200 µL of the supernatant was transferred into a 96-well plate. Absorbance was monitored at 540 nm using a microplate reader (Bio-Tek, China) and normalized against bacterial growth by measuring optical density at 600 nm (OD₆₀₀).

RT-qPCR analysis

Total cellular RNAs from untreated and 2% glucose-induced XN108 cultures were extracted using the RNAprep Pure Cell/Bacteria Kit (TIANGEN, China) and reverse-transcribed via the RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, USA) following the recommended methods. The cDNA was amplified with TB Green™ Premix Ex Taq™ II (TaKaRa, China) and analyzed by the CFX96 Manager (Bio-Rad, USA). The primers used for RT-qPCR are listed in Table S2. Gene expression levels were determined using the 2^−ΔΔCT method with gyrA as endogenous control.

Electrophoretic mobility shift assay (EMSA)

The His-tagged recombinant protein CcpA was purified and stored in our previous work [22]. The biotin-labeled DNA probe encompassing the 160-bp region upstream of cI was amplified from the XN108 genome using primers listed in Table S2, and incubated with increasing concentrations of His-tagged CcpA. EMSA was conducted as previously described [22], in which unlabeled probes were added in a 200-fold excess as specific competitors. DNA probes, generated by annealing two complementary 50-bp primers (Table S2) that cover the upstream sequences of rocD and proC, served as positive and negative controls, respectively, according to a previous study [23].

Detection of cI promoter activity

The fragment containing the upstream 300-bp region and the start codon of cI was amplified from XN108 with primers listed in Table S2. The PCR product was inserted into the pGFP vector, which had been pre-digested with Kpn I and Xba I, using the Gibson assembly master mix (NEB, USA), creating pGFPcI that contains the cI promoter driving the expression of green fluorescent protein. The plasmid pGFPcI was initially transformed into the S. aureus strain RN4220 for modification and subsequently electroporated into XN108, generating the reporter strain XN108/pGFPcI. The activity of the cI promoter was monitored by measuring fluorescence intensity and normalized by bacterial growth of OD₆₀₀.

SDS-PAGE and Western blot analysis

Overnight cultures of XN108 and derivatives that grown under indicated conditions were subjected to exoprotein isolation as previously described [24]. The protein samples were solubilized in SDS-PAGE buffer (Beyotime, China), boiled for 10 min, and subsequently separated on a 10% SDS-PAGE gel. For Western blot analysis, the proteins were transferred onto a PVDF membrane (Beyotime, China) and blocked with 5% skim milk in high-salt Tris-buffered saline (HS-TBS; 20 mM Tris, 500 mM NaCl, pH7.5) at room temperature for 1 h. The membrane was then incubated overnight at 4 °C with a rabbit polyclonal anti-Hlb antibody (1:10,000; Sangon Biotech, China). After washing five times with PBS-T (PBS containing 0.05% Tween-20), the membrane was incubated with a goat anti-rabbit IgG-horseradish peroxidase conjugate (1:10,000; Solarbio, China) for 1 h at 37 °C. The blots were visualized using the Pierce ECL Western Blotting Substrate (Thermo Scientific, USA) and photographed using a ChemiDoc XRS Imaging System (Bio-Rad, USA).

Mouse experiments

Eight-week-old male BALB/c mice were purchased from HUNAN SJA laboratory animal Co., LTD (Hunan, China) and housed in two-way housing on a 12-h light/dark cycle. Hyperglycemic mice were induced as previously described with minor modifications [25–27]. Briefly, BALB/c male mice were treated by intraperitoneal injection of streptozocin (STZ; Sigma-Aldrich, USA) at a dosage of 180 mg/kg. On the tenth day following STZ injection, the 6-h fasting blood glucose (FBG) levels were analyzed from the tail vein using a glucometer (Yuwell, China). Mice exhibiting FBG levels exceeding 200 mg/dL (11.1 mmol/L) in two consecutive measurements were classified as hyperglycemic and selected for further experimentation.

