Targeted lysosomal activation in bladder epithelium enhances clearance of intracellular uropathogenic Escherichia coli

Abstract

Recurrent urinary tract infections (UTIs) are a major clinical burden, driven in part by the ability of uropathogenic Escherichia coli (UPEC) to establish intracellular niches within the bladder epithelium, where bacteria withstand antibiotics and host defenses. The oral bacterial lysate OM-89 (Uro-Vaxom), a clinically approved and globally used therapy for the prevention and management of recurrent UTIs for several decades, reduces recurrence rates, but its cellular mechanisms of action remain incompletely understood. Here, we demonstrate that OM-89 strengthens antimicrobial defenses in bladder epithelial cells and, in combination with antibiotic therapy, limits post-antibiotic bacterial regrowth in epithelial infection models. OM-89 promotes lysosomal acidification and increases lysosomal protease activity in bladder organoids and differentiated epithelial monolayers, thereby directing intracellular UPEC toward degradative compartments. In parallel, OM-89 enhances intracellular accumulation of multiple antibiotic classes. These effects are conserved across distinct UPEC strains and in both murine and human epithelial models. Our findings demonstrate that bladder epithelial antimicrobial pathways can be pharmacologically reinforced to influence treatment outcomes by enhancing intracellular bacterial clearance.

Author summary

Urinary tract infections are common, and for many people recur after antibiotic treatment appears to succeed. This happens partly because some bacteria survive treatment hidden inside the cells lining the bladder, allowing infection to return once treatment stops. OM-89, an orally-given bacterial lysate already used to prevent recurrent infections, is thought to work mainly by stimulating the immune system, but whether it also acts directly on bladder cells has been unclear. Using bladder organoids and cell models from both mice and humans, we show that OM-89 strengthens the bladder’s own antibacterial defenses. It boosts the activity of lysosomes, the cellular compartments degrading harmful material, helping infected cells destroy bacteria more effectively. OM-89 also increased how much antibiotic entered bladder cells. Together, these effects improved bacterial clearance and reduced regrowth after antibiotic treatment. This protective effect held across multiple bacterial strains and in cell models from both mice and humans, and the underlying pathway is linked to natural immune defense activity in independent human bladder tissue datasets, supporting its relevance beyond our own models. Our findings offer a possible cellular explanation for how OM-89 helps prevent recurrent urinary tract infections and point toward more effective antibiotic treatment against this common infection.

Introduction

Intracellular persistence of bacterial pathogens represents a major barrier to successful infection clearance, as antibiotics often fail to efficiently accumulate within host cells [1,2] and due to the pathogens’ ability to escape degradative compartments [3,4]. This limitation is particularly relevant for mucosal epithelia, which act as the first physical barrier during many infections. Whether epithelial cell-intrinsic antimicrobial pathways can be therapeutically reinforced to enhance intracellular bacterial clearance remains insufficiently explored.

Urinary tract infections (UTIs) exemplify this challenge [5–7]. Uropathogenic Escherichia coli (UPEC), the predominant pathogen of UTIs, invades bladder epithelial cells and establishes protected intracellular niches that promote persistence and recurrence [8]. These include rapidly replicating intracellular bacterial communities [9] and long-lived quiescent reservoirs [10], both of which are shielded from immune responses and incompletely eradicated by antibiotics [8]. Recurrent UTIs therefore drive repeated antibiotic exposure, accelerating emergence of antimicrobial resistance and increasing the risk of treatment failure [11]. Despite the clear role of intracellular bacterial reservoirs in the recurrence of UTIs, strategies aimed at strengthening epithelial antimicrobial mechanisms rather than directly targeting UPEC remain rare.

The bladder epithelium is an immunologically active tissue capable of pathogen sensing, trafficking control and lysosomal degradation [12]. For example, bladder epithelial cells expel invading UPEC via fusiform vesicles or traffic invading bacteria via endo-lysosomal pathways or autophagosomes into degradative compartments and secrete antimicrobial peptides into the urine [13]. Experimental and clinical data suggest that modulating mucosal immunity can reduce recurrent UTIs [14]. While this observation highlights the coordinated contribution of epithelial barriers and mucosal immune cells to host defense, the epithelial cell-intrinsic effector mechanisms underlying protection remain poorly defined.

OM-89 (Active Pharmaceutical Ingredient of Uro-Vaxom), an orally administered lysate derived from 18 different E. coli strains, is clinically approved and has been used for nearly four decades for the prevention of recurrent UTIs [15,16]. OM-89 is approved in numerous countries and is widely prescribed as prophylaxis in patients with recurrent lower UTIs, either alone or in combination with antibiotic therapy. Clinical trials have demonstrated that OM-89 significantly reduces recurrence rates [17–19] and experimental studies suggest that it exerts immunomodulatory effects through activation of innate immune cells and stimulation of antimicrobial humoral responses [20–22]. However, despite extensive clinical use and experimental investigation, the cellular mechanisms underlying its protective effects remain incompletely understood. For instance, rodent infection studies have demonstrated protective effects of OM-89 alone [23,24] and in combination with antibiotic therapy [25,26], although this observed in vivo protection could not be linked to any major quantitative changes in bladder immune cell infiltration [25], leaving the underlying molecular mechanism not fully resolved. However, Canton et al. speculated that the bladder epithelium itself may act as a direct effector site, supported by pharmacokinetic evidence showing urinary accumulation of OM-89-derived components [27]. In parallel, other bacterial lysates have been shown to directly modulate epithelial responses in the respiratory tract [28–30], supporting the concept of epithelial conditioning across different mucosal surfaces. Together, these observations raise the possibility that the bladder epithelium, and specifically epithelial cell-intrinsic antimicrobial pathways, may contribute to OM-89-mediated protection.

Here, we investigated whether OM-89 directly modulates antimicrobial pathways in bladder epithelial cells using a previously established murine bladder organoid model [31], a human-derived bladder organoid model and differentiated epithelial monolayers that recapitulate urothelial differentiation. We show that OM-89 remodels endo-lysosomal networks and enhances lysosomal acidification and protease activity, indicating that epithelial antimicrobial pathways can be pharmacologically reinforced to enhance intracellular bacterial clearance. Together with increased intracellular accumulation of antibiotics across different classes, these changes are associated with improved intracellular clearance and reduced bacterial regrowth across diverse UPEC strains. The key features of this phenotype, including lysosomal expansion, acidification, protease activation and enhanced antibiotic uptake, are conserved in human bladder epithelial monolayers and organoid cultures, despite species-specific differences in autophagic flux regulation. Together, these findings reveal a previously unrecognized lysosome-centered epithelial mechanism by which OM-89 strengthens bladder epithelial antimicrobial defenses and enhances intracellular bacterial clearance, identifying enhanced lysosomal function as a therapeutically targetable component of host defense.

