The impact of EmrA deficiency on antibiotic-susceptibility and virulence- and stone-associated phenotypes in Proteus mirabilis

Abstract

Background

Proteus mirabilis is a common uropathogen responsible for urinary tract infections (UTIs). The emrA gene encodes the EmrA protein, which is a critical component of the EmrAB-TolC multidrug efflux pump. In addition to its established role in antibiotic resistance, emerging evidence indicates that EmrA is also implicated in bacterial stress adaptation and virulence-associated processes. Nevertheless, its specific functions in the pathogenesis of P. mirabilis remain unclear.

Results

To define the function of EmrA in pathogenicity of P. mirabilis, this study combined phenotypic assays, transcriptomics, and a rat model of urinary tract infection. Comparison of the WT and ΔemrA strains showed that emrA deletion was associated with decreased MICs for carbenicillin and ampicillin, whereas the susceptibilities to most other antibiotics remained unchanged. Because MICs were not determined for the complemented strain, these results identify a deletion-associated susceptibility phenotype but do not establish an emrA-specific resistance mechanism. The study also found that EmrA promotes its adaptation to nutrient limitation, heat, and acid stress. However, the deletion of emrA significantly increased bacterial tolerance to oxidative stress. The mutant strain exhibited enhanced virulence-associated phenotypes, including increased swarming motility, biofilm formation, and improved adhesion and invasion capabilities of urothelial cells. Transcriptomic analysis revealed that emrA deletion upregulated multiple virulence-associated genes, including those encoding fimbriae components. In the rat UTI model, infection with the mutant was associated with higher bacterial burdens in the bladder and kidneys, and more severe inflammatory tissue damage; stone-associated outcomes were also increased in the models used. These stone-associated findings were accompanied by increased urease activity in the mutant strain.

Conclusion

This study identifies associations between emrA deletion and susceptibility changes to selected antibiotics, altered stress responses, and enhanced virulence- and stone-associated phenotypes in P. mirabilis. Because MICs were not measured for the complemented strain, the specific contribution of EmrA to the antibiotic-susceptibility phenotype remains unresolved.

Graphical abstract

Introduction

Urinary tract infections (UTIs) represent a major global health burden, ranking as the second most common bacterial infection after respiratory illnesses, with an estimated annual incidence affecting 150 million people (Flores-Mireles et al., 2015; Foxman, 2014). Among these, catheter-associated urinary tract infections (CAUTIs), as the most frequent hospital-acquired infections, have a mortality rate as high as 28.2%, more than twice that of non-CAUTIs (Hunt et al., 2024; Jacobsen et al., 2008). If not controlled promptly, UTIs can progress to life-threatening urosepsis. A severe complication of UTIs is the formation of infection stones, which constitute 10–15% of all urinary calculi. These stones, which can rapidly develop into staghorn calculi obstructing the renal tract, drive a cycle of urinary obstruction, hydronephrosis, and recurrent pyelonephritis, ultimately risking permanent kidney damage. Critically, bacteria within the stone matrix can evade antibiotics and host immune defense, rendering standard antimicrobial therapy ineffective and often making surgical removal necessary, despite recurrence rates of 30–50% post-operation (Schwartz and Stoller, 1999; Bichler et al., 2002; Flores-Mireles et al., 2015). The management of these complex infections is further complicated by the rising global prevalence of multidrug-resistant (MDR) pathogens (ElTaweel et al., 2024; Chakkour et al., 2024), highlighting an urgent need to develop novel therapeutic strategies.

Proteus mirabilis is a major pathogen in complicated urinary tract infections (UTIs), particularly those associated with catheters (CAUTIs), where it accounts for 10–44% of isolates (Armbruster et al., 2018; Jacobsen et al., 2008). Its pathogenicity is driven by distinct virulence mechanisms, most notably its characteristic swarming motility, which facilitates rapid surface colonization and retrograde dissemination from the urethra to the upper urinary tract. Additional key virulence factors include flagella, fimbriae, hemolysin, and specialized secretion systems, which collectively mediate adhesion, invasion, toxin production, and immune evasion (Debnath et al., 2018; Pearson et al., 2008; Armbruster et al., 2018). Notably, P. mirabilis is among the most potent urease-producing bacteria, exhibiting significantly higher activity than other common uropathogens such as Escherichia coli (Mobley and Hausinger, 1989; Yang et al., 2024; Chakkour et al., 2024). This robust urease activity is the primary driver of infection stone formation, a severe complication that exacerbates disease and complicates treatment. The synergistic effect of its multifaceted virulence and stone-forming capacity solidifies its role as a major pathogen in UTIs and related urolithiasis (Norsworthy and Pearson, 2017; Flores-Mireles et al., 2015). The increasing prevalence of multidrug-resistant P. mirabilis strains globally further underscores the urgent need to elucidate its pathogenesis, as understanding these mechanisms is vital for developing novel therapeutic strategies (ElTaweel et al., 2024; Chakkour et al., 2024).

Bacterial efflux pumps are fundamental antimicrobial resistance mechanisms in Gram-negative bacteria, actively expelling various toxic compounds, including antibiotics. Based on structure and energy coupling, efflux pumps are classified into six major families, including the Resistance-Nodulation-Division (RND) and Major Facilitator Superfamily (MFS) groups—based on their structure and energy coupling (Li and Nikaido, 2009; Nikaido and Takatsuka, 2009; Darby et al., 2023; Li et al., 2015). Among these, tripartite RND-type pumps, which span the inner and outer membranes, are particularly clinically significant due to their broad substrate specificity. EmrA is a membrane fusion protein that functions as a critical adaptor, linking the inner membrane transporter to the outer membrane channel in efflux complexes such as EmrAB-TolC (Hinchliffe et al., 2014; Tanabe et al., 2009; Li and Nikaido, 2009). Beyond its canonical role in multidrug efflux, emerging evidence indicates that EmrA is involved in broader physiological processes, including metabolite transport, biofilm formation, heavy metal tolerance, and can influence bacterial virulence in some species (Lin et al., 2017; Ma et al., 2014). However, its specific function in the pathogenesis of P. mirabilis, an important uropathogen, remains unclear. Given the critical need to understand the virulence mechanisms of this pathogen, elucidating the role of EmrA in P. mirabilis may provide insight into the biology of P. mirabilis urinary tract infection and identify hypotheses for future mechanistic investigation. Our preliminary pan-genome analysis of P. mirabilis revealed that emrA gene, encoding the membrane fusion protein EmrA, belongs to the species-specific core genome, prompting the hypothesis that EmrA plays a significant role in P. mirabilis pathogenicity. We therefore examined the association of emrA deletion with antibiotic susceptibility, stress adaptation, motility, virulence-associated phenotypes, and infection stone formation. Using the clinical isolate P. mirabilis DP2019 as the wild-type background, we constructed an isogenic emrA deletion mutant (ΔemrA) and its genetically complemented counterpart (CΔemrA). Comparative phenotypic analyses were integrated with transcriptomic profiling and RT-qPCR validation to identify pathways associated with the observed phenotypes.