Log-phase cultures of S. aureus strain XN108 and derivatives that grown in TSB were collected, enumerated by viable bacterial counts, and diluted to a concentration of 5 × 10⁸ CFU/mL in PBS. Hyperglycemic mice were anesthetized with 2,2,2-tribromoethanol (Sigma-Aldrich, USA), and their dorsal regions were shaved. Subsequently, 100 µL of XN108 and derivatives, equal to 5 × 10⁷ CFU, were subcutaneously injected into the left and right flanks of each mouse’s back, respectively. The mice were monitored daily for changes in body weight and abscess formation, with photographic documentation conducted over 14 days. The size of the skin abscesses was quantified using ImageJ software (ImageJ, USA). On the sixth day post-infection, skin lesions were excised and fixed in 4% paraformaldehyde. Hematoxylin-eosin (HE) staining and histopathological analyses were conducted by Servicebio Technology Co., Ltd. (Servicebio, China).

Statistical analysis

Statistical analysis was performed by GraphPad Prism v8.0. Student’s t-test was used for the comparison of two independent datasets. One-way or two-way analysis of variance (ANOVA) was utilized for multiple comparisons. A P-value less than 0.05 was considered significant.

Results

Features of prophages in S. aureus strain XN108

The methicillin-resistant S. aureus strain XN108 belonging to sequence type 239 was originally isolated from wound specimens of a 34-year-old steam-burned patient, with complete genome sequenced previously [28, 29]. To profile the prophage elements present in XN108, the whole genomic sequence of XN108 was surveyed using the PHASTER web tool [30], resulting in the prediction of six putative prophages (designated as pp1 to pp6) (Fig. 1A). These prophages exhibit considerable variation in size, ranging from 9.6 kb to 128.9 kb, and differ in GC content compared to the XN108 genome (32.8%), indicative of the nature of mobile genetic elements (Fig. 1B). Further analysis of the predicted attachment (att) sites for each prophage revealed that all prophages integrate into intergenic regions, except for pp5 which inserts into the intragenic region of hlb encoding β-hemolysin (alternative β-toxin) (Fig. 1B and C). Based on the polymorphism of phage-encoded integrase, S. aureus prophages are categorized into 10 distinct groups, with those integrating into hlb referred to as Sa3int phages [13]. Consistently, sequence alignment of the integrase encoded by pp5 demonstrated 98.0% and 97.7% amino acid identity with the Sa3int representatives Φ13 and ΦSa3mw, respectively, but only 24.1% and 25.0% similarity to Sa5int Φ11 and Sa2int Φ12 (Fig. S1). Additionally, pp5 possesses several typical characteristics belonging to Sa3int phages, such as encoding a putative CI/Mor regulatory switch and several known immune evasion genes including scn, sak, and sea, which encode staphylococcal complement inhibitor (SCIN), staphylokinase (SAK), and staphylococcal enterotoxin A (SEA), respectively (Fig. 1D). Thus, we infer that pp5 belongs to the Sa3int family and renamed it ΦSa3XN.

Fig. 1.

Prediction and characteristics of prophages in S. aureus strain XN108. (A) Circular representation of the S. aureus XN108 genome with six predicted prophages indicated by blue rectangles in the outermost concentric ring. (B) Features of the six prophages in the XN108 genome. These prophages differ in size and GC content, in which only prophage pp5 (alternative ΦSa3XN) inserts into the intragenic region and disrupts hlb encoding β-hemolysin. (C) Schematic of ΦSa3XN integration that splits hlb into 5’ and 3’-terminal. Arrows indicate the location and orientation of genes that are not drawn to scale. (D) Genetic organization of the ΦSa3XN genome. Grey arrows represent genes of unknown function