Results

OM-89 enhances epithelial control of intracellular UPEC and limits post-antibiotic bacterial regrowth

To determine whether OM-89 enhances bladder epithelial control of intracellular UPEC during antibiotic treatment, we used a previously established and characterized mouse bladder organoid model [31] infected with the well-characterized UPEC strain CFT073 [32], originally isolated from a pyelonephritis patient. We monitored bacterial growth, antibiotic-mediated bacterial clearance and post-antibiotic regrowth within the organoids. To model relevant treatment scenarios with the clinically used therapy Uro-Vaxom, we applied OM-89 under three different regimens (Fig 1A): (i) pre-application – OM-89 exposure for 72 hours followed by a 24-hour rest period before microinjection of UPEC; (ii) co-application – OM-89 exposure initiated at the same time as antibiotic treatment; and (iii) continuous application – OM-89 exposure both 72 hours before microinjection and throughout the entire experiment. We quantified UPEC fluorescence area overlapping with the organoid area as a proxy for viable burden and regrowth inside each organoid, normalized within experiments to PBS controls. Hence, the microscopy data represent the total “intra-organoid” bacterial burden at each experimental stage, without distinguishing the exact localization of the bacteria – which can be luminal, intracellular or tissue-associated. Notably, OM-89 did not influence bacterial loads during the initial infection phase compared to the PBS control groups (4h post-infection, pi; S1A-C Fig), indicating no antimicrobial effects of OM-89 alone. However, OM-89 significantly reduced bacterial regrowth following antibiotic treatment (10h pi; 10x MIC of ampicillin) compared to the PBS control in all three treatment regimens (Fig 1B-D), with the effect being most pronounced when OM-89 was present during the antibiotic treatment phase (co-application, Fig 1C; and continuous application, Fig 1D). These positive effects persisted up to eight hours post-antibiotic withdrawal (15h pi) in the continuous application regimen (Fig 1E).

Fig 1. OM-89 reduces regrowth of CFT073 in mouse bladder organoids due to enhanced antibiotic-mediated clearance.

(A) Mouse bladder organoids treated with OM-89 under different regimes (pre-application: OM-89 exposure for 72h and 24h rest period; co-application: OM-89 exposure together with antibiotic treatment (at 7h post-infection, pi); continuous application: OM-89 exposure both 72h before microinjection and throughout the entire experiment). After microinjection of fluorescently labelled UPEC (0h pi), bacterial growth (4h pi), antibiotic-mediated clearance (7h pi), and regrowth 3 hours after antibiotic withdrawal (10h pi) were monitored. Antibiotic treatment was initiated straight after monitoring bacterial growth at 4h pi and kept for 3 hours. Created in BioRender. Tomasek, K. (2026) https://BioRender.com/syqz10y. (B)-(D) CFT073 signal 10h pi (3h regrowth) at 10x MIC in (B) pre-application, (C) co-application, and (D) continuous application regimes in mouse bladder organoids. Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test. N ≥ 15 per condition for (B), n ≥ 17 per condition for (C) and n ≥ 23 per condition for (D). Images show representative organoids (grey, surface reconstruction) and UPEC (magenta, some marked with arrows). (E) CFT073 signal 15h pi (8h regrowth) at 10x MIC in the continuous application regime in mouse bladder organoids. Values normalized to PBS control (dashed line). Mean ± 95% CI. Mann-Whitney test. N ≥ 15 per condition. Images show representative organoids (grey, surface reconstruction) and UPEC (magenta, some marked with arrows). (F)-(H) CFT073 signal 7h pi (antibiotic-mediated clearance) at 10x MIC in (F) pre-application, (G) co-application, and (H) continuous application regimes in mouse bladder organoids. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test for (F), Mann-Whitney test for (G), (H). N ≥ 15 per condition for (F), n ≥ 17 per condition for (G) and n ≥ 14 per condition for (H).

We next examined the effect of OM-89 during antibiotic treatment (7h pi). While the pre-application regimen showed modest modulation of antibiotic effects (Fig 1F), bacterial loads within organoids were significantly reduced in conditions where OM-89 was present at the time of antibiotic treatment (co-application, Fig 1G; continuous application, Fig 1H). Following bacterial burden in individual organoids in the continuous application regimen over time, we observed that OM-89 consistently enhanced bacterial clearance throughout antibiotic treatment and regrowth phases (S1D Fig). Finally, as observed previously [31], we found that regrowth of UPEC occurred preferentially from the organoid wall – irrespective of treatment with OM-89 (S1E Fig, S1 and S2 Video).

To further investigate OM-89’s modulating effect, we focused on the continuous application regime for the rest of the manuscript as this treatment regime most likely reflects the clinical application in patients.

First, given the enhanced antibiotic-mediated clearance in OM-89-treated cells, we asked if OM-89 could alter antibiotic accumulation inside bladder epithelial cells, as many conventional antibiotics, including β-lactam antibiotics such as ampicillin, display low intracellular accumulation levels [33]. We therefore examined the uptake of fluorescently labeled ampicillin into differentiated monolayers of mouse bladder epithelial cells (S2 Fig) and observed significantly increased accumulation of fluorescently labelled ampicillin in bladder epithelial cells treated with OM-89 (Fig 2A). Importantly, UPEC infection alone did not increase intracellular antibiotic accumulation compared to uninfected controls (Fig 2A), indicating that enhanced uptake is a specific response to OM-89 exposure rather than a general E. coli-driven stimulation response. Furthermore, the increased antibiotic uptake was predominantly observed in umbrella-like cells (cytokeratin (CK)20 positive cells), whereas intermediate cells (CK13 positive cells) showed little uptake (S1F Fig), identifying terminally differentiated cells as the main contributors to the uptake phenotype. The effect of increased antibiotic uptake extended to gentamicin, an aminoglycoside antibiotic with usually slow and insufficient intracellular accumulation dynamics commonly used in infection assays to eradicate only extracellular bacteria [33–35] (S1G Fig). It is important to note that the TFP ester of the fluorescent dye used to label ampicillin and gentamicin preferentially reacts with the non-protonated form of free amine groups, which may alter the antibiotics’ intracellular behavior and potentially affect their antimicrobial activity. Nonetheless, OM-89 treatment led to increased intracellular accumulation of both fluorescently labelled ampicillin and gentamicin, as well as Dextran-TMR, a polysaccharide uptake marker (S1H Fig), suggesting a generalized enhancement of cellular uptake in bladder epithelial cells compared to PBS-treated controls. However, because the FITC labeling may alter antibiotic behavior, our uptake measurements reflect fluorescent proxies rather than native drug activity.