Materials and methods

Bacterial strains and culture conditions

P. mirabilis DP2019 (GenBank CP110673), a clinical urinary isolate, was used as the wild-type (WT) strain. The isogenic emrA deletion mutant was constructed by allelic exchange using the suicide vector pCVD442Gm. Briefly, the upstream and downstream homologous fragments flanking emrA were amplified, fused, cloned into pCVD442Gm, introduced into the donor strain for conjugation, and transferred into P. mirabilis DP2019. Candidate double-crossover mutants were screened by antibiotic selection and sucrose counterselection and verified by PCR and Sanger sequencing. For complementation, the native emrA gene with its promoter was cloned into pRK415 and transferred into ΔemrA via conjugation, and CΔemrA was routinely grown in Lysogeny Broth (LB) at 37 °C. Tetracycline (10 μg/mL) was used during routine culture and inoculum preparation of CΔemrA for plasmid maintenance.

Minimum inhibitory concentration

MICs were determined by the broth microdilution method in Mueller-Hinton broth (Macklin) according to CLSI guidelines (Schuetz et al., 2025). Experiments were performed in triplicate.

Growth curve analysis

Overnight cultures were normalized to OD₆₀₀ = 0.6, diluted 1:100 into fresh LB or M9 minimal medium, and incubated at 37 °C with shaking. OD₆₀₀ was recorded every 2 h for 24 h. Oxygen-limited growth assays were performed under anaerobic conditions, and OD₆₀₀ was recorded every 2 h for 8 h. Three independent biological replicates were performed.

Stress survival assays

Bacterial suspensions (OD₆₀₀ ≈ 0.5 in PBS) were subjected to: (i) heat stress at 56 °C for 2 and 4 min; (ii) acid stress in LB (pH 4.0) at 37 °C for 2, 4 and 6 h; (iii) oxidative stress in 20 or 40 mM H₂O₂ at 37 °C for 15 min. Viable counts were determined by plating on LB agar. Survival rate (%) = (CFU after treatment/CFU before treatment) × 100. At least three biological replicates were performed.

Swarming motility assay

A 10 μL aliquot of normalized bacterial suspension (OD₆₀₀ = 0.5) was spot-inoculated onto blood agar plates (1.5% agar, 5% defibrinated sheep blood). Plates were incubated at 37 °C in a humidified atmosphere. Swarming colony diameters were measured at 2-h intervals for up to 12 h. Three independent replicates were performed.

Biofilm formation assay

Overnight cultures were diluted 1:100 in LB, and 1.5 mL aliquots were added to 24-well polystyrene plates. After static incubation at 37 °C for 24 h, planktonic cells were removed, and adherent biofilms were washed with PBS, stained with 0.1% crystal violet, and solubilized in extraction solution (ethanol: acetone, 8:2). Absorbance was measured at 570 nm. Experiments were performed in triplicate with technical duplicates.

Cell culture and in vitro infection assays

Cell culture. SV-HUC-1 (urinary bladder epithelial cells) and HK-2 (renal tubular epithelial cells) were cultured in DMEM/F-12 (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (Procell) at 37 °C in 5% CO₂.

Adhesion and invasion assays. Approximately 1 × 10⁵ cells/well were cultured in a 24-well plate with 1 × 10⁶ CFU/well of bacteria for 1 h. The inoculum of each strain was verified by CFU plating to ensure equivalent bacterial challenge. For adhesion, monolayers were washed and lysed with 0.5% Triton X-100. For invasion, after the 1-h co-incubation, cells were treated with gentamicin (100 μg/mL) for 1 h to kill extracellular bacteria, then lysed. Serial dilutions were plated for CFU enumeration. Adhesion and invasion rates were expressed as percentages of the initial inoculum, and mock-infected controls were included to monitor baseline host-cell viability.

Cytotoxicity assays. Approximately 1 × 10⁴ cells/well were cultured in a 96-well plate with 1 × 10⁶ CFU/well of bacteria for 2 h. Cell metabolic activity was assessed using the CCK-8 assay (NCM Biotech), and membrane damage was quantified by LDH release (Beyotime), following the manufacturers’ protocols. Cells were infected at an MOI of 100:1 for 2 h.

Rat model of acute urinary tract infection

Six- to seven-week-old female Sprague–Dawley rats were randomly divided into four groups: PBS-injected control, wild-type DP2019, ΔemrA and CΔemrA, with 6 rats per group. Animals were randomly assigned to treatment groups, and histological assessment was performed using consistent scoring criteria. The animals were acclimatized for 3 days and then deprived of water for 24 h prior to infection. Overnight cultures of DP2019, ΔemrA, and CΔemrA were grown in LB broth at 37 °C with shaking at 220 rpm. Bacterial cells were harvested, washed twice with sterile PBS, and resuspended to a final concentration of 1 × 10¹⁰ CFU/mL. The inoculum was verified by serial dilution and plating. For inoculation, rats were anesthetized by intraperitoneal injection of 1% sodium pentobarbital (40 mg/kg). Care was taken to avoid injecting into the thoracic cavity or bladder, which could cause death or insufficient anesthesia. Once the pedal reflex disappeared, the rat was placed in a supine position on a sterile surgical drape. The bladder was gently compressed to expel residual urine, the urethral orifice was disinfected with an alcohol swab, and a sterile drape was placed to expose the urethra. A 1-mL syringe was filled with the bacterial suspension or PBS and connected to a 24-gauge intravenous catheter. The catheter was primed with the suspension, and then the syringe was reloaded to the 0.1-mL mark. The catheter tip was lubricated with sterile glycerol and gently inserted into the urethra. When resistance was felt at the pubic symphysis, the catheter was flattened and carefully advanced into the bladder. A total of 100 μL of inoculum was slowly injected. Leakage from the urethra or catheter was monitored; if any leakage occurred, the corresponding volume was supplemented. After inoculation, the catheter was withdrawn slowly to minimize leakage and allow bacterial stabilization in the bladder. The rat was kept in a supine position with its pelvis elevated for 30 s, and any further leakage was observed. The animals were returned to their cages and kept warm; water was withheld until the following morning. At 24 h post-infection, rats were euthanized by intraperitoneal injection of 200 mg/kg pentobarbital sodium. Blood, bladder and kidneys were collected. The bladder was bisected longitudinally: one half was placed in 4% paraformaldehyde for fixation, and the other half in PBS for bacterial enumeration. The kidneys were also excised from the retroperitoneal space. The left kidney was decapsulated and fixed in 4% paraformaldehyde; half of the right kidney was placed in PBS for bacterial load determination. Bacterial loads were determined by plating tissue homogenates. Cytokine levels (IL-1β, IL-6, TNF-α) in bladder homogenates and serum were measured by ELISA. Tissues were processed for H&E and immunohistochemistry.

Vitro crystallization assays

Bacterial lysates were prepared by sonication, and protein concentrations were normalized using the BCA assay (Beyotime). For crystal formation (nucleation), 200 μL of lysate (0.5 mg/mL) was mixed with 400 μL of improved sterile artificial urine (KH₂PO₄ 1.36 g/L, NH₄Cl 0.54 g/L, urea 25 g/L, 0.1 mol/L MgCl₂ solution 100 mL, adjust to 1 L with purified water, then adjust pH to 7.0) (Sarigul et al., 2019) and incubated at 37 °C for 1 h. Crystals were counted under light microscopy (200×) in ten randomly selected fields per sample, using the same field-selection and counting criteria across all groups. For crystal aggregation, 2 mg of struvite seed crystals were suspended in artificial urine and mixed with lysate (0.3 mg/mL, 1:10 ratio). After incubation at 37 °C for 1 h, aggregates were quantified microscopically. The tight assembly of two or more individual crystals is defined as crystal aggregation. At least three biological replicates were performed.