ΦSa3XN is capable of excision but unable to produce phage virions

Given that ΦSa3XN integrates site-specifically into hlb, disrupting the expression of β-toxin, we initially determined whether ΦSa3XN could excise from the XN108 genome. To differentiate between the integrated and excised forms of ΦSa3XN, two sets of primers were employed: one flanking the attB site, which is present only in the XN108 chromosome after ΦSa3XN excision, and the other flanking the attP site, present exclusively in the circularized form of ΦSa3XN (Fig. 2A). Combinatorial PCR analysis demonstrated that ΦSa3XN is capable of excision and circularization, resulting in the formation of both intact attB and attP sites, as verified by DNA sequencing (Fig. 2B and C). To quantify the rate of excision, qPCR was conducted by comparing the number of attB sites to attR sites (present only in the integrated ΦSa3XN-bacterial DNA junction), and revealed a spontaneous excision frequency of ΦSa3XN from the overnight-cultured XN108 genome at 4.18 × 10⁻⁴ (equivalent to 1 in 2,392 cells), which significantly increased to 4.62 × 10⁻³ (equivalent to 1 in 216 cells) following treatment with the prophage activation-inducing agent MMC (Fig. 2D). In alignment with capable of excision, we screened a XN108-derived mutant via plating the MMC-induced XN108 cultures on Columbia blood agar, which lacks the entire ΦSa3XN region (designated ΔΦSa3XN) and demonstrates significantly increased hemolytic activity (Fig. 2E and F; Fig. S2A). SDS-PAGE analysis of overnight culture supernatants showed that a band with a molecular size between 35 and 40 kDa was present in ΔΦSa3XN but not in XN108, which was further analyzed by mass spectrometry and Western blot analysis, confirming it as Hlb with a theoretical size of 37.2 kDa (Fig. 2G and H; Fig. S2B). These findings suggest that ΦSa3XN can excise from the XN108 genome, thereby restoring Hlb expression.

Fig. 2.

ΦSa3XN can excise but is unable to produce phage particles. (A) Schematic illustrating the excision and circularization of ΦSa3XN. The primers used for amplification of attB, attP, attL, and attR are shown. (B) The excised and integrated forms of ΦSa3XN were analyzed for XN108 genomic DNA by combinatorial PCR. The primer pair int-F/R was applied for amplification of the ΦSa3XN-encoded integrase gene. (C) Sequencing of PCR products across attB and attP sites following ΦSa3XN excision as obtained in (B). The 14-bp att site recognized by ΦSa3XN is labeled with the red box. (D) The excision frequency of ΦSa3XN from overnight cultures of XN108 that uninduced or induced with 1 µg/mL MMC was quantified via qPCR. Data are expressed as the mean ± standard deviation (SD) from three independent replicates and analyzed by Student’s t-test, ***P < 0.001. (E) Hemolytic activity of XN108 and derivatives. Data are the mean ± SD from three independent replicates and determined by one-way ANOVA with Tukey’s multiple comparisons test, ****P < 0.0001. (F) Representative hemolysis of XN108 and derivatives cultured on Columbia blood agar. ΔΦSa3XN forms a clear hemolytic ring. (G) SDS-PAGE analysis of whole exoproteins from overnight cultures of XN108 and ΔΦSa3XN. The red arrow indicates the band between 35 and 40 kDa that was present in ΔΦSa3XN but not in XN108. (H) Western blot analysis for Hlb expression in XN108 and ΔΦSa3XN. Representative gel is shown from three independent experiments. LC: loading control. (I) TEM analysis for overnight culture supernatants of XN108 and ΔΦSa3XN that treated with 1 µg/mL MMC. No obvious difference for phage particles was detected between XN108 and ΔΦSa3XN