Fig 2. OM-89 enhances uptake of antibiotics into mouse bladder epithelial cells and reduces intracellular UPEC burden.

(A) Uptake of FITC-labelled ampicillin into differentiated monolayers of mouse bladder epithelial cells with and without UPEC infection. Quantification of fluorescence after background subtraction of PBS- or OM-89-treated cells without labelled antibiotic. Each dot represents one field of view. Mean ± 95% CI. Kruskal-Wallis test with Dunn’s correction. N ≥ 6 per condition. Z-projection (maximum intensity) of representative images. FITC-ampicillin shown in cyan. (B) CFT073 signal 7h pi (antibiotic-mediated clearance) and 10h pi (regrowth) in mouse bladder organoids using 1x MIC of ampicillin. Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Mann-Whitney test. N ≥ 15 per condition. Images show representative organoids (grey, surface reconstruction) and UPEC (magenta, some marked with arrows). (C, D) Intracellular colony-forming unit (CFU) enumeration in differentiated mouse bladder epithelial monolayers following the continuous application regimen. Intracellular CFUs were quantified in PBS- and OM-89-treated cells treated with (C) 100 µg/ml gentamicin for 1h or (D) with an additional 3h treatment with ampicillin (10x MIC). Mean ± 95% CI. Welch’s t-test for (C) and Mann-Whitney test for (D). N = 15 per condition.

To determine whether the increased uptake of antibiotics contributed to OM-89’s protective effect, we reduced the antibiotic concentration from 10-fold to 1-fold the MIC. Even under these conditions, OM-89 maintained its ability to significantly reduced bacterial burden within three hours (7h pi) and suppress regrowth up to three hours post-antibiotic treatment (10h pi) (Fig 2B), suggesting that OM-89 improves antibiotic efficacy potentially by increasing intracellular accumulation.

To determine whether increased intracellular antibiotic accumulation contributed to the enhanced bacterial clearance observed in organoids, we quantified intracellular bacteria by colony-forming unit (CFU) assays in differentiated bladder epithelial monolayers, as CFU measurements in Matrigel-embedded organoids are technically challenging. Following infection, extracellular bacteria were removed performing a gentamicin protection assay and intracellular bacteria were quantified by plating cell lysates. Surprisingly, under these conditions OM-89 significantly reduced intracellular bacterial burden compared with PBS-treated controls (Fig 2C). Given the usually weak antimicrobial effects of gentamicin against intracellular bacteria, this result suggests an additional antimicrobial effect triggered by OM-89 early after UPEC invasion. Finally, extending the antibiotic treatment phase with ampicillin, after gentamicin removal, resulted in a moderate, but non-significant additional reduction in intracellular CFUs in OM-89-treated cells (Fig 2D). OM-89 did not alter the MIC values of either gentamicin or ampicillin (Table 1), indicating that the observed effects do not arise from direct modulation of bacterial susceptibility.

Table 1. MICs for CFT073 with OM-89 treatment.

Antibiotic	PBS	OM-89
Gentamicin	0.61	1.63
Ampicillin	6.43	7.00

Together, these data show that OM-89 enhances epithelial control of intracellular UPEC, promotes bacterial clearance during antibiotic treatment and sustains suppression of post-treatment regrowth, consistent with additionally improved intracellular antimicrobial activity.

OM-89 reroutes epithelial degradative pathways toward a lysosome-dominant antimicrobial state

To define how OM-89 alters epithelial antimicrobial pathways during infection, we performed transcriptomic profiling of infected and treated bladder organoids.

Principal component analysis revealed a strong separation between control and infected organoids (Fig 3A), consistent with widespread differential gene expression across conditions (Fig 3B, S3A Fig). Gene set enrichment analysis (GSEA) confirmed that CFT073 infection induced pathways associated with inflammation, bacterial immune signaling, and cell death, whereas OM-89 treatment attenuated the magnitude of these infection-associated signatures (Fig 3C and S3B and S3C Fig). Notably, OM-89-treated organoids clustered closer to infected samples than to PBS controls (Fig 3A), indicating that OM-89 induces a transcriptional program partially overlapping with infection-responsive pathways. This overlap is further supported by elevated cellular responses to pathogenic stimuli in OM-89-treated organoids compared to uninfected controls (S3D Fig). However, OM-89-treated organoids following infection remained transcriptionally distinct from PBS-infected samples, suggesting that OM-89 does not simply recapitulate an infection state but reconfigures epithelial responses in a qualitatively distinct, host defense-oriented manner.

Fig 3. OM-89 modulates inflammatory signaling and promotes lysosomal pathways in bladder organoids.

(A) Principal component analysis (PCA) of mouse bladder organoids treated with OM-89 or PBS, with and without CFT073 infection (PBS – dark blue, OM-89 – cyan, PBS+UPEC – light blue, OM-89 + UPEC – green). (B) Volcano plots of differentially expressed genes (log₂FC ≥ 1; adjusted p ≤ 0.05) comparing PBS-treated organoids with (light blue) and without infection (dark blue), and OM-89- (green) versus PBS-treated organoids during infection (light blue). (C) Gene set enrichment analysis (GSEA) of treatment and infection groups. GO-BP terms were clustered into infection-relevant pathways. NES, normalized enrichment score. Non-significant changes not highlighted. (D) GSEA of the top 10 Kyoto Encyclopedia of Genes and Genomes (KEGG) terms comparing OM-89- and PBS-treated organoids during infection. Ns changes not highlighted. (E) Enrichment plot of lysosomal pathway genes (KEGG term) in OM-89- versus PBS-treated organoids during infection (FDR corrected p-value = 1.88e-14 and uncorrected p-value = 1.21e-16). (A)-(E) Data from four independent wells with n ≥ 50 organoids each.

GSEA analysis additionally indicated that the lysosomal pathway was the most significantly induced gene signature following OM-89 treatment during infection (Fig 3D) and further gene set enrichment analysis of lysosomal genes confirmed their involvement in OM-89’s response to infection (Fig 3E). Although lysosomal pathways were also among the top 10 upregulated gene sets in uninfected organoids exposed to OM-89 (S3D Fig), the strong upregulation of lysosome-associated terms, together with autophagy-related pathways, was driven by the combination of infection and OM-89 treatment, as infection in the PBS control group did not upregulate these pathways (S3C Fig). These data suggest that OM-89 promotes pathways linked to intracellular cargo degradation, particularly under infectious conditions.