Urease activity assays

Urease activity was measured by three methods: (i) phenol red colorimetric assay (0.001% phenol red in artificial urine, pH 6.0, absorbance at 550 nm); (ii) pH shift assay (direct pH measurement over 6 h); (iii) indophenol blue assay (Solarbio) for ammonia quantification. Bacterial lysates were normalized to 0.3–0.5 mg/mL protein as indicated.

RNA extraction and RT-qPCR

Total RNA was extracted using the GeneJET RNA Purification Kit (Thermo Fisher). cDNA was synthesized with the PrimeScript RT Reagent Kit (TaKaRa). RT-qPCR was performed using TB Green Premix Ex Taq II (TaKaRa) on a CFX96 Touch system (Bio-Rad). The gene gyrB was used as an internal control. Relative expression was calculated by the 2^−ΔΔCt method.

Rat model of chronic infection stone formation

All animal procedures were approved by the Institutional Animal Care and Use Committee of South China Agricultural University (Approval No. GZTH2020008; approval date: 2026-03-18). Female Sprague–Dawley rats (6 ~ 7 weeks old, 180 ~ 220 g) were randomly divided into four groups (n = 6 per group): PBS-injected control, WT, ΔemrA and CΔemrA. After 3 days of acclimatization, rats underwent surgical implantation of a zinc disc into the bladder. Anesthesia was induced by intraperitoneal injection of 1% sodium pentobarbital (40 mg/kg). The lower abdominal fur was shaved, and the skin was disinfected with iodophor. A suprapubic midline incision was made to expose and open the bladder, and a sterile zinc disc (4 mm diameter) was immediately inserted into the bladder through the incision. The bladder and abdominal muscle layers were closed using absorbable sutures, and the skin incision was closed with non-absorbable sutures. Ampicillin (20 mg/kg) was administered postoperatively to prevent surgical site infections. After a 7-day recovery period, rats were inoculated transurethrally with 10⁸ CFU of the respective bacterial strains, following the same procedure as described for the acute UTI model. This was designated as day 0 of infection. Booster inoculations of 10⁸ CFU and 10⁹ CFU were administered on days 6 and 12, respectively, at the same volume. On day 14 post-initial infection, rats were euthanised, and bladder stones, blood, kidney tissues and bladder tissues were collected for subsequent analyses. The excised stones were dried and weighed using an electronic balance. Bacterial loads in bladder, kidney and urine, as well as cytokine levels, were determined as described above.

RNA isolation, library construction, and sequencing

The WT and ΔemrA were inoculated into LB liquid medium at a ratio of 1:100 (vol/vol) and cultured at 37 °C with shaking at 220 rpm until the mid-logarithmic phase. The bacterial cells were then harvested by centrifugation at 8,000 rpm for 20 min, washed twice with phosphate-buffered saline (PBS), and the supernatant was discarded. The resulting cell pellets were stored at −80 °C for further use. Three independent biological replicates were performed for each strain. Total RNA was extracted from bacterial samples using TRIzol® Reagent according to the manufacturer’s protocol. The concentration and purity of the extracted RNA were assessed using a NanoDrop™ 2000 spectrophotometer, RNA integrity was evaluated by agarose gel electrophoresis, and the RNA integrity number (RIN) was determined using an Agilent 2,100 Bioanalyzer. Ribosomal RNA was removed and the enriched mRNA was randomly fragmented by the addition of fragmentation buffer. First-strand cDNA was synthesized from the fragmented mRNA using random hexamer primers. For second-strand synthesis, dUTP was substituted for dTTP in the dNTP mix, resulting in the incorporation of A/U/C/G bases into the second cDNA strand. End Repair Mix was then added to convert the sticky ends into blunt ends, followed by the addition of a single adenine (A) base to the 3′ ends for adapter ligation. The second-strand cDNA containing uracil was digested with UNG enzyme, ensuring that only the first-strand cDNA was retained in the library, which was subsequently amplified by PCR. After adapter ligation, the products were purified and subjected to size selection. The size-selected products were then amplified by PCR, and the final library was obtained after purification. Library quality was assessed, and sequencing was performed on the NovaSeq platform (Illumina). The RNA-seq data have been deposited in the SRA database under accession number PRJNA1472729.

Results

EmrA relates to resistance to some β-lactam antibiotics and environmental stress adaptability

To examine whether emrA deletion was associated with altered antimicrobial susceptibility, MICs were compared between the WT and ΔemrA strains. Relative to WT, Minimum inhibitory concentration (MIC) experiments revealed that the ΔemrA mutant exhibited significantly increased sensitivity to β-lactam antibiotics, showed an 8-fold reduction in the MIC for carbenicillin and a 2-fold reduction for ampicillin. In contrast, the MIC for tetracycline increased, whereas the MIC of most other tested antibiotics remained unchanged (Table 1). These results identify an antibiotic-specific susceptibility phenotype associated with the ΔemrA genotype rather than a generalized change in antimicrobial susceptibility. However, because MICs were not determined for CΔemrA, the present data do not demonstrate restoration of the ampicillin or carbenicillin phenotype and cannot establish that these MIC changes are specifically attributable to loss of EmrA.

Table 1.

Determination of minimum inhibitory concentration.

Antibiotics	DP2019	ΔemrA
Levofloxacin	0.125 μg/mL	0.125 μg/mL
Imipenem	2 μg/mL	2 μg/mL
Ceftriaxone Sodium	<0.125 μg/mL	<0.125 μg/mL
Gentamicin	2 μg/mL	2 μg/mL
Polymyxin B	>128 μg/mL	>128 μg/mL
tigecycline	16 μg/mL	16 μg/mL
Furantoin	>128 μg/mL	>128 μg/mL
Tetracycline	32 μg/mL	64 μg/mL
Compound Sulfamethoxazole	>128 μg/mL	>128 μg/mL
Ampicillin	1 μg/mL	0.5 μg/mL
Carbenicillin	4 μg/mL	0.5 μg/mL
Nalidixic Acid	32 μg/mL	32 μg/mL
Chloramphenicol	>128 μg/mL	>128 μg/mL

Subsequently, deletion of emrA is associated with antibiotic-specific susceptibility changes and altered stress adaptation. The growth of the ΔemrA mutant was comparable to WT in nutrient-rich LB medium, but showed a significant growth defect during the log phase in nutrient-limited minimal M9 medium (Figures 1A,B). On this basis, anaerobic environment was added, and it was found that the mutant strain grew significantly slower (Figures 1C,D). Furthermore, the mutant showed markedly reduced survival under heat shock (56 °C) and acidic (pH 4.0) conditions compared to the wild-type (WT) and complemented strains (CΔemrA) (Figures 1E–H). In contrast, the survival rate of the ΔemrA mutant under oxidative stress (20 and 40 mM H₂O₂) exhibited significantly enhanced trend (Figures 1I,J). These data indicate that EmrA contributes to adaptation to selected environmental stresses, including nutrient limitation, heat, and acid, but unexpectedly suppresses bacterial survival under oxidative conditions.

