For this study, we collected 4,405 non-repetitive S. aureus isolates from various clinical sources during 2008–2020 from a general Hospital in Shanghai, China. Thereafter, we randomly selected 10 isolates of UTI-ST1, UTI-ST5, nUTI-ST1, UTI-ST7, and UTI-ST398 to perform the phenotypic experiments. All the strains were grown in TSB (Oxoid, United States) in the presence or absence of urea supplementation (2% w/v; Yeasen, Shanghai, China). A nutrient-deficient medium, containing 1 g of peptone (Diamond, Shanghai, China), 1 g of D(+)-glucose (Sangon Biotech, Shanghai, China), 2 g of potassium dihydrogen phosphate (BBI, Shanghai), and 5 g of sodium chloride (Diamond, Shanghai) in 1 L of sterilized water, with or without 2% (w/v) urea-supplementation (Yeason), was prepared as described previously (Duran Ramirez et al., 2022), to conduct the urease activity test. The nutrient-deficient medium without phenol red was used to conduct the growth test.

We selected 10 isolates of UTI-ST1 and UTI-ST5 for the growth curve test. The isolates were grown overnight in 3 mL TSB at 37°C and 220 rpm. The overnight cultures were then washed two times with phosphate-buffered saline (PBS) and diluted (1:10) in 3 mL fresh nutrient-deficient medium with or without urea supplementation (2% w/v). Thereafter, the diluted cultures were inoculated into sterile 96-well flat-bottom tissue culture plates (200 μL/well; Corning) and incubated at 37°C with shaking in Micro-ELISA Autoreader (Synergy 2, Bio-TeK, United States), and the OD was measured at 600 nm, every 30 min. The assay was performed in triplicate.

Semi-quantitative biofilm assay was performed as previously described (Vuong et al., 2003; He et al., 2019). Briefly, the overnight bacterial cultures were diluted with TSB, containing 0.5% glucose, to obtain a final OD of 0.05. Then, the diluted cultures were aliquoted to 96-well plates (200 μL/well) and incubated at 37°C for 24 h. The wells were washed with PBS after the gentle removal of the culture supernatants. Thereafter, the Bouin fixative was added to the wells to treat the biofilms for 1 h. The fixative was then gently aspirated and the wells were washed thrice with PBS, and then stained with 0.4% (w/v) crystal violet (Sangon). Biofilm formation was determined by measuring the OD at 570 nm using the Micro-ELISA autoreader.

The hemolysis test was conducted as described previously (He et al., 2018; Ridder et al., 2021). Hemolytic activities were determined by incubating the overnight culture-supernatants with human red blood cells (RBCs; 2% v/v in Dulbecco’s phosphate-buffered saline, DPBS) for 1 h at 37°C and measuring the OD at 540 nm using an enzyme-linked immunosorbent assay (ELISA) reader. The assay was performed in triplicate. MRSA strain USA300 was used as a reference, as it displayed relatively strong virulence. A standard curve was made by using serial dilutions (1,2, 1:4, 1:8, and 1:16) of the supernatants of the RBC lysates.

The adhesion and invasion test was performed as described previously (Wang et al., 2018). The human bladder carcinoma 5,637 cells were cultured in Roswell Park Memorial Institute (RPMI) medium, supplemented with 10% fetal bovine serum (FBS) and 0.1% (w/v) urea, at 37°C and 5% CO₂ and inoculated (2 × 10⁵ cells/well) in a 24-well plate. The UTI-ST1 and UTI-ST5 isolates were cultivated in TSB, with or without urea supplementation, to the mid-logarithmic growth phase and washed twice with PBS. Thereafter, the epithelial cells were infected with the UTI isolates at a 1:10 ratio. The adhesion assay was performed by co-incubating bacterial cells and epithelial cells for 2 h at 37°C and 5% CO₂. The cells were washed three times with PBS to remove the planktonic bacteria and lysed with 0.1% sodium deoxycholate (Sangon) to release the adhered bacteria. Bacterial CFU was determined by serial dilutions of epithelial cell lysates on TSA plates. For the invasion test, the cells were incubated with bacteria for 4 h, and gentamicin (100 μg/mL) was added to the culture for 30 min to digest the bacteria outside the cells. Thereafter, the bacterial CFU was enumerated to assess the invasion ability.