A criterion for active lysogeny is that prophage excision does not result in the production of progeny virions [3]. To investigate whether ΦSa3XN could produce infectious phage particles, we first conducted the spot assay to assess whether supernatants from XN108 and ΔΦSa3XN could form plaques on lawns of indicator S. aureus strains. If ΦSa3XN could release virions, the supernatants from XN108 should form plaques on mutant ΔΦSa3XN, but not on XN108, due to the mechanism of superinfection exclusion. S. aureus strain RN4220 is a prophage-free and restriction-deficiency strain known to be susceptible to phages, both the supernatants of XN108 and ΔΦSa3XN produce clear plaques on RN4220 lawns with a comparable efficiency of plating (Fig. S2C). Conversely, when these supernatants were plated on lawns of XN108 and ΔΦSa3XN, no plaques were observed, indicating that ΦSa3XN is incapable of producing phages (Fig. S2C). We also utilized TEM to visualize potential phage particles collected in MMC-treated supernatants. Although intact phages were detected, no obvious morphological differences were observed between the supernatants from XN108 and ΔΦSa3XN under tested conditions (Fig. 2I). To further elucidate the origin of phages, the DNase protection assay was performed, wherein DNA encapsulated within phage capsids is protected from DNase digestion and can be subsequently amplified via PCR. As controls, all six prophages were PCR-detected in MMC-induced XN108 supernatants before DNase treatment. However, following DNase digestion, only prophage pp1 exhibited DNase-resistant DNA (Fig. S2D). Similarly, all prophages, except for ΦSa3XN, were present in the MMC-induced supernatants of mutant ΔΦSa3XN prior to DNase treatment, but only pp1 was detected after DNase treatment (Fig. S2D). These results indicate that prophage pp1, but not ΦSa3XN, may contribute to the observed plaques and phages. Altogether, we infer that ΦSa3XN can excise from the bacterial chromosome spontaneously or induced by MMC, but is unable to generate infectious phage particles, thus functioning in a role of active lysogeny.

Glucose promotes ΦSa3XN excision and enhances the hemolytic activity of S. aureus

Although Sa3int phages are widely distributed in human-associated S. aureus strains and have been shown to increase excision under several clinical infectious conditions [15], the cues triggering Sa3int phage excision remain poorly understood. A previous study showed that besides MMC and H₂O₂, both known to activate prophage through DNA damage-dependent mechanism, biofilms also promote Sa3int phage ΦSa3mw excision with concomitant Hlb production, in which the bacteria were cultivated in a medium supplemented with 2% glucose, 2% NaCl, and human plasma [18]. Considering that glucose is a preferred carbon source crucial for S. aureus infections, we assessed whether glucose could function as an environmental signal regulating the ΦSa3XN life cycle. qPCR analysis of XN108 overnight cultures that untreated or supplemented with 2% glucose revealed that ΦSa3XN indeed responds to glucose and excises under glucose-treated conditions (Fig. 3A). Compared to uninduced control, the ΦSa3XN excision frequency after 2% glucose induction was significantly increased by approximately 15.6 times (Fig. 3A). In addition, the hemolytic activities of XN108 supernatants that induced with 2% glucose were also significantly enhanced (Fig. 3B).

Fig. 3.

Glucose-induced CcpA directly regulates ΦSa3XN excision. (A) The relative excision rates of ΦSa3XN from overnight cultures of XN108 and ΔccpA that untreated or supplemented with 2% glucose were determined by qPCR. The ΦSa3XN excision rate for untreated XN108 was set to 1. Data are the mean ± SD from three independent replicates and analyzed by two-way ANOVA with Sidak’s multiple comparisons test, **P < 0.01; ***P < 0.001; ns, not significant. (B) Hemolytic activities of XN108 and ΔccpA as treated in (A). The hemolytic activity of untreated XN108 was set to 1. Data are the mean ± SD from three independent replicates and determined by two-way ANOVA with Sidak’s multiple comparisons test, ***P < 0.001; ****P < 0.0001; ns, not significant. (C) RT-qPCR analysis of ccpA, cI, and mor expression in XN108 after 2% glucose treatment. Data are the mean ± SD from three independent replicates and analyzed by Student’s t-test, *P < 0.05. (D) RT-qPCR analysis of cI expression in XN108 and mutant ΔccpA. Data are the mean ± SD from three independent replicates and determined by Student’s t-test, **P < 0.01. (E) The frequency plot of CcpA-binding cre sites. Examples of known cre sites positioned in the promoter regions of corresponding genes are shown below. The predicted CcpA-binding site in cI promoter is highlighted in red characters. (F) EMSA analysis of specific binding of CcpA to cI promoter. A 200-fold excess of unlabeled probe (PcI) was included as a competitor. Probes Biotin-ProcD and Biotin-PproC served as positive and negative controls, respectively. Representative gel from three independent experiments is shown. (G) Assessment of cI promoter activity. S. aureus XN108 containing the GFP-based reporter vector was cultivated overnight that untreated or supplemented with 2% glucose, and fluorescence was measured and normalized to bacterial growth. Data are the mean ± SD from three independent replicates and analyzed by Student’s t-test, ****P < 0.0001