Figure 1.

Phenotypic characterization of the ΔemrA mutant under stress conditions. (A,B) Growth curves of WT, ΔemrA, and CΔemrA in nutrient-rich LB medium (A) and M9 minimal medium (B) (comparisons are versus the WT strain). Optical density at 600 nm (OD₆₀₀) was measured every 2 hours for 24 h. Data are shown as mean ± SD (n = 3). (C,D) Growth curves of WT, ΔemrA, and CΔemrA in nutrient-rich LB medium (C) and M9 minimal medium (D) (comparisons are versus the WT strain) under oxygen stress conditions. Optical density at 600 nm (OD₆₀₀) was measured every 2 hours for 8 h. Data are shown as mean ± SD (n = 6). (E,F) Bacterial counts expressed as log₁₀ CFU/mL (E) and corresponding survival rates (F) after exposure to heat stress (56 °C for 2 min and 4 min). Survival rate was calculated as (CFU after heat treatment/CFU of untreated control) × 100%. (G,H) Bacterial counts (log₁₀ CFU/mL) (G) and survival rates (H) following exposure to oxidative stress (20 mM and 40 mM H₂O₂ for 15 min). (I,J) Bacterial counts (log₁₀ CFU/mL) (I) and survival rates (J) under acid stress (pH 4 stress at 2 h and 4 h). Survival rates at 4 h and 6 h are shown relative to the 2 h time point. Statistical analysis: All experiments were performed at least three independent times with triplicate samples. Data are presented as mean ± SD. Comparisons among groups were performed using one-way ANOVA followed by Tukey’s post-hoc test (for survival rate comparisons) or two-way ANOVA with repeated measures (for growth curves). Significance levels are indicated as: * P < 0.05, ** P < 0.01, *** P < 0.001; n.s., not significant.

EmrA deletion promotes a hyper-motile and hyper-biofilm-forming phenotype

Based on established connections between efflux systems and surface-associated behaviors, we investigated the role of EmrA in swarming motility and biofilm formation. Results showed that the ΔemrA mutant displayed a significantly larger swarming diameter than the WT and CΔemrA (Figures 2A,B), demonstrating that EmrA is consistent with a negative association of this form of motility. Simultaneously, quantified by crystal violet staining revealed that the biofilm-forming ability of the ΔemrA mutant was significantly enhanced (Figures 2C,D). These results indicate that loss of emrA is associated with enhanced motility and biofilm-related phenotypes in P. mirabilis.

Figure 2.

Effects of emrA deletion on swarming motility, biofilm formation, and host cell interactions. (A,B) Swarming motility of P. mirabilis WT (DP2019), ΔemrA mutant, and CΔemrA on blood agar plates. (A) Quantification of swarming diameters (cm) (B) Representative images of swarming rings after 24 h of incubation. Data are presented as mean ± SD (n = 3 independent experiments). (C,D) Biofilm formation capacity after 24 h of static culture. (C) Quantification of biofilm biomass measured as optical density at 595 nm (OD570). (D) The three panels of each strain show representative wells after crystal violet staining, respectively. Data are shown as mean ± SD (n = 3 independent experiments). (E–L) Adhesion and invasion assays using SV-HUC-1 human urothelial cells and HK-2 human kidney proximal tubular epithelial cells. Bacterial adhesion (E,G) and invasion (I,K) were determined by plate counting (log₁₀ CFU/well). The relative adhesion rates (F,H) and relative invasion rates (J,L) were calculated by normalizing the values of the WT (DP2019) group to 100%. Data are presented as mean ± SD (n = 3 independent experiments). (M,N) Cell viability assessed by CCK-8 assay after infection of SV-HUC-1 (M) and HK-2 (N) cells with the indicated strains. Viability of uninfected control cells was set to 100%. Data are shown as mean ± SD (n = 3). (O,P) Lactate dehydrogenase (LDH) release in the supernatants of SV-HUC-1 (O) and HK-2 (P) cells after bacterial infection. The LDH level of the WT (DP2019) group was set to 100% to calculate relative release rates. Data are presented as mean ± SD (n = 6). Statistical analysis: All experiments were performed at least three independent times. Comparisons among three or more groups were performed using one-way ANOVA followed by Tukey’s post-hoc test. Significance levels are indicated as: * P < 0.05, ** P < 0.01, *** P < 0.001; n.s., not significant.

EmrA deletion is associated with enhanced cellular pathogenicity in vitro

The virulence-associated phenotype of the ΔemrA mutant was evaluated using human urothelial cell lines. Compared to the WT and CΔemrA, the ΔemrA mutant exhibited significantly enhanced adhesion to and invasion of both SV-HUC-1 (bladder) and HK-2 (kidney) cells (Figures 2E–L). Cytotoxicity experiments (CCK-8 viability and LDH release assays) further indicated that ΔemrA infection led to reduced host cell survival and increased cell membrane damage (Figures 2M–P). These data suggest that emrA deletion enhances several in vitro cell-associated pathogenic phenotypes of P. mirabilis.

The ΔemrA mutant shows enhanced virulence-associated phenotypes and exacerbated tissue damage in vivo

A rat model of ascending urinary tract infection was employed to assess the in vivo phenotype associated with emrA deletion. Results showed that the bacterial loads in the bladder and kidneys of rats in the ΔemrA infection group were significantly higher than those in the WT (DP2019) and CΔemrA group (Figures 3A,B). Accompanied by significantly elevated levels of the pro-inflammatory cytokines (IL-1β, IL-6, and TNF-α) in the bladder tissues of ΔemrA-infected rats (Figures 3C–E). Histopathological analysis revealed that ΔemrA infection caused bladder epithelial exfoliation, edema, and extensive inflammatory cell infiltration, whereas the WT and CΔemrA groups showed only mild focal inflammation (Figure 3H). Immunohistochemistry confirmed strong, widespread expression of IL-6 and TNF-α in the bladders of ΔemrA-infected group (Figures 3F,G,I,J). The infection with the ΔemrA mutant was also associated with significant renal dysfunction, manifested as elevated serum creatinine and urea levels, indicating impaired glomerular filtration (Figures 3K,L). Concurrently, serum magnesium and phosphorus levels were significantly decreased (Figures 3M,N). Histology of kidney tissues from the mutant group showed epithelial shedding and interstitial inflammation, consistent with the biochemical evidence of injury (Figure 3O).

Figure 3.