We next examined whether OM-89 modulates intracellular trafficking of UPEC in differentiated monolayers of mouse bladder epithelial cells during infection. We first focused on the endosomal pathway by analyzing Rab11A and Rab27B, two small GTPases implicated in UPEC expulsion from bladder epithelial cells [36]. Rab11A marks recycling endosomes [37], while Rab27B is involved in endosomal secretion [38] and late endosome/lysosome trafficking [39]. OM-89 treatment significantly increased the number of Rab11A- and Rab27B-positive vesicles, without affecting vesicle size (Fig 4A, 4B).

Fig 4. OM-89 remodels vesicle organization and turnover in bladder epithelial cells during UPEC infection.

(A)-(D), (F), (G) Quantification of vesicle numbers and size for (A) RAB11A, (B) RAB27B, (C) P62, (D) LC3B, (F) RAB7, and (G) LAMP1 in infected monolayers of mouse bladder epithelial cells with and without OM-89 treatment. Values normalized to PBS control (dashed line). Mean ± 95% CI. Mann-Whitney test for RAB11A and RAB7 (A), (F); Welch’s t test for RAB27B, P62, LC3B and LAMP1 (B)-(D), (G). N = 12 per condition. Z-projection (maximum intensity) of representative images. Foci, cyan; DAPI, grey. (E) Western blot analysis of LC3B during infection of monolayers. LC3B-II/GAPDH ratio normalized to PBS control (dashed line). Mean ± 95% CI. Kruskal-Wallis test with Dunn’s correction. Autophagic flux (BafA1-blocked minus unblocked LC3B-II/GAPDH): PBS = 24.75 ± 10.68; OM-89 = 65.87 ± 9.11. N = 3 per condition.

To assess autophagy responses to OM-89 during UPEC infection, we next quantified canonical autophagy markers [40,41]. OM-89 significantly increased both the number and size of ubiquitin-binding protein p62 vesicles (Fig 4C), suggesting enhanced cargo sequestration into autophagosomes. In parallel, OM-89 did not significantly change the number of microtubule-associated protein 1A/1B-light chain 3B (LC3b) vesicles, but it significantly increased LC3b vesicle size (Fig 4D), indicating the formation of larger autophagosomes. To determine whether these morphological changes translated into altered autophagic flux during UPEC infection and OM-89 treatment, we performed immunoblotting for membrane-bound LC3b-II in the presence and absence of V-ATPase inhibitor bafilomycin A1 (BafA1) blocking the autophagosome-lysosome fusion. In PBS-treated bladder epithelial cells, UPEC infection stalled flux, as membrane-bound LC3b-II levels did not increase further upon lysosomal inhibition (Fig 4E), consistent with previously described mechanisms of UPEC-mediated autophagy blockade [42]. In contrast, OM-89 treatment overruled this pathogen-induced stalling, as LC3b-II accumulated significantly with BafA1, demonstrating restoration of autophagic flux upon OM-89 treatment.

Finally, we assessed late endosomal [43] and lysosomal markers [44]. OM-89 treatment substantially increased both the number and size of Rab7-positive vesicles during UPEC infection, consistent with the expansion of the late endosomal compartment (Fig 4F). Lysosomal-associated membrane protein 1 (Lamp1)-positive vesicles were also strongly increased in number during infection; however, their size was reduced (Fig 4G), potentially indicating modulation of the lysosomal network. Interestingly, in agreement with the induction of lysosome-associated genes observed in OM-89-treated uninfected organoids (S3D Fig), OM-89 treatment alone also increased the number of Lamp1-positive vesicles in the absence of infection, while vesicle size remained unchanged (S4A Fig).

Together, these findings show that OM-89 shifts epithelial trafficking toward a degradative state by restoring autophagic flux and expanding lysosome-associated compartments.

Lysosomal activation is required for OM-89-mediated control of intracellular bacteria

To test whether the observed effects on lysosomal pathways could mechanistically, at least in parts, explain OM-89-mediated protection, we first used Genebridge analysis [45] to examine how the lysosomal gene signature identified in our RNA-seq data relates to host defense programs in the human bladder. To evaluate the translational relevance of our experimental findings, we used a computational Module-Module Association Score (MMAS) analysis across eight independent human bladder transcriptomic datasets comprising over 1,400 clinical samples. This network-based approach evaluates the transcriptional correlation between the lysosomal gene network and functional biological pathways across diverse human cohorts. Module-Module association analysis performed on these human bladder datasets indicated that the lysosome module has strong positive associations with specific host defense modules, including “response to molecule of bacterial origin”, “cell activation involved in immune response”, and “innate immune response” (Fig 5A), highlighting a conserved functional link between lysosomal activity and immune defense pathways in the bladder epithelium. Altogether, these positive correlations suggest that enhanced lysosomal function represents a conserved pathway integrated within mucosal immunity across species, rather than an isolated cellular response unique to our experimental models.

Fig 5. OM-89 enhances lysosomal activity and acidification, while disruption of acidification alters UPEC clearance.

(A) GeneBridge MMAS (Module-Module Association Score) analysis of KEGG lysosomal terms in human bladder datasets. Positively correlated modules linked to immune responses and bacterial defense are highlighted. (B) Quantification of vesicle numbers and size of LysoTracker DND-99 in infected monolayers of mouse bladder epithelial cells with and without OM-89 treatment. Values normalized to PBS controls (dashed line). Mean ± 95% CI. Vesicle numbers analyzed with Welch’s t test; vesicle size with Mann-Whitney test. N = 12 per condition. (C) Intraorganelle pH in infected monolayers with and without OM-89 treatment. Mean ± 95% CI. Welch’s t test. N = 12 per condition. (D)-(G) Colocalization of intracellular UPEC with (D) LysoTracker, (E) LAMP1, (F) Cathepsin L and (G) Cathepsin L with bafilomycin A1 (BafA1) blocking in monolayers of mouse bladder epithelial cells. For quantification: Intensities normalized to ROI area drawn around intracellular UPEC (see S4B-S4D and S4F Fig). Mann-Whitney test. N ≥ 24 per condition for LysoTracker, n ≥ 18 per condition for LAMP1, n ≥ 19 per condition for Cathepsin L, n ≥ 15 per condition for Cathepsin L + BafA1 blocking. Enrichment score (= observed overlap/random overlap) between OM-89 vs PBS for (D) 0.68 vs 0.71, (E) 0.94 vs 0.89, (F) 1.25 vs 0.91 and (G) 1.06 vs 1.17. Random overlap from S4B-S4D and S4F Fig. For images: Cross-section of representative images focused on intracellular bacteria. Intracellularity of bacteria confirmed by inspecting each position visually in XYZ with either phalloidin (D, E) or CellMask (F, G) staining. White inserts in cross-section (XYZ): intracellular UPEC with (D) LysoTracker, (E) LAMP1, (F) Cathepsin L and (G) Cathepsin L with bafilomycin A1 (BafA1) in YZ and XZ. Scale bar of inserts 5 µm. UPEC, magenta; LysoTracker/LAMP1/Cathepsin L, cyan; Phalloidin or CellMask, yellow; DAPI or Hoechst, grey. (H), (I) CFT073 signal 7h pi (antibiotic-mediated clearance) and 10h pi (regrowth) in mouse bladder organoids after blocking with (H) BafA1 or (I) CQ. Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Mann-Whitney test. N = 15 per condition. Images show representative organoids (grey, surface reconstruction) and UPEC (magenta, some marked with arrows).