CcpA responds to glucose and directly modulates ΦSa3XN excision

Prophages integrate stably into the bacterial genome in a dormant state, mediated by the master repressor CI, which represses lytic genes while facilitating lysogenic gene expression. Generally, the lysis–lysogeny conversion involves intricate regulatory mechanisms wherein prophages not only sense the viability of the bacterial host through the DNA damage-induced SOS response, but also can employ bacterial or phage-encoded quorum-sensing systems to synchronize with the bacterial population, leading to lysis–lysogeny switch accordingly [31]. For Sa3int family phages, recent studies have characterized the genetic switch of phage Φ13, which constitutes cI, mor encoding the modulator of repression (Mor), and three high-affinity operator sites for CI binding [32, 33]. Sequence analysis indicated that ΦSa3XN encodes a putative CI/Mor regulatory switch, but no homologous CI binding sites was identified. Furthermore, the regulation of the Sa3int phage life cycle is limited illustrated at the molecular level, and bacteria-derived factors that control the dynamics, as well as the underlying mechanisms, remain largely unexplored.

In S. aureus, CcpA represents a global regulator of carbon catabolite repression, which binds to catabolite-responsive elements (cre) upon glucose addition to control the transcription of genes involved in central metabolism and virulence [34]. Considering that glucose promotes ΦSa3XN excision, we thus assessed whether CcpA participates in this process. RT-qPCR analysis showed that expression of both ccpA and mor were significantly increased, and cI was significantly decreased for XN108 cultures treated with 2% glucose compared to controls (Fig. 3C). Moreover, the expression of cI was significantly elevated in XN108 when ccpA was in-frame deleted (Fig. 3D), indicating the negative regulation of CI expression by CcpA. Consistently, the effect of glucose-induced ΦSa3XN excision and hemolytic activity was significantly attenuated in mutant ΔccpA compared to XN108 (Fig. 3A and B).

To further uncover the cues of CcpA-controlled ΦSa3XN activation, we first analyzed the promoter sequence of cI and identified a putative CcpA-binding cre site (5’-ATGAAAATTTCGTATA-3’) located 25-bp upstream of the start codon of cI (Fig. 3E). EMSA demonstrated that recombinant His-tagged protein CcpA specifically binds to biotin-labeled probes containing the predicted cre site in a dose-dependent manner; however, this binding was greatly blocked in the presence of a 200-fold excess of unlabeled probes (Fig. 3F). Additionally, the GFP-based reporter assay in which the expression of GFP is directly controlled by cI promoter consistently demonstrated the reduced expression of CI in XN108 after 2% glucose induction (Fig. 3G). Taken together, CcpA binds directly to the promoter of cI and represses CI expression, regulating lysis-lysogeny conversion of ΦSa3XN.