In vivo pathogenicity of the ΔemrA mutant in a rat model of urinary tract infection. (A,B) Colonization capacity of P. mirabilis strains in rat bladder (A) and kidney (B) tissues. Bacterial loads are expressed as log₁₀ CFU per gram of tissue. Each symbol represents an individual animal (n = 6 per group). Horizontal lines indicate the median. (C–E) Levels of pro-inflammatory cytokines in bladder tissue homogenates measured by ELISA: IL-6 (C), IL-1β (D), and TNF-α (E). Data are presented as pg./mg total protein, mean ± SD (n = 6 per group). (H) Representative hematoxylin and eosin (H&E) staining of bladder tissue sections from rats infected with the indicated strains. Scale bar = 200 μm (left panels) and 50 μm (right panels). (F,G,I,J) Immunohistochemical analysis of IL-6 and TNF-α expression in bladder tissue sections. (F,G) Quantitative analysis of IL-6-positive area (%) and TNF-α-positive area (%). (I,J) Representative images of IL-6 and TNF-α staining (brown signal). Data are presented as mean ± SD (n = 5 random fields per section from 3 rats per group). (K–N) Serum levels of biochemical markers indicative of renal function and electrolyte balance. (K) Creatinine (μmol/L). (L) Magnesium (mmol/L). (M) Phosphorus (mmol/L). (N) Urea (mmol/L). Each symbol represents an individual animal (n = 6 per group). Horizontal lines indicate the median. (O) Representative hematoxylin and eosin (H&E) staining of kidney tissue sections from rats infected with the indicated strains. Scale bar = 200 μm (left panels) and 50 μm (right panels). Statistical analysis: Normality of the data was assessed using the Shapiro–Wilk test. For non-normally distributed data (bacterial loads, cytokine levels, and IHC quantification), pairwise comparisons between groups were performed using the Mann–Whitney U test. For normally distributed data, one-way ANOVA followed by Tukey’s post-hoc test was applied. All tests were two-tailed. Significance levels are indicated as: * P < 0.05, ** P < 0.01, *** P < 0.01; n.s., not significant. For IHC data, * indicates a statistically significant difference compared to the DP2019 (WT) group (P < 0.05).

Notably, macroscopic bladder stones were observed exclusively in rats infected with the ΔemrA mutant. The urine from these animals was significantly more alkaline, and the characteristic decrease in serum magnesium and phosphorus—a biochemical hallmark of active struvite stone formation—provided further corroboration. Collectively, these in vivo data demonstrate that EmrA deletion is associated with increased urinary tract colonization, tissue damage and a microenvironment conducive to infection stone formation.

The enhanced ability of Proteus mirabilis to form stones after EmrA deficiency is related to the increased urease activity

To evaluate the association between EmrA and infection stone-related phenotype, we performed complementary in vitro analyses. In vitro crystallization experiments showed that the crystal nucleation and aggregation abilities of the ΔemrA mutant were significantly enhanced compared with the WT and CΔemrA (Figures 4A–D). This pro-calculogenic phenotype was linked to a fundamental increase in ureolytic activity. Further detection revealed that Urease activity of the ΔemrA mutant, measured by multiple independent assays (measured via pH changes, phenol red colorimetry, indophenol blue method), was significantly higher than that of WT and CΔemrA (Figures 4E–G). At the molecular level, RT-qPCR analysis revealed that the mutant exhibited upregulated expression of key urease-associated genes, including the structural subunits (ureA, ureB, ureC), accessory proteins (ureE, ureG), and the transcriptional activator ureR (Figures 4H,I).

Figure 4.

EmrA deficiency enhances crystal formation, aggregation and urease activity in P. mirabilis. (A,B) Crystal formation assay. (A) Quantification of crystal formation in artificial urine. (B) Representative microscopic images of crystals formed by each strain (scale bar = 50 μm). (C,D) Crystal aggregation assay. (C) Quantification of crystal aggregation. (D) Representative images of crystal aggregates. (E–G) Urease activity measured by three independent methods. (E) pH increase over time in urea-containing medium. (F) Phenol red colorimetric assay (OD550). (G) Indophenol blue colorimetric method, quantifying ammonia production. All data are presented as mean ± SD (n = 3 independent experiments). (H,I) Relative expression levels of urease subunit genes and the transcriptional regulator ureR determined by quantitative real-time PCR (qPCR). Expression levels were normalized to the housekeeping gene (gurB) and are shown relative to the WT (DP2019) strain, which was set to 1. Data are presented as mean ± SD (n = 3 biological replicates). Statistical analysis: Normality of the data was assessed using the Shapiro–Wilk test. Comparisons among three or more groups were performed using one-way ANOVA followed by Tukey’s post-hoc test. Between-group differences were assessed using an independent-samples t-test. Significance levels are indicated as: * P < 0.05, ** P < 0.01, *** P < 0.001; n.s., not significant.

EmrA deficiency is associated with increased stone formation in a chronic infection model

A rat zinc disc implantation model was used to assess stone formation during chronic infection. The results showed that the stone mass of the ΔemrA mutant infection group showed an increasing trend compared to the WT group, and was significantly higher than that of the CΔemrA group (Figures 5A,B). This enhanced stone burden was associated with significantly higher bacterial loads in the bladder, kidneys, and urine of rats infected with the ΔemrA mutant, consistent with enhanced colonization and persistence (Figures 5C–E). Analysis of the host response in this chronic model revealed a distinct profile. Contrary to the acute model, chronic ΔemrA infection showed reduced bladder and serum IL-1β, IL-6, and TNF-α despite high bacterial loads (Figures 5F–I). This discrepancy may reflect the inflammatory peak may have already subsided in the chronic phase or prolonged bacterial persistence might drive immune exhaustion or modulation. Meanwhile, the formation of stones may physically isolate bacteria from immune cells, which can also affect the occurrence of immune responses.

Figure 5.

EmrA deficiency promotes bladder stone formation, enhances bacterial colonization, and modulates inflammatory responses in a chronic infection stone model. (A,B) Bladder stone formation in rats infected with P. mirabilis strains. (A) Representative images of bladder stones retrieved from rats infected with WT (DP2019), ΔemrA mutant, or complemented (CΔemrA) strain. (B) Quantification of stone weight (mg). Each symbol represents an individual animal (n = 6 per group). Horizontal lines indicate the median. (C–E) Bacterial colonization capacity. Bacterial loads in bladder tissue (C), kidney tissue (D) and urine (E) are expressed as log₁₀ CFU per gram of tissue or per mL of urine. Each symbol represents an individual animal; horizontal lines indicate the median. (F–I) Inflammatory cytokine levels. (F) IL-6, (G) IL-1β, and (H) TNF-α in bladder tissue homogenates (pg/mg total protein). (I) Serum IL-6 levels (pg/mL). Each symbol represents an individual animal; horizontal lines indicate the median. Statistical analysis: Normality of the data was assessed using the Shapiro–Wilk test. As the data were not normally distributed, pairwise comparisons between groups (ΔemrA vs. WT, ΔemrA vs. CΔemrA, and WT vs. CΔemrA) were performed using the Mann–Whitney U test. All tests were two-tailed. Significance levels are indicated as: * P < 0.05, ** P < 0.01, *** P < 0.001; n.s., not significant.