To further investigate the role of lysosomal modulation in OM-89-mediated effects against UPEC infection, we examined the involvement of lysosomal activity and intracellular pH. In line with remodeling of the late endosomal and lysosomal network (Fig 4F, 4G), OM-89 treatment led to a significant increase in the number of enlarged acidic vesicles (Fig 5B), accompanied by a significant reduction in overall intracellular pH (Fig 5C). Notably, while CFT073 infection elevated intraorganellar pH in PBS-treated controls (pH of 7.5 ± 0.38) compared to uninfected controls (pH of 7.0 ± 0.24), consistently with pathogen-mediated disruption of vesicle acidification [3,36], OM-89 treatment maintained a more acidic environment in both conditions (pH of 6.58 ± 0.35 during infection and 6.45 ± 0.28 without infection; Fig 5C, S4B Fig). When focusing on intracellular UPEC, OM-89-treated cells showed a significant increase in signal intensity of acidic compartments (Fig 5D), Lamp1-positive vesicles (Fig 5E) and enzymatically active lysosomes (cathepsin L-positive vesicles; Fig 5F) in close proximity to internalized bacteria compared to PBS controls. While OM-89 expanded Lamp1- and LysoTracker-positive compartments (Fig 4G, 5B), colocalization of intracellular UPEC with these markers was largely proportional to vesicle abundance and consistent with stochastic overlap (enrichment scores ≈1 or <1) (Fig 5D, 5E and S4C and S4D Fig). However, OM-89 significantly increased the enrichment of UPEC in proximity to cathepsin L-positive vesicles (enrichment score 1.25 in OM-89 vs 0.91 in PBS) (Fig 5F, S4E Fig), indicating enhanced routing of UPEC into proteolytically competent autolysosomes. Overall, OM-89-treated bladder epithelial cells exhibited significantly elevated cathepsin L levels in the presence and absence of infection (S4F and S4G Fig), indicating that the highly abundant acidic and lysosomal compartments were functionally active. The observed lysosomal expansion, acidification (Fig 5B, 5C), and increased cathepsin L activity (Fig 5F) independently confirm the transcriptomic enrichment of lysosome-associated pathways due to OM-89 treatment.

Since lysosomal proteases such as cathepsins require an acidic environment for activation of their antimicrobial effects against intracellular bacteria [46], we next aimed to abolish acidification with pharmacological inhibition. BafA1 [47] collapsed the OM-89-driven enrichment of intracellular UPEC in proximity to cathepsin L-positive vesicles to PBS-control levels, demonstrating a clear acidification dependence (Fig 5G, S4H Fig). Overall, baf1 and chloroquine (CQ) [48] treatment reduced cathepsin activity to comparable levels in OM-89- and PBS-treated cells (S4F Fig). Interestingly, CQ treatment led to a partial increase in cathepsin activity, particularly in infected PBS-treated cells.

Finally, given the enhanced lysosomal function and increased localization of intracellular UPEC to protease active compartments during OM-89 treatment, we next asked whether this effect was essential for OM-89-mediated protection. Importantly, inhibition of acidification using either bafilomycin A1 or chloroquine did not affect initial bacterial load during infection in OM-89-treated versus PBS-treated organoids (S4I and S4J Fig). However, blocking acidification using either of the two inhibitors fully abolished OM-89’s protective effect, restoring bacterial regrowth to levels observed in PBS-treated controls (Fig 5H, 5I; 10h pi – regrowth). Interestingly, while OM-89 still promoted antibiotic-mediated clearance in bafilomycin A1-treated organoids (Fig 5H, clearance), this effect was completely lost with chloroquine treatment (Fig 5I, 7h pi – clearance).

Together, these data establish lysosomal acidification and protease activity as essential for the OM-89-dependent enhancement of intracellular bacterial clearance.

OM-89 enhances intracellular activity across antimicrobial classes and UPEC strains

To validate OM-89’s ability to reduce regrowth of CFT073 after clinically relevant antibiotic exposure, we investigated its effect following treatment with antibiotics that have different mechanisms of action than ampicillin and different intracellular accumulation levels [33,49]. We selected two commonly prescribed antibiotics for uncomplicated UTIs: Fosfomycin (FOS), a bacterial cell wall-targeting antibiotic displaying weak host cell accumulation [50] and trimethoprim/sulfamethoxazole (TMP-SMX), a folate synthesis inhibitor showing more effective intracellular accumulation [49]. OM-89 effectively reduced regrowth of CFT073 in bladder organoids (10h pi) after both FOS (Fig 6A) and TMP-SMX (Fig 6B) treatment and reduced bacterial burden (7h pi), indicating that its effect on bladder epithelial cells is independent of the antibiotic’s mode of action.

Fig 6. OM-89 reduces UPEC recurrence in bladder organoids across antibiotics and strains.

(A), (B) CFT073 signal during treatment with (A) Fosfomycin (FOS) or (B) trimethoprim-sulfamethoxazole (TMP-SMX). Quantification of bacterial fluorescence area in mouse bladder organoids 7h pi (antibiotic-mediated clearance) and 10h pi (regrowth). Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Mann-Whitney test for antibiotic-mediated clearance in (B); Welch’s t test for (A) and regrowth in (B). N ≥ 14 per condition for (A) and n ≥ 15 per condition for (B). Representative organoids shown (grey, surface reconstruction) with UPEC (magenta, some marked with arrows). (C)-(F) Regrowth of (C) UTI89, (D) J96, (E) 721536-18 and (F) 721402-18 after ampicillin treatment (10x strain-specific MIC). Quantification of bacterial fluorescence area 7h pi (antibiotic-mediated clearance) and 10h pi (regrowth). Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test for antibiotic-mediated clearance in (E), (F) and regrowth in (C); Mann-Whitney test for antibiotic-mediated clearance in (C), (D) and regrowth in (D), (E), (F). N ≥ 17 per condition for (C), n ≥ 15 per condition for (D), n ≥ 13 per condition for (E) and n ≥ 14 per condition for (F). Representative organoids shown (grey, surface reconstruction) with UPEC (magenta, some marked with arrows).