ΦSa3XN excision-restored Hlb augments S. aureus virulence in hyperglycemic mice

S. aureus is one of the most frequently isolated bacteria in SSTIs, particularly in patients with diabetes, where hyperglycemia exacerbates the severity of the infections [35]. The excessive glucose availability in diabetes greatly triggers metabolic adaptations for S. aureus following infection and alters expression profiles of several virulence factors, such as α-toxin, biofilm, aureolysin (Aur), Panton-Valentine leukocidin (PVL), toxic shock syndrome toxin-1 (TSST-1), and caseinolytic proteases (Clp) [36]. While Hlb has been demonstrated to promote S. aureus skin colonization by damaging keratinocytes [37], the pathogenic role of Hlb in diabetic SSTIs remains unexplored. To explore this, we first constructed an in-frame deletion of the 3’-terminal of hlb that interrupted by ΦSa3XN integration in XN108, wherein the stop codon of hlb was retained ensuring the correct translational termination of the 5’-terminal after ΦSa3XN excision, and the resultant mutant was designated Δhlb^C (Fig. S3A). Combinational PCR analysis confirmed that ΦSa3XN could still spontaneously excise and circularize from the Δhlb^C chromosome (Fig. S3B). While ΦSa3XN excision from Δhlb^C was significantly increased upon 2% glucose induction, the hemolytic activity of Δhlb^C was significantly attenuated, exhibiting a comparable level relative to both XN108 and Δhlb^C that without glucose treatment (Fig. 4A and B).

Fig. 4.

ΦSa3XN excision-restored Hlb promotes S. aureus virulence in hyperglycemic mice. (A) The relative excision rates of ΦSa3XN from overnight cultures of XN108 and Δhlb^C that untreated or supplemented with 2% glucose were determined by qPCR. The ΦSa3XN excision rate for untreated XN108 was set to 1. Data are the mean ± SD from three independent replicates and analyzed by two-way ANOVA with Sidak’s multiple comparisons test, **P < 0.01; ns, not significant. (B) Hemolytic activities of XN108 and Δhlb^C as treated in (A). The hemolytic activity of untreated XN108 was set to 1. Data are the mean ± SD from three independent replicates and determined by two-way ANOVA with Sidak’s multiple comparisons test, ****P < 0.0001; ns, not significant. (C) Schematic illustrating the animal experimental procedure. (D) The 6-h fasting blood glucose levels of mice before and after streptozotocin injection. Five mice successfully developed hyperglycemia with FBG more than 11.1 mmol/L (the dotted line). (E) Monitoring for abscess formation. Data are the mean ± SD (n = 5 mice) and analyzed by two-way ANOVA with Sidak’s multiple comparisons test, *P < 0.05. (F) Representative images of mouse skin damage. (G) Representative HE staining of mouse skin tissues collected on day 6 post-infection. The red triangles indicate the infiltration of inflammatory cells, black triangles show the vascular congestion, and blue triangles point to the loosed arrangement of connective tissues. (H) The severity of the skin lesions in histological sections of (G) was numerically classified with ImageJ software, and the relative inflammatory cell infiltration area (%) was calculated and indicated. Three sections with the bars of 200 μm were used. Statistical significance was determined by Student’s t-test, *P < 0.05

To further clarify whether ΦSa3XN excision and subsequent restoration of Hlb increase the virulence of S. aureus in vivo, we utilized a streptozotocin (STZ)-induced hyperglycemic mouse model to compare the severity of S. aureus SSTIs between XN108-derived strains that differ in Hlb expression. Each mouse was inoculated subcutaneously with XN108 and Δhlb^C on the left and right dorsal skin, respectively, and was monitored 14 days post-infection for abscess formation (Fig. 4C and D). The result showed that XN108-infected mice developed more severe infections with significantly increased lesion size compared to that of mutant Δhlb^C (Fig. 4E and F). Histopathological analysis also demonstrated that XN108-infected mouse displayed more pronounced subcutaneous skin damage than those infected with Δhlb^C, as evidenced by increased inflammatory cell infiltration, a looser arrangement of connective tissues, increased vascular congestion, and partial epidermal disruption (Fig. 4G and H). In contrast, no significant difference was observed between XN108 and Δhlb^C when the skin infection model was applied to normal (non-hyperglycemic) mice (Fig. S3C-G). Taken together, our results suggest that hyperglycemia-induced ΦSa3XN excision and subsequent Hlb expression promote the virulence of S. aureus.