Transcriptomic profiling identifies expression changes associated with EmrA deletion

To explore pathways associated with the ΔemrA mutant phenotypes, we conducted comparative transcriptomic analysis (RNA-seq). Sample-level expression distributions were first evaluated (Figure 6A). Basic functional annotation analysis was performed based on the reference genome GCF_026167565.1. RNA-seq analysis showed that emrA deletion led to differential expression of 134 genes (|log₂FC| ≥ 1, padj <0.05), including 54 upregulated and 80 downregulated genes (Figure 6B). Functional enrichment analysis revealed that upregulated genes were primarily enriched in virulence-related pathways. Key virulence-associated operons were significantly induced, including genes encoding the Type VI secretion system (tssB, tssC, tssF), fimbrial components (fimA), and hemolysin (hpmB). Correspondingly, Gene Ontology (GO) and KEGG pathway analyses highlighted significant enrichment for biological processes related to fimbrial assembly, cell adhesion, biofilm formation, and protein secretion (Figures 6C–F). The urease subunit genes (ureA, ureB, ureC, ureE, ureG) showed an upregulation trend but did not reach a statistically significant level (padj>0.05). In contrast, downregulated genes were predominantly enriched in core metabolic pathways, most notably in amino acid biosynthesis and metabolism. Genes involved in arginine synthesis (argA, argB, argC, argF) and transport (artP) were significantly suppressed (Figures 6G–J). This transcriptional signature indicates that EmrA deletion is associated with changes in bacterial resource allocation, and is associated with virulence gene expression while inhibiting basic metabolism.

Figure 6.

Transcriptomic profiling of the mutant compared to the WT strain. (A) Box plot showing the distribution of gene expression levels (log₁₀(TPM + 1)) in WT and ΔemrA samples. The boxes represent the interquartile range (IQR), with the line inside each box indicating the median. Whiskers denote the 5th and 95th percentiles. (B) Volcano plot of differentially expressed genes (DEGs). The x-axis represents log2(fold change) (log2FC) between ΔemrA and WT, and the y-axis represents −log10(adjusted p-value). Red dots: Significantly upregulated genes (log2FC ≥ 1, adjusted p < 0.05). Green dots: Significantly downregulated genes (log2FC ≤ −1, adjusted p < 0.05). Blue dots: Non-significant genes. (C,D) Functional annotation of upregulated DEGs. (C) GO annotation analysis (level 2) showing the distribution of upregulated DEGs across biological process, cellular component, and molecular function categories. (D) KEGG annotation analysis showing the top enriched pathways for upregulated DEGs. (E,F) Enrichment analysis of upregulated DEGs. (E) GO enrichment analysis displaying significantly enriched GO terms. (F) KEGG enrichment analysis displaying significantly enriched KEGG pathways. (G,H) Functional annotation of downregulated DEGs. (G) GO annotation analysis (level 2) showing the distribution of downregulated DEGs across biological process, cellular component, and molecular function categories. (H) KEGG annotation analysis showing the top enriched pathways for downregulated DEGs. (I,J) Enrichment analysis of downregulated DEGs. (I) GO enrichment analysis displaying significantly enriched GO terms. (J) KEGG enrichment analysis displaying significantly enriched KEGG pathways. Statistical analysis: Differential expression analysis was performed using DESeq2 (V1.42.0). Genes with |log2(fold change)| ≥ 1 and an adjusted p-value (Benjamini-Hochberg correction) < 0.05 were considered significantly differentially expressed. For GO and KEGG enrichment analyses, significance was defined as an adjusted p-value < 0.05 (Benjamini-Hochberg corrected). All RNA-seq experiments were performed with three biological replicates per condition.

Discussion

In this study, we investigated the association of the membrane fusion protein EmrA of P. mirabilis with antimicrobial susceptibility, stress adaptation, multiple virulence-associated phenotypes, and infection stone formation in P. mirabilis. Using an isogenic emrA deletion mutant, complemented strain, in vitro phenotypic assays, transcriptomic profiling, and rat infection models, we found that deletion of emrA decreased MICs for carbenicillin and ampicillin, increased the MIC for tetracycline, and left most tested antibiotics unchanged; emrA deletion also displayed impaired adaptation to nutrient limitation, heat, and acidic stress conditions, but increased survival under oxidative-stress; In addition, emrA deletion was associated with enhanced swarming motility, biofilm formation, host-cell adhesion and invasion, and urease activity, and infection stone-associated phenotypes. In a rat model of urinary tract infection and an infectious stone model, the mutant was associated with more severe bladder and kidney colonization, a stronger inflammatory response and increased bladder struvite stone formation. Transcriptome analysis revealed that EmrA deficiency upregulated virulence gene clusters such as T6SS, fimbriae, hemolysin, urease, stress-response, and metabolic genes. While downregulating universal stress protein genes. The complemented strain often shifted phenotypes toward WT, supporting involvement of emrA. Together, this complex profile suggests that EmrA may shape the balance among membrane transport, stress physiology and virulence-associated programs rather than controlling a single linear resistance pathway, although the underlying regulatory mechanisms remain to be defined.

Bacteria exposed to antibiotic and host immune pressures may display resource-based or regulatory trade-off between antimicrobial resistance and virulence (Beceiro et al., 2013). In the present study, comparison of the WT and ΔemrA strains showed that emrA deletion was associated with lower MICs for ampicillin and carbenicillin, a higher MIC for tetracycline, while most other antibiotics were unaffected. This pattern indicates an antibiotic-specific susceptibility phenotype rather than a generalized change in antimicrobial resistance. However, these findings should not be interpreted as definitive evidence that EmrA directly mediates resistance to ampicillin or carbenicillin, because restoration of the MIC phenotype was not evaluated in CΔemrA. Beyond antimicrobial susceptibility, the ΔemrA mutant grew comparably to WT in nutrient-rich LB medium but displayed reduced growth under nutrient-limited and oxygen-limited conditions and reduced survival following heat and acid stress. In contrast, survival under oxidative-stress was increased. These condition-dependent responses suggest that emrA genotype is associated with altered stress adaptation rather than a generalized increase or decrease in bacterial fitness. Such environmental conditions may be relevant to the urinary tract infection, where bacteria can encounter nutrient restriction, variable oxygen tension, acidic microenvironments, and host-derived oxidative stress. The ΔemrA mutant also exhibited enhanced virulence-associated phenotypes. We therefore hypothesize that the physiological changes associated with emrA deletion may shift the balance between selected stress-adaptation functions and virulence-associated traits. This interpretation remains provisional because the underlying regulatory mechanism has not been established. Context-dependent relationships between antimicrobial resistance determinants and virulence have also been reported in other bacteria. For example, loss of the porins OmpC/OmpD in Enterobacter hormaechei confers carbapenem resistance but severely attenuates virulence (Perault et al., 2025). The cfr-like genes drive linezolid resistance while simultaneously impairing sporulation, biofilm formation, and toxin production in C. difficile, illustrating a resistance-virulence association. Similar links between efflux systems and virulence-associated physiology have been described in other pathogens, including Pseudomonas aeruginosa, deletion of the MexAB-OprM efflux pump also affects quorum sensing and virulence gene expression (Alcalde-Rico et al., 2020). These studies provide conceptual evidence that resistance-associated systems can influence broader bacterial physiology, but they do not establish the mechanism operating in P. mirabilis. The coexistence of reduced MICs for selected β-lactam antibiotics and enhanced virulence-associated phenotypes in the ΔemrA mutant is consistent with a possible resistance–virulence association. However, the current findings do not establish a genetically validated mechanistic link between the antibiotic-susceptibility phenotype and virulence-associated changes, because the antibiotic-susceptibility phenotype was evaluated only in the WT and deletion strains. The increased oxidative-stress tolerance of the mutant may partly explain its enhanced persistence in vivo, whereas its reduced heat and acid tolerance indicates that emrA deletion does not globally increase stress resistance. Additional genetic controls and mechanistic experiments will be required to determine whether and how EmrA links antimicrobial susceptibility, stress adaptation, and virulence-associated physiology.