Next, we assessed the regrowth of two additional well-characterized UPEC strains of phylogroup B2, UTI89 (a cystitis isolate) [51] and J96 (a pyelonephritis isolate) [52], following ampicillin treatment. As observed with CFT073, OM-89 significantly enhanced antibiotic-mediated clearance (7h pi) of both UTI89 (Fig 6C) and J96 (Fig 6D) leading to reduced regrowth (10h pi), without affecting their initial growth within the organoids (S5A and S5B Fig).

Finally, we extended our analysis to two ampicillin-sensitive primary UPEC isolates obtained from cystitis patients, 721536–18 and 721402–18, which were identified among hyper-invasive strains in a genome-wide association study and belong to phylogroups A and B1, respectively [53]. Similar to the laboratory strains, OM-89 did not affect initial bacterial loads within organoids (S5C and S5D Fig), indicating no direct antibacterial activity. However, during antibiotic treatment, OM-89 significantly enhanced bacterial clearance (7h pi; Fig 6E, 6F) and reduced post-antibiotic regrowth of both primary isolates 721536–18 and 721402–18 (10h pi; Fig 6E, 6F), demonstrating that the positive effects of OM-89 extend beyond the classical B2 UPEC phylogroup to genetically and phylogenetically distinct clinical strains.

OM-89 enhances lysosome-associated vesicle remodeling, acidification and protease activity in human bladder epithelial cells

The bladder epithelium differs in substantial ways between mouse and human – not only in the number of cell layers and the expression of specific markers, such as cytokeratin (CK)5, CK13, CK14, CK17 and toll like receptor (TLR)11 [54], but also cell-intrinsically for example in pathogen defenses [55,56]. Such species-specific factors could influence the outcome of infection and antibiotic treatment processes.

Therefore, we asked whether the key hallmarks observed by the OM-89-driven phenotype, such as lysosome-associated vesicle modulation, acidification, protease activation and antibiotic uptake observed in mouse bladder epithelium, are conserved in human cells. Using differentiated monolayers of primary human bladder epithelial cells (S6 Fig), we quantified canonical autophagy and endo-lysosomal markers during infection and OM-89 treatment. OM-89 did not significantly change the number of p62- or LC3b-positive vesicles during UPEC infection but significantly increased their size (S7A and S7B Fig), indicating formation of enlarged autophagosomal structures. In contrast to mouse bladder epithelial cells, OM-89 did not restore autophagic flux in infected human cells (S7C Fig), despite the significant increase in LC3b vesicle size.

We next assessed late endosomal and lysosomal compartments. OM-89 treatment significantly increased both the number and size of Rab7-positive vesicles and Lamp1-positive vesicles (Fig 7A, 7B), consistent with expansion and remodeling of the late endosomal-lysosomal network. These structural changes were accompanied by functional alterations: OM-89 significantly decreased intraorganellar pH (Fig 7C) and increased cathepsin L activity (Fig 7D), consistent with lysosome-centered functional activation. The increase in protease activity was fully reversible upon inhibition of acidification with bafilomycin A1 or chloroquine (Fig 7D), demonstrating dependence on lysosomal acidification. Similar to the effects observed in mouse bladder epithelial cells, enhanced lysosomal function was already apparent in the absence of infection. Indeed, OM-89 treatment alone increased both the number and, more prominently, the size of Lamp1-positive vesicles (S7D Fig), decreased intraorganellar pH (S7E Fig), and increased cathepsin L activity (S7F Fig). Finally, OM-89 markedly enhanced intracellular uptake of fluorescently labeled ampicillin in uninfected human bladder epithelial cells (Fig 7E), mirroring the uptake phenotype observed in mouse cells.

Fig 7. OM-89 enhances lysosomal activity, intracellular acidification, antibiotic uptake and bacterial clearance in human bladder epithelial systems.

(A), (B) Quantification of vesicle numbers and size for (A) RAB7 and (B) LAMP1 in infected monolayers of human bladder epithelial cells with and without OM-89 treatment. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test for RAB7 (A); Mann-Whitney test for LAMP1 (B). N = 12 per condition for RAB7 and ≥ 11 per condition for LAMP1. Z-projection (maximum intensity) of representative images. Foci, cyan; DAPI, grey. (C) Intraorganelle pH in infected human monolayers with and without OM-89 treatment. Mean ± 95% CI. Welch’s t test. N = 12 per condition. (D) Cathepsin L activity in human monolayers during infection with or without OM-89, and with BafA1 or CQ blocking. Values normalized to unblocked PBS controls (dashed line). Mean ± 95% CI. Brown-Forsythe ANOVA with Welch’s correction. N = 11 per condition. (E) Uptake of FITC-labelled ampicillin into monolayers of human bladder epithelial cells. Quantification of fluorescence after background subtraction of PBS- or OM-89-treated cells without labelled antibiotic. Each dot represents one field of view. Mean ± 95% CI. Mann-Whitney test. N = 16 per condition. Z-projection (maximum intensity) of representative images. FITC-ampicillin shown in cyan. (F) Representative H&E (Hematoxylin & Eosin) staining of human bladder organoids derived from primary H-6215 cells, showing stratified epithelial architecture with 4-5 layers and thin, elongated cells lining the central lumen. Frequently binucleated cells are indicated by arrows. Scale bar, 50 µm. Inserts (I and II) highlight luminal cell morphology; black lines outline elongated umbrella-like cells. Scale bar, 10 µm. (G), (H) Human bladder organoids express markers specific to umbrella cells, cytokeratin (CK)20 (G), and important for UPEC infection, UP3A (H), towards the lumen. Binucleated cells are marked by asterisks. Representative images shown (partially same images as in S9A Fig). (I) CFT073 signal 7h pi (antibiotic-mediated clearance) and 10h pi (regrowth) in human bladder organoids. Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Mann-Whitney test for antibiotic-mediated clearance and Welch’s t-test for regrowth. N ≥ 14 per condition. Images show representative organoids (grey, surface reconstruction) and UPEC (magenta, some marked with arrows).