Discussion

Active lysogeny represents a newly described mechanism of phage-bacteria interactions wherein prophages function as molecular regulatory switches without producing progeny phage virions [3]. Several forms of active lysogeny have been documented, predominantly in the species L. monocytogenes [4, 5], S. pyogenes [6], and B. subtilis [7]. Sa3int family phages are by far the most prevalent S. aureus integrating prophages that typically insert into hlb and concurrently carry several functional factors beneficial for S. aureus infections [13]. While stable maintenance of Sa3int phages within bacterial chromosomes under non-infectious conditions, growing compelling evidences showed that infections can greatly select for the loss of Sa3int phages and reconstitution of Hlb-positive variants of S. aureus with increased virulence [15, 37–39], suggesting the potential of active lysogeny for Sa3int phages. Previously, the prophage ΦSa3mw, a member of the Sa3int family phages, was first shown to act as a regulatory switch for the conditional expression of hlb [18]. Here, we present a novel instance of active lysogeny for Sa3int phages, in which ΦSa3XN can excise from the bacterial chromosome without forming infectious phage particles. Furthermore, we revealed a new mechanism that the S. aureus-encoded master regulator CcpA responds to glucose and promotes ΦSa3XN excision by directly suppressing the expression of CI repressor, leading to the induction of ΦSa3XN, elevated Hlb expression and augmented virulence for S. aureus SSTIs (Fig. 5).

Fig. 5.

Proposed model for glucose-induced ΦSa3XN excision and modulation of Hlb-mediated virulence. Under hyperglycemic conditions, the glucose-responsive regulator CcpA is activated and binds directly to the cre site within the cI promoter, leading to the repression of CI expression. Consequently, ΦSa3XN transitions from lysogenic to lytic life cycle, resulting in increased ΦSa3XN excision. This process restores the expression of hlb, leading to a substantial production of Hlb at the population level, which ultimately enhances the virulence of S. aureus in skin infections in hyperglycemic mice. Conversely, when ΦSa3XN remains stably integrated into hlb, only spontaneous ΦSa3XN excision and Hlb expression occurs, resulting in limited S. aureus virulence

Although widely distributed in S. aureus, the ability to form phage virions for Sa3int phages differs greatly in distinct S. aureus strains. A recent study showed that both two prototypic Sa3int phages, namely Φ13 and ΦN315 that derived from the S. aureus reference strain 8325 and the methicillin-resistant strain N315, respectively, can excise and enter lytic production to release infectious phage particles based on phage transfer experiments [40]. Replication of Φ13 and ΦN315 was significantly increased in S. aureus strain Newman and MW2 compared to strain 8325 and USA300, indicating the life cycle switches of Sa3int phages are additionally influenced by bacterial host factors [40]. In comparison, prophage ΦSa3mw carried in S. aureus strain MW2 was found to be able to excise, but did not release infectious virions upon induction with H₂O₂ or MMC [18]. However, this finding is somewhat perplexing, as MW2 harbors additional putative prophages that could potentially be activated to form phages; yet, no phage particles were detected in the culture supernatants, and the study regrettably did not include results from plaque assays or TEM analysis in the article [18]. In our study, we assessed the phage-producing capability of prophage ΦSa3XN using a combination of DNase protection assay, plaque assay, and TEM analysis. Although phage particles were detected in the culture supernatants, no obvious difference was observed between wild-type XN108 and ΦSa3XN-cured mutant ΔΦSa3XN, indicating the inability of ΦSa3XN to release phage virions. Note that ΦSa3XN exhibits very high average nucleotide identity to other Sa3int phages, such as 97.07%, 97.72%, and 95.17% to that of Φ13, ΦN315, and ΦSa3mw, respectively. Further comprehensive genomic analysis is needed, which may elucidate the factors contributing to the varied phage-forming capacity for the Sa3int family phages.