Our study found that emrA deficiency is associated with enhanced virulence-associated phenotypes in P. mirabilis. Swarming motility is a key mechanism for the retrograde spread of P. mirabilis from the bladder to the kidneys, and its enhancement may facilitate bacterial colonization of the bladder from the urethra and subsequent ascent to the kidneys. Biofilms serve as a protective barrier for bacteria against host immunity and antibiotic attack; enhanced biofilm formation increases bacterial colonization efficiency on catheter surfaces and the urothelium, thereby promoting resistance to antibiotic killing and immune clearance and serving as a source of recurrent infection. Our results showed that the mutant exhibited significantly increased swarming migration on solid media and enhanced biofilm formation. This regulatory pattern has been validated in other bacteria. In E. coli, deletion of multiple TolC-dependent efflux pumps, including EmrAB-TolC, resulted in increased swimming motility except ΔemrY and ΔacrZ (Di Maso and Ruiz, 2024). Moreover, it was observed that in the presence of cationic antimicrobial compounds, deletion mutants of several secondary multidrug efflux pumps, including EmrA, displayed significantly increased biofilm formation (Bay et al., 2017). These studies indicate that efflux pump components potentially influence bacterial motility and biofilm formation.

Adhesion and invasion are critical steps in bacterial infection of host cells. Adhesion is a prerequisite for bacterial colonization of the urothelium, while invasion allows bacteria to evade host immune clearance and cause deep tissue damage. Our study showed that the mutant exhibited significantly higher adhesion and invasion rates to both SV-HUC-1 and HK-2 urothelial cells compared to the WT and CΔemrA, suggesting that emrA deletion may enhance bacterial-host cell interactions through altered expression of fimbriae, adhesins or other surface-associated factors. Cytotoxicity assays using CCK-8 and LDH release confirmed that the survival of urothelial cells treated with the knockout strain was significantly decreased, while LDH release was increased, indicating that emrA deletion enhances the direct damage capability of bacteria to host cells. To validate the physiological relevance of these in vitro phenotypes, we established a rat model of urinary tract infection. The results showed that emrA deletion was associated with enhanced in vivo pathogenicity-associated phenotypes. The bacterial loads of the knockout strain in bladder and kidney tissues were significantly higher than those of the WT and CΔemrA, indicating that EmrA loss is associated with increased bacterial colonization and proliferation in the urinary tract, thereby laying the foundation for subsequent infection progression. In the mutant group, the levels of pro-inflammatory cytokines such as IL-6, IL-1β and TNF-α in bladder tissue homogenates were significantly elevated. Histopathological examination (HE staining) revealed more severe inflammatory cell infiltration and epithelial damage and immunohistochemistry further confirmed the local enrichment of inflammatory factors, suggesting that emrA deletion is associated with exacerbates inflammatory damage to the urinary tract mucosa either by enhancing bacterial pathogenicity or by directly modulating the host immune response. Furthermore, serum creatinine and urea levels were significantly increased in the mutant group, indicating impaired renal function. Together, these results demonstrate that EmrA deficiency is associated with more severe pathological outcomes of urinary tract infection by enhancing in vivo colonization capacity and pro-inflammatory effects. EmrA is associated with the virulence in P. mirabilis and its deletion results in a consistent virulence gain effect across the entire chain from motility and colonization to tissue damage.

Combined with transcriptomic sequencing results, we observed that multiple virulence-related genes were significantly upregulated in the mutant strain, which may contribute to the enhanced virulence-associated phenotypes. First, the expression of core components of the Type VI secretion system (T6SS), including tssB, tssC, tssF and Vgr family protein genes, was upregulated in the mutant strain, consistent with a molecular explanation for the increased cytotoxicity. T6SS is a contact-dependent killing apparatus that can directly disrupt host cell membrane integrity through perforin-like effector proteins, and its activation leads to increased LDH release and reduced cell viability (Reglinski et al., 2025). Second, potential fimbriae genes were significantly upregulated, consistent with a structural basis for the enhanced adhesion and invasion. Fimbriae are the primary tools for bacterial adhesion to host cells, and their increased expression may improve binding efficiency to receptors on urothelial cells, thereby promoting invasion and initial biofilm formation (Gahlot et al., 2022). In addition, upregulation of the haemolysin accessory protein gene hpmB suggests enhanced haemolysin activity, which together with T6SS may further aggravate cell damage (Uphoff and Welch, 1990). The coordinated activation of these virulence genes partially explains at the molecular level the phenotypes of increased adhesion/invasion, elevated cytotoxicity and enhanced biofilm formation after EmrA deletion.

Infectious stones are a serious complication of P. mirabilis infection, and their formation depends on urease-mediated urea hydrolysis and urine alkalinisation (Fitzgerald et al., 2024). In this study, the mutant strain exhibited significantly enhanced capacity in both crystal nucleation and aggregation assays, along with markedly increased urease activity. Mechanistically, this was attributed to upregulation of the urease transcriptional activator UreR and the urease structural genes. Although the transcriptomic data revealed a similar trend for urease-encoding genes, the increase did not reach statistical significance, which we attribute to differences in detection sensitivity between the two methods. This finding is consistent with EmrA to the transcriptional regulation of urease, whereas previously the activation of the urease operon was known to rely mainly on UreR and nitrogen sensing. These results suggest a possible relationship between EmrA and the urease system, but they do not demonstrate direct transcriptional regulation. It should be noted that the crystal formation and aggregation assays were performed using bacterial lysates rather than live bacteria, which may not fully reflect the dynamic physiological processes, such as active metabolism and continuous urease production, that occur during in vivo infection and stone formation. But, in a chronic infectious stone model, the stones formed by the mutant showed a certain upward trend compared to WT and CΔemrA, supporting an association with the in vitro findings without establishing a direct causal pathway. Of note, in this chronic model, the mutant group displayed a lower local inflammatory response than the WT group, despite higher bacterial loads. This seemingly paradoxical phenomenon may reflect immune exhaustion induced by persistent infection with a higher bacterial burden, or a physical barrier effect resulting from extensive stone formation (Weißelberg et al., 2024), suggesting that EmrA deficiency not only enhances stone-forming capacity but may also promote chronic infection by reshaping the host immune response. This apparent discrepancy between enhanced virulence-associated phenotypes and attenuated inflammation precisely illustrates the complexity of bacterial resource reallocation strategies and reminds us that different dimensions of pathogenicity should be distinguished when evaluating bacterial virulence.