The nutrient-deficient medium with phenol red was used to perform the bacterial urease activity test. The overnight cultures of different isolates were centrifuged at 4,000 rpm and washed with PBS to remove the secreted urease. Bacteria were resuspended in the nutrient-deficient medium, with or without urea supplementation, and diluted to a final OD of 0.03–0.04. The urease activity was measured by continuously monitoring the OD at 560 and 415 nm, every 30 min, on the Micro-ELISA autoreader. Proteus mirabilis and E. coli were used as urease-positive and -negative controls, respectively.

The transcription of urease (ureACD) and accessory gene regulator (agr) effector (RNAIII) genes in UTI-ST1 and UTI-ST5 was detected by Quantitative reverse transcription PCR (RT-qPCR). Complementary DNA (cDNA) was synthesized from total RNA using the PrimeScript™ reverse transcriptase kit (Takara), according to the manufacturer’s instructions. Thereafter, the cDNA samples were amplified using the FastStart Universal SYBR Green Master kit (Roche). The reactions were performed on a 7500 Sequence Detector (Applied Biosystems). Purified chromosomal DNA (0.005–50 ng/mL) was used to construct a standard curve. The reactions were performed in triplicate, and DNA gyrase subunit B (gyrB) was used as an internal reference.

MLST was performed using seven housekeeping genes (arcC, areO, yqiL, glp, pta, tpi, and gmk). The sequences of the PCR products were compared with the references on the MLST website¹ for S. aureus.

The murine UTI model was constructed as described previously (Hagberg et al., 1983; Armbruster et al., 2018; Rashid et al., 2021). Female C57BL/6 mice (aged 6–8 weeks, housed 5/cage) were obtained from JSJ Laboratory Animals LTD (Shanghai, China) and subjected to bacterial infection after 3 days, after overnight water-deprivation, to void the bladders in case of immediate micturition after transurethral infection. Subsequently, the bacterial culture (5 × 10⁸ CFU/100 μL) was transurethrally injected into the anesthetized mice with a catheter needle (28G). Each mouse was inoculated with one kind of bacterial culture. Urine was collected every 24 h by massaging the abdomen of the mice and 10 μL of urine was used for serial dilution to calculate bacterial CFU. Bladders and kidneys were harvested for the quantification of bacterial burden (CFU/mL) at 72 hpi, by serial dilution plating and colony enumeration of homogenized organs after 24-h incubation.

Animal experiments were performed in ABSL2 facilities following the Guide for the Care and Use of Laboratory Animal Sciences (CALAS) and approved by the ethics committee of Renji Hospital, School of Medicine, Shanghai Jiaotong University, Shanghai, China (Approval Number: KY2021-225-B).

All statistical tests were performed with GraphPad Prism 8.0 software. Percentage values were analyzed pairwise by the two-tailed chi-square test or Fisher’s exact test. For comparisons of two groups, paired and unpaired two-tailed Student’s t-tests were applied. Error bars in all graphs indicated the standard error of the mean (mean ± SEM), and a p-value of <0.05 was considered statistically significant.

In this prevalence survey, a total of 4,405 non-repetitive S. aureus strains were collected from various clinical sources during 2008–2020 from a general Hospital in Shanghai, China (Table 1). Among these strains, UTI-ST5 (33.37%) was the most prevalent ST, followed by UTI-ST239 (12.10%), which has shown a gradual decline in recent years (Dai et al., 2019), and UTI-ST1 (6.74%; Table 1). Screening of each ST by specimen type revealed that the midstream urine specimen isolations of UTI-ST1 (11.59%, p < 0.001) were significantly higher than the other STs (Table 1). In addition, among the 193 UTI-SA strains, the isolation rate of UTI-ST5 (29.53%) was the highest, followed by the isolation rates of UTI-ST1 (18.13%) and UTI-ST7 (9.84%; Figure 1A; Table 2). Furthermore, both UTI-ST1 and UTI-ST5 were prevalent in the elderly population (aged ≥65) compared to the young population (88.57% vs. 11.43%, p < 0.05 and 63.16% vs. 36.84%, p < 0.05, respectively), while UTI-ST7, UTI-ST398, and UTI-ST188 were more frequently found in the younger populations (aged < 65; Table 2) and UTI-ST239 showed a slight advantage in elderly people (Table 2). Moreover, the majority of UTI-ST1 (65.85%) and UTI-ST5 (81.67%) isolates were recovered from men (Supplementary Figure S1). Therefore, owing to the high prevalence and isolation rates of UTI-ST1 and UTI-ST5, these samples were selected to further characterize the UTI-SA strains.