Together, these data show that OM-89 drives conserved lysosome-centered activation and enhanced antibiotic uptake in human bladder epithelium, while revealing species-specific differences in autophagic flux regulation and lysosomal network architecture.

OM-89 enhances antibiotic-mediated clearance and limits UPEC regrowth in human bladder organoids

Finally, we asked whether OM-89 exerts similar effects on bacterial clearance in human bladder epithelium. We therefore established human bladder organoids (hBOs) from primary bladder epithelial cells which already showed robust stratification and differentiation in another 3D microtissue model [57]. Cells were cultured in 3D under conditions adapted from the mouse bladder organoid model [31] and other bladder epithelial models [58]. Briefly, following an initial expansion phase of 9 days, differentiation was induced for 4–5 days to generate stratified organoids that recapitulated key aspects of human urothelium (S8A and S8B Fig). By day 13–14, hBOs displayed 3–4 epithelial layers surrounding a central lumen (Fig 7F, S9A Fig). Luminal cells exhibited hallmark features of umbrella cells, including thin morphology, frequent binucleation, and expression of CK20+ and UP3A+ cells [59,60] (Fig 7G, 7H, S9A Fig), while CK13 + intermediate and CK5 + basal-like cells [61] populated deeper layers (S9A Fig). RT-qPCR confirmed upregulation of umbrella (KRT20, UPK3A) and intermediate/progenitor (KRT13, TP63 [62]) markers (S9B Fig). The organoids formed a functional epithelial barrier, as evidenced by ZO-1 expression [63] and restricted dextran diffusion in both freshly cultured and cryopreserved-then-differentiated (thawed) organoids (S9C and S9D Fig).

To test OM-89’s effect on human cells, hBOs were treated under the continuous application regimen for 72 hours prior to microinjection of UPEC strain CFT073, with OM-89 maintained throughout the experiment. As in mouse organoids, OM-89 did not alter bacterial burden during the initial infection phase (4h pi; S10A Fig). However, although OM-89 did not significantly reduce bacterial burden in hBOs (7h pi; 10x MIC ampicillin), it significantly reduced bacterial regrowth following antibiotic treatment (10h pi) (Fig 7I). Consistent with our observations in mBOs, UPEC regrowth in hBOs preferentially originated from the organoid wall rather than from luminal compartments (S10B Fig, S3 and S4 Video), indicating that the epithelial layer serves as the primary site of bacterial regrowth following antibiotic withdrawal.

Together, these findings demonstrate that OM-89 restricts intracellular UPEC regrowth in a human organoid model, suggesting conservation of its positive effects across host species.

Discussion

Intracellular reservoirs of UPEC within bladder epithelial cells are a major driver for the reactivation of UTIs from the bladder epithelium due to their capacity to evade both antibiotics and immune responses [8]. Here, we identify a host-directed drug-based approach that strengthens epithelial antimicrobial mechanisms and enhances intracellular bacterial clearance.

Using the bacterial lysate OM-89 – a clinically approved therapy for recurrent UTIs with decades of clinical use – as a mechanistic probe of epithelial responses, we show that bladder epithelial cells can be modulated to accelerate intracellular bacterial clearance through coordinated enhancement of lysosomal function and enhanced intracellular accessibility of antibiotics (Fig 8). Transcriptomic profiling further revealed that OM-89 induces a transcriptional program partially overlapping with infection-associated responses, yet remaining clearly distinct from the infected state, suggesting that OM-89 establishes a modified epithelial defense state rather than mimicking infection-driven inflammation. Together, these findings provide a mechanistic framework for the long-observed clinical efficacy of OM-89 and identify epithelial lysosomal pathways as a therapeutically targetable component of host defense that can be used to improve intracellular bacterial clearance.

Fig 8. Working model hypothesis: OM-89 reinforces a lysosome-centered antimicrobial defense against intracellular UPEC in the bladder epithelium.

Schematic summary of the OM-89-induced antimicrobial response. OM-89 triggers coordinated activation of the endo-lysosomal and autophagic pathways in bladder epithelial cells. First, intracellular UPEC, internalized into early vacuoles or routed through p62/LC3B-II-positive autophagosomes, are trafficked through Rab7 ⁺ late endosomes toward an expanded lysosomal compartment (LAMP1⁺). Second, autophagic flux restoration is observed in mouse but not in human epithelial cells, but with enlarged lysosomal vesicles still observed in human cells. Third, OM-89 treatment increases lysosomal acidification and cathepsin L protease activity, enriching intracellular UPEC in proximity to proteolytically active autolysosomes. Fourth, in parallel, and independently of lysosomal routing, OM-89 increases intracellular accumulation of multiple antibiotic classes by an unknown mechanism. Together, these coordinated epithelial responses result in increased intracellular bacterial clearance and reduced post-antibiotic UPEC regrowth. Steps triggered by OM-89 are highlighted in orange. UPEC, magenta. Bladder epithelial cell, grey. Antibiotics, cyan. Created in BioRender. Tomasek, K. (2026) https://BioRender.com/wa2tl4q.

OM-89 induced coordinated lysosome-centered functional remodeling in both mouse and human bladder epithelial cells. In mouse cells, the observed increased Rab11A and Rab27B vesicles are consistent with enhanced endosomal turnover and secretion pathways implicated in expulsion of intracellular pathogens [36]. Despite accumulation of p62, likely reflecting infection-associated cellular stress [41], enlarged autophagosomes together with increased autophagic flux indicate that the ubiquitinated cargo is efficiently routed toward lysosomal degradation in mouse epithelial cells, overcoming typical UPEC-mediated blockade of autophagic maturation [42]. Increased numbers of Rab7-positive late endosomes and Lamp1-positive vesicles further support expanded lysosomal biogenesis and remodeling. While OM-89-induced lysosomal remodeling, autophagic vesicle enlargement and acidification were also preserved in human bladder epithelial cells, the effect on canonical autophagic flux appears to diverge, with restoration of flux in mouse but not in human cells. One underlying reason could be the difference in basal levels of autophagy between the two species, as at least mouse models seem to have generally a higher basal level of autophagy than human lung epithelial cells [64]. Such species-specific variability likely reflects differences in regulatory wiring upstream of autophagy, leading to differences in the tuneability of these networks by OM-89, but suggest that the core lysosome-directed antimicrobial mechanism appears to be preserved even when certain pathway features differ between species.