An intriguing area of research for Sa3int phages is the identification of potential triggers for lysis-lysogeny switches. Several studies have demonstrated that DNA-damaging agents, such as MMC, H₂O₂, and ciprofloxacin, which are well-known for phage-inducing properties, can promote the excision of Sa3int phages in an SOS response-dependent manner [18, 41]. Sub-lethal concentrations of commercial biocides containing H₂O₂ have also been found to stimulate the excision and transfer of prophage Φ13 from S. aureus human strains to livestock-associated methicillin-resistant S. aureus CC398 strains [42]. Additionally, biofilm and infectious situations are assumed to play crucial roles in modulating the life cycle of Sa3int phages [39]. For instance, ΦSa3mw excised from the MW2 chromosome with increased frequency under in vivo mouse skin colonization conditions [37], and β-toxin-producing variants are prevalent in S. aureus strains where the Sa3int prophage is differentially induced in endocarditis vegetations, ischemic livers, and kidney abscesses in rabbit models [38]. Nevertheless, the infection-related cues and bacterial factors that regulate the switching dynamics of Sa3int phages remain largely unexplored. Here, we discovered that glucose, a primary carbon source for both S. aureus and host cells, can induce the lysis-lysogeny conversion of ΦSa3XN through the master regulator CcpA, revealing a novel paradigm of life cycle regulation for Sa3int phages.

While frequently disrupted by Sa3int phage insertions, growing evidence has demonstrated that Hlb, as an important virulence factor, plays a pivotal role in S. aureus pathogenesis [43]. Currently, various functions have been attributed to Hlb, including the selective killing of human monocytes [44], inducing mouse lung injury through ectodomain shedding of syndecan-1 in epithelial cells [45], inhibition of endothelial interleukin-8 production and neutrophil transmigration [46], promoting mouse skin colonization [37], intensifying infective endocarditis vegetation formation in rabbit [47], modulating human aortic endothelial cell and platelet function [48], and causing mouse skin inflammation [49]. Here, we demonstrate that glucose-induced Hlb expression significantly promotes skin inflammation and subcutaneous abscess formation in hyperglycemic mice. Given that elevated blood glucose levels are a hallmark of advanced diabetes, which is highly susceptible to S. aureus infections and the potential regulatory switch of Sa3int phages, we speculate that glucose-mediated active lysogeny of Sa3int phages may play a critical role in diabetic S. aureus infections. A primary focus of our ongoing research is to further investigate the causal relationship between Sa3int phage carriage, Hlb expression, and disease severity in clinical diabetic specimens.

Conclusions

Our findings provide evidence that ΦSa3XN functions as a bona fide regulatory switch. Glucose as an environmental cue triggers ΦSa3XN excision and reinstates Hlb expression under the regulation of CcpA, leading to increased virulence of S. aureus in hyperglycemic mice SSTIs. This study illustrates a novel example of active lysogeny for Sa3int family phages, and uncovers a new mechanism by which CcpA responds to glucose and mediates ΦSa3XN excision to promote S. aureus pathogenesis, highlighting the dynamic and sophisticated interactions between S. aureus and phages.

Author contributions

GL, RJZ, and ML conceived and designed the experiments. WH, XT, and KTZ performed the experiments. GL, WH, and SGL analyzed the data. YZ, JZ, TYH, and CL contributed reagents/materials/analysis tools. GL and ML wrote and supervised the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China under Grant number 31900145 to Gang Li.

Data availability

All data generated or analyzed during this study are included in this published article [and its supplementary information files].

Declarations

Ethics approval and consent to participate

Animal experiments were approved by the Laboratory Animal Welfare and Ethics Committee of the Third Military Medical University and complied with the Institutional Animal Welfare and Ethical guidelines (AMUWEC2019421).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.