The antimicrobial pattern of UTI-ST1 was different from that of UTI-ST5 and other STs (Table 2). In general, UTI-ST5 and UTI-ST1 isolates were highly resistant to penicillin (100% and 97.14%, respectively) and cefoxitin (94.74% and 91.43%, respectively; Table 2). In contrast, gentamicin, erythromycin, and trimethoprim-sulfamethoxazole displayed potent antibiotic activity against UTI-ST1 (8.57%, 11.43%, and 2.86%, respectively), but not against UTI-ST5 (Table 2). In addition, among the STs, only UTI-ST1 exhibited resistance to levofloxacin (91.43%, p < 0.05; Table 2). The wide use of antibiotics in recent years has led to the generation of multiple antibiotic-resistance genes in bacteria, causing severe health problems. UTI-ST5 is a highly drug-resistant and less virulent ST, which is prevalent in the hospital, while UTI-ST398 and UTI-ST188 are community-associated STs with hypervirulence and low drug resistance. UTI-ST5 showed high drug resistance rates against most antibiotics, while UTI-ST1 displayed lower resistance to antibiotics, except for levofloxacin. Nitrofurantoin, trimethoprim-sulfamethoxazole, fosfomycin, and pivmecillinam are recommended for first-line therapy (Gupta et al., 2011) in UTI treatment, and β-lactams and fluoroquinolones are used as alternative treatments (Gupta et al., 2011, 2017), due to the prevalence of high antibiotic-resistance. This suggests that fluoroquinolone resistance in UTI-SA isolates is a major concern in UTI treatment.

The urinary tract is a urine-discharging system and human urine primarily consists of water (95%) and urea (2%; Sarigul et al., 2019; Volpin et al., 2019). The liver produces urea via the urea cycle, which is then released into the blood (Ha et al., 1947; Morris, 2002). The kidneys then filter the blood and concentrate urea and waste metabolites in the urine (Morris, 2002; Weiner et al., 2015). A previous study reported that urea supplementation in tryptic soy broth (TSB) supplemented with excess glucose, could reduce bacterial cell death during the stationary growth phase, due to ammonia generation and reduced intracellular reactive oxygen species levels (Zhou et al., 2019). To determine how UTI-ST1 and UTI-ST5 adapt to the high-urea environment of the bladder, we randomly selected 10 isolates from each group to test their growth, hemolytic ability, biofilm formation, and adhesion.

To investigate their growth competence, we conducted the growth assay in a modified nutrient-deficient medium, with or without urea supplementation (2%), at 37°C and 220 rpm. During the 8-h monitoring, the optical density (OD) at 600 nm was recorded every 30 min. The UTI-ST1 isolates grew significantly faster in the urea-supplemented medium than in the urea-unsupplemented medium (Figure 1B). In contrast, UTI-ST5 isolates did not show growth improvement with urea supplementation until 6-h growth (Figure 1C). These results suggest that UTI-ST1 can utilize urea to promote survival and growth, while UTI-ST5 displays a delayed utilization of urea. Furthermore, to assess the effect of urea supplementation on the pathogenicity of UTI-ST1 and UTI-ST5, we performed hemolysis and biofilm formation tests in TSB, with or without urea supplementation. After 10–12 h of cultivation, the supernatants of UTI-ST1 and UTI-ST5 from the urea-supplemented and -unsupplemented media were collected and incubated with erythrocytes (Ridder et al., 2021). Our results revealed that the cytolytic abilities of the UTI-ST1 isolates were highly weakened in the presence of urea (Figure 1D), while the hemolytic abilities of the UTI-ST5 isolates were not impacted by urea supplementation (Figure 1D). Similarly, UTI-ST1 biofilm was thicker in the urea-supplemented media, compared to that in the unsupplemented media (Figure 1E), while no significant difference was observed in the biofilms of the UTI-ST5 isolates, in the presence or absence of urea supplementation (Figure 1E). Since the UTI-ST5 strain is characterized by aggressive biofilm formation and hypovirulence (Figures 1D,E), we compared the hemolytic and biofilm-forming abilities of the non-UTI-derived (nUTI)-ST1 strains, which showed no significant differences in the presence and absence of urea supplementation (Supplementary Figures S2A,B). These results demonstrate that UTI-ST1 shows reduced virulence and thicker biofilm formation on polystyrene surfaces, such as an implant in the urea environment. However, the weakened virulence, although may cause less severe inflammatory reactions (Gong et al., 2014), biofilm formation in the urinary tract can reduce the flow of urine and antimicrobial agents (Scotland et al., 2019), facilitating persistent infection. Considering the elevated adhesion of UTI-ST1 to polystyrene in the urea-supplemented medium, we hypothesized that the adhesion of UTI-ST1 to the urinary epithelium could be improved by urea supplementation. Thus, 0.1% (w/w) urea was supplied to the cell medium to mimic the bladder microenvironment and it was observed that the UTI-ST1 isolates displayed stronger adhesion to the bladder epithelium 5,637 cells (Figure 1F), although their invasion ability was not improved (Supplementary Figure S2C).