Further analysis of the transcriptomic data revealed that, in contrast to the significantly upregulated virulence genes, several metabolism- and stress-related genes were markedly downregulated in the mutant strain. Downregulation of the arginine synthesis pathway and the arginine transporter gene artP suggests that the mutant strain may be associated with reduced resource investmentin energy-consuming amino acid synthesis. Moreover, significant downregulation of the universal stress protein genes OOZ57_RS09740(KO name: uspG) and OOZ57_RS12380 (KO name: uspF) reveals a dual role for EmrA in environmental adaptation. Universal stress proteins are induced under various environmental stresses such as acid, heat and nutrient starvation (Zhang et al., 2025; Weißelberg et al., 2024). Therefore, their downregulation may explain the reduced survival of the knockout strain under heat and acid stress. The paradoxical phenotype of increased oxidative stress tolerance suggests that compensatory activation of the antioxidant system may occur independently of the regulation of universal stress proteins, a mechanism that warrants further investigation. Arginine is a precursor for polyamine synthesis, and polyamines are involved in biofilm formation and motility regulation (Liu et al., 2022; Scribani Rossi et al., 2022; Shi et al., 2019). However, the mutant strain did not show reduced biofilm or motility. On the contrary, these were significantly enhanced. We propose the following explanations. First, the marked upregulation of T6SS core components, fimbriae genes and the haemolysin accessory protein may directly enhance bacterial adhesion, aggregation and surface migration, and this positive driving effect overcomes any potential negative impact of reduced polyamines. Second, bacteria may maintain basal polyamine levels through compensatory activation of the ornithine decarboxylase pathway or exogenous polyamine uptake to support virulence functions. Finally, the downregulation of the arginine pathway reflects a resource redistribution strategy that actively reduces energy intensive arginine synthesis and reinvests resources in the effective production of virulence factors, which may reflect a resource redistribution strategy associated with enhanced virulence gene expression (Kratz and Banerjee, 2024; Whittle et al., 2024). This seemingly contradictory phenomenon precisely illustrates the complexity of global transcriptional reprogramming in the knockout strain, which is associated with enhanced virulence-associated phenotypes through differential gene expression. This resource reallocation strategy is consistent with the possible association between antibiotic susceptibility and virulence-associated phenotypes. In summary, the transcriptomic data identify gene-expression changes associated with emrA deletion and provide a basis for mechanistic hypotheses. However, these data do not establish direct regulation or a complete molecular pathway, which will require targeted biochemical investigation and further genetic validation.

Traditional antibiotics impose strong selective pressure and can promote resistance spread (Ayres, 2024; Lu et al., 2025). Anti-virulence strategies have been proposed as an approach that aims to attenuate pathogens without directly challenging their survival, potentially reducing the selection for resistance (Hopkins et al., 2024). Our observation that emrA deletion enhances virulence-associated phenotypes raises the question of whether maintaining EmrA function could, in principle, be explored as a concept for limiting pathogenicity. However, this idea remains entirely speculative and would require substantial experimental validation—including gain-of-function studies and direct mechanistic investigation—before any therapeutic implications could be considered. Therefore, we do not propose EmrA as a target for therapeutic development at this stage, but rather highlight the association as a starting point for future mechanistic research.

The direct interaction between EmrA and known virulence-associated regulatory proteins has not been clarified. While transcriptomic data revealed differential expression of multiple virulence gene clusters, we have not yet identified the protein interaction partners of EmrA using methods such as co-immunoprecipitation, bacterial two-hybrid screening or crosslinking mass spectrometry. Whether EmrA directly binds UreR, TssB or other regulatory proteins, or acts through indirect pathways, remains unclear. Construction of deletion mutants of key regulators and examination of their genetic epistatic relationships with emrA will help to define the regulatory hierarchy in which EmrA operates. Furthermore, validation of the anti-virulence target is still at a conceptual stage. We have not performed small-molecule screening, nor have we constructed strains with conditional overexpression of emrA to assess the therapeutic effect of enhancing EmrA function. In addition, epidemiological surveys have not yet been conducted to determine whether natural mutations or polymorphisms in emrA exist in clinical isolates, and whether such mutations are associated with enhanced virulence or stone-forming capacity. In summary, this study reveals an association between EmrA and multiple phenotypic traits in P. mirabilis, including specific antibiotic susceptibility changes, stress adaptation, and virulence- and stone-associated phenotypes. EmrA deficiency is associated with reduced expression of stress-response genes and increased expression of virulence-associated genes, which may contribute to, leading to more severe infection and stone complications in animal models. Transcriptomic analysis identified gene expression changes associated with emrA deletion, involving pathways such as T6SS, fimbriae, and haemolysin, but the underlying molecular mechanisms remain to be determined. The observation that emrA deletion correlates with both reduced β-lactam susceptibility and enhanced virulence-associated phenotypes raises the question of whether the EmrAB-TolC system might be explored as a target for anti-virulence research. However, such a possibility remains hypothetical and would require substantial experimental validation, including gain-of-function studies and small-molecule screening, before any therapeutic implications can be considered. This study also acknowledges the current limitations of the evidence, particularly the lack of direct molecular interactions between EmrA and known virulence regulators, thereby pointing to specific directions for future mechanistic investigation. Overall, our findings expand the understanding of the biological functions of bacterial efflux pumps and provide a basis for further research into the role of EmrA in the pathogenesis of urinary tract infections.

Conclusion

This study shows that emrA deletion is associated with antibiotic-specific susceptibility changes, altered stress responses, and enhanced virulence- and stone-associated phenotypes in P. mirabilis. Because the MIC phenotype was not tested in the complemented strain, the direct contribution of EmrA to ampicillin and carbenicillin resistance remains unresolved. These findings support further investigation of the emrAB locus but do not yet establish EmrA as a genetically validated regulator of antimicrobial resistance. Although EmrA appears to play a role in intrinsic resistance to some β-lactam antibiotics and in stress adaptation, its genetic inactivation was associated with a phenotype exhibiting enhanced virulence-associated traits, including increased swarming, biofilm formation, adhesion, invasion, urease activity, and stone formation. The combined evidence from phenotypic, transcriptomic, and animal studies suggests that the absence of EmrA is associated with a phenotype showing enhanced virulence-associated traits, with higher expression of virulence determinants and lower expression of stress-response genes. These observations provide additional insight into the pleiotropic effects associated with emrA deletion in P. mirabilis. The association between EmrA and virulence-associated phenotypes may warrant further investigation of the EmrAB-TolC system in the context of P. mirabilis pathogenesis, although any therapeutic implications remain entirely speculative at this stage.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the Guangdong Provincial College Students Innovation and Entrepreneurship Training Program (grant S202510570071). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm.nih.gov/, PRJNA1472729.

Ethics statement

The animal study was approved by Laboratory Animal Ethics Committee of South China Agricultural University. The study was conducted in accordance with the local legislation and institutional requirements.

Author contributions

XJ: Methodology, Validation, Data curation, Writing – original draft. LiD: Writing – original draft, Validation, Methodology. TX: Validation, Methodology, Writing – original draft. LinD: Writing – original draft, Validation. HL: Validation, Writing – original draft. QH: Methodology, Writing – original draft, Investigation. HY: Writing – original draft, Validation. LC: Validation, Writing – original draft. ZM: Writing – original draft, Validation, Methodology. YZ: Methodology, Investigation, Writing – original draft. PL: Methodology, Resources, Project administration, Writing – review & editing. DX: Funding acquisition, Resources, Project administration, Writing – review & editing, Methodology, Supervision.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb.2026.1895236/full#supplementary-material