We found that UTI-ST1 may shift to hypo virulent status and display increased biofilm formation on polystyrene, while the increased adhesion to epithelium helps UTI-ST1 to colonize and persist in the urinary tract. In addition, we hypothesized that UTI-ST5 may use its biofilm formation ability to survive in the urinary tract.

Based on the previous results, we hypothesized that UTI-ST1 and UTI-ST5 may have disparate urease activities. However, the conventional urease detection assay—Christensen urea agar—is qualitative and has limited sensitivity. Therefore, in this study, we used a modified semi-quantitative method to evaluate the urease activity precisely and easily (Roberts et al., 1978; Duran Ramirez et al., 2022). We inoculated diluted overnight culture in a modified urea-supplemented (2%, w/v) medium (pH 6.5) containing phenol red, a pH indicator, which would turn red under alkaline conditions, such as when ammonia is generated. The OD at 560 and 415 nm was recorded to monitor the color change of the solution and the results revealed that the urease activity of UTI-ST1 reached its peak at 2 h (Figure 2A), while all the selected ST5 strains, albeit harboring the wwurease operon confirmed by DNA sequencing (data not shown) and other strains did not show any obvious change in the urease activities (Figure 2A; Supplementary Figure S3). Thereafter, we quantified the urease activities of common UTI and nUTI strains at 2 h (Figure 2B) and found that UTI-ST1 and UTI-ST5 strains varied significantly in their urea-utilizing ability.

Thus, to determine the cause of the differences in the urease activities of the S. aureus strains, we selected 2 strains of UTI-ST1 and nUTI-ST1 each, for sequencing analysis. However, we found no variations in the sequences (mutations, insertions, or deletions) of urease genes between the two strains (data not shown). Therefore, we hypothesized that the variations in urease activities of the S. aureus strains may arise at the transcriptional level. S. aureus urease is a complex nickel-containing enzyme (Krajewska and Ureases, 2009). It consists of three subunits, α, β, and γ, which are encoded by UreC, ureB, and ureA, respectively, and coenzymes encoded by ureDEFH (Krajewska and Ureases, 2009; Konieczna et al., 2012). The results of quantitative reverse transcription PCR (RT-qPCR) revealed that the transcription of ureA, ureC, and ureD was significantly higher in UTI-ST1, compared to that in UTI-ST5, and nUTI-ST1 (Figures 2C–E), indicating the positive urease activities of UTI-ST1. Since the α-subunit, encoded by ureC, is the active site of the urease enzyme, we constructed the UTI-ST1 ureC mutant and a complemented strain to confirm the function of urease in ST1-induce UTI. The results revealed that the UTI-ST1 ureC mutant lost the urease activity (Figure 3A) and growth improvement in the urea environment (Figure 3B), indicating that the mutant could not utilize urea. However, the mutant showed no significant difference in the hemolytic and biofilm-forming phenotypes in the presence or absence of urea in the TSB medium (Figures 3C,D). While, the complemented strain recovered the urease activities (Figure 3A) and exhibited a similar hemolysis phenotype change as the WT strain (Figure 3C), confirming the function of urease. In addition, to verify the function of urease in vivo, we constructed a murine UTI model, by transurethrally injecting the same inoculum amount of UTI-ST1 WT, UTI-ST1 ureC mutant, and UTI-ST5 (Rashid et al., 2021). On the first day post-infection (24 hpi), both UTI-ST1 WT and UTI-ST1 ureC mutants were successfully established in most of the mice, with the mutant group displaying higher bacterial burden in the urine (Supplementary Figure S4), although the difference was not statistically significant. This was possibly because the mutant strain showed stronger virulence compared to the WT, leading to an intense immune response and discharge of more bacterial particles. Therefore, the CFU of the urine from the UTI-ST5-infected mice was slightly lower and relatively stable than that of the urine from the UTI-ST1-infected mice (Figure 3E; Supplementary Figure S4), which was consistent with its high isolation rate compared to all the STs. At 48 hpi, the bacterial burden of the ureC mutant-infected group declined rapidly (~10-fold) compared to that of the WT-infected group (Figure 3E), confirming the increased adhesion results that were obtained earlier. While at 72 hpi, the CFUs in the urine of UIT-ST1-WT dropped significantly probably because of the reduced virulence in urine for UTI-ST1 leading to phenotype like UTI-ST5. The bacterial burden of the bladder and kidney at 72 hpi suggested that the UTI-ST1 WT and UTI-ST1 ureC mutant mainly focused on the lower urinary tract, probably leading to cystitis (Figure 3F). However, UTI-ST5 displayed kidney colonization, suggesting that UTI-ST5 can cause ascending infections, such as pyelonephritis or bloodstream infection.

Previous studies have outlined urease as more than a virulence factor. Urease also acts as an acid resistance regulator to combat the low-pH environment (Cotter and Hill, 2003). With the growth of the bacteria, the pH of the culture drops, due to the accumulation of acid metabolites (Somerville and Proctor, 2009; Bronesky et al., 2016). We monitored the pH of the UTI-ST1 culture in urea-supplemented and unsupplemented TSB media for 8 h (Figures 4A,B) and found that in the absence of urea, the pH of the culture reduced to approximately 6.0 from 7.50 (Figure 4A). In the urea-supplemented media, the pH of the UTI-ST1 culture gradually rose to approximately 8.20 after 8 h, due to the urea hydrolysis caused by UTI-ST1 urease (Figure 4B); however, the pH of the UTI-ST5 and nUTI-ST1 cultures remained low (~6.5; Figure 4B). Thereafter, we monitored the continuous transcription of UTI-ST1 ureACD at 2, 4, and 6 h, on TSB without urea. At 4 h, when the culture pH was approximately 7.0, ureACD transcription was approximately 2–3 times higher than that at 2 h (pH 7.50). Furthermore, at 6 h, ureACD transcription declined 3–10 times compared to that at 4 h, with a drop in the culture pH (~6.0; Figures 4C–E). Meanwhile, both UTI-ST5 and nUTI-ST1 showed low urease gene transcription and low culture pH (Figures 4C–E). These results indicate that the urease gene transcription increases with the drop in pH and peaks in a neutral to mildly acidic environment. When the environmental pH continuously decreases, the transcription of urease would be repressed.

Urease gene expression is a complex network, and ccpA (Zhou et al., 2019), codY (Huang et al., 2014), and agr (Queck et al., 2008; Bronesky et al., 2016) have been predicted to be involved in the regulation of the urease gene. A study has reported ccpA and codY as positive and negative regulators of the urease gene expression, respectively (Zhou et al., 2019). In addition, Agr, a critical quorum sensing system in S. aureus, has been proven to upregulate urease gene expression; however, its role in urease gene expression during UTI pathogenesis has not yet been determined. The Agr system consists of two adjacent transcripts, RNAII and RNAIII, whose expression is driven by P2 and P3 promoters, respectively. RNAII transcript is an operon of four genes (agrBDCA) encoding the machinery of the quorum sensing system, whereas RNAIII transcript encodes the primary effector that regulates the expression of most agr-dependent downstream target genes (Peng et al., 1988). Therefore, we evaluated the transcription of Agr system-associated genes and ureAD. At 4 h of incubation, the transcription of RNAIII and hla, a downstream regulatory virulence gene of the Agr system, was reduced by approximately 2-fold and 10-fold in the urea-supplemented media (pH ~7.7) compared to that in the urea-unsupplemented media (pH ~7.0; Figures 5A,B), indicating that the Agr system was inhibited in the alkaline environment, due to ammonia generation from urea degradation. Meanwhile, at 4 h of incubation, the transcription of ureA and ureD in the urea-supplemented media (pH ~7) also reduced by approximately 2.5-fold and 7.4-fold, respectively, compared to that in the urea-unsupplemented media (Figures 5C,D). Combined with the phenotypic changes in the urea-supplemented media, the Agr system senses the variations in the environmental pH to adapt to the environmental conditions during the bacterial growth process. The Agr system functions actively under neutral pH conditions and declines under alkaline and acidic conditions (Regassa and Betley, 1992), suggesting that it may downregulate virulence genes and upregulate biofilm-forming genes, during UTI pathogenesis (Queck et al., 2008; Arciola et al., 2012). Finally, we constructed UTI-ST1 agrA mutant and complemented strains and found that the ureACD expression of the mutant was significantly declined compared to that of the UTI-ST1 WT (Figures 5E–G), suggesting that agr plays a role in upregulating the urease genes, which might help in maintaining the optimal environmental pH (Figure 6).