Strikingly, OM-89 did not only enlarge the lysosomal network but enhanced its functional capacity in both species. Lysosomal proteases became concentrated in proximity to intracellular pathogens, suggesting targeted trafficking of UPEC into proteolytically active autolysosomes. Pharmacological inhibition of lysosomal acidification abolished OM-89-mediated protection, demonstrating a direct link between lysosomal function and reduced UPEC regrowth. These coordinated changes likely undermine UPEC’s strategies to evade degradation [3,4], revealing a yet underappreciated mechanism by which OM-89 fine-tunes epithelial antimicrobial defenses through lysosomal functional enhancement.

Although this enhanced host activity did not affect UPEC replication during primary infection, its significance emerged during antibiotic treatment as observed already in a mouse study using repeated UPEC infections [25]. Using bladder organoid and monolayers models, we showed that OM-89 increased intracellular uptake and/or retention of structurally distinct, poorly permeant antibiotics, as well as dextran, indicating a broad effect on intracellular delivery pathways. Because antibiotics used to treat UTIs display varied intracellular activity profiles [33,49,50], improving intracellular access may significantly expand their therapeutic efficacy. Supporting this, OM-89 reduced bacterial regrowth even at low antibiotic concentrations, indicating that host-directed boosting of intracellular pharmacokinetics can enhance antibiotic performance.

Interestingly, CFU quantification in differentiated bladder epithelial cells revealed that OM-89-mediated reduction of intracellular bacteria was already apparent following the gentamicin protection phase, before prolonged ampicillin exposure. This observation suggests that the protective effect of OM-89 is initiated early after bacterial invasion and cannot be explained solely by enhanced intracellular antibiotic accumulation. Instead, increased antibiotic uptake likely acts together with epithelial antimicrobial mechanisms, including enhanced lysosomal function, to restrict intracellular bacterial survival. Notably, several hallmarks of enhanced lysosomal function, including increased Lamp1-positive vesicles, enhanced acidification and elevated cathepsin L activity, were already evident in the absence of infection. Together with the transcriptomic changes induced by OM-89 alone, these observations suggest that OM-89 establishes a generalized epithelial defense state rather than merely augmenting infection-induced responses. The absence of differences in total bacterial burden during the growth phase in organoids does not necessarily contradict these findings, as our fluorescence-based quantification captures the overall intra-organoid bacterial population without distinguishing luminal from intracellular bacteria. Thus, early changes affecting the intracellular niche may remain masked by the larger luminal bacterial population. Whether additional epithelial mechanisms contribute to this early restriction phase remains an interesting question. OM-89 has previously been shown to activate innate immune pathways and enhance antimicrobial responses [21,36,65–67], suggesting that multiple epithelial defense pathways may cooperate to limit intracellular persistence. OM-89’s protective effect extended across multiple antibiotics and genetically diverse UPEC strains – from classical UTI-associated B2 phylogroup isolates to phylogroups A and B1, more commonly linked to gut strains [68] – indicating robustness against heterogeneous intracellular persistence strategies. Importantly, these epithelial effects were not restricted to murine systems. Human bladder organoids recapitulated the functional phenotype, showing reduced intracellular bacterial burden during antibiotic treatment and reduced regrowth after antibiotic treatment. Together with the conserved core lysosome-centered antimicrobial response, our findings support the translational relevance of OM-89 as a clinically used therapy whose mechanism of action involves direct modulation of epithelial antimicrobial pathways.

One limitation of our study is that OM-89 was applied directly to epithelial cultures and organoids, whereas in clinical use it is administered orally. Although pharmacokinetic studies have demonstrated systemic distribution and urinary accumulation of OM-89-derived components following oral administration [27], our experimental setup does not recapitulate the exact route, kinetics or concentration profiles encountered in vivo. Rather, our models were designed to determine whether bladder epithelial cells are capable of responding directly to OM-89-mediated signals and to identify the intracellular pathways involved. Given the documented systemic exposure following oral administration, direct effects on the urothelium are biologically plausible. However, future studies will be required to determine how the epithelial responses identified here integrate with the complex systemic and immune-mediated effects of OM-89 under physiological administration conditions.

Although a clear effect of OM-89 on spontaneous recurrences and bacteriuria was demonstrated in mice previously, the study did not find depletion of UPEC from the bladder reservoir [25]. Our findings suggest that enhanced endo-lysosomal trafficking of intracellular UPEC together with increased acidification may restrict bacterial reactivation and therefore luminal shedding. Such modulation of epithelial trafficking of intracellular pathogens could therefore mechanistically contribute to reduced recurrences of UTIs by limiting the epithelial source of reseeding events. Such host-targeted strategies at the site of infection, the bladder epithelium, could enhance the effectiveness of existing antibiotics at the cellular level and help limit intracellular pathogen survival, shorten antibiotic treatment courses, and reduce overall antibiotic exposure. In the face of rising antimicrobial resistance [69], our findings provide a mechanistic rationale for the clinical use of OM-89 and support epithelial lysosomal pathways as a promising target for host-directed therapeutic strategies that enhance intracellular bacterial clearance and improve the efficacy of existing antibiotics.

Methods

Bacteria

The well-characterized UPEC strains CFT073 [32], UTI89 [51] and J96 [52], and primary clinical isolates 721536–18 and 721402–18 [53] were used. Strains were grown overnight in DMEM/F-12 supplemented with 20 mM HEPES without (for organoid injections) or with (for monolayer infections) 10% heat-inactivated Fetal Bovine Serum (FBS) at 37°C statically.

Chromosomal sfGFP integration.

To enable real-time monitoring of the infection dynamics, all UPEC strains were genetically engineered to stably express sfGFP under control of a strong P_σ70 promoter at the chromosomal att_HK022 site using scarless cloning [70,71]. Briefly, the kan^R-P_RhaB-relE toxin cassette was amplified from plasmid pSLC217 using primers P4 and P5 and integrated into the genome of the strains using λ Red recombineering [72,73]. Primers P6 and P7 were then used to amplify the sfGFP-containing fragment from plasmid pSLC293, and these PCR amplicons were used to replace the kan^R-P_RhaB-relE toxin cassette by λ Red recombineering with negative selection on M9 agar supplemented with 2 mM MgSO₄, 0.1 mM CaCl₂ and 0.2% L-rhamnose. Successful recombination at each step was confirmed by PCR using primers attHK_F and attHK_R.

Used strains, cells, reagents and equipment are listed in Table 2. Primers are listed in Table 3.

Table 2. Reagents and tools table.

Reagent/Resource	Reference or Source	Identifier or Catalog Number
Experimental Models
CFT073	[32]	N/A
UTI89	[51]	N/A
J96	[52]	N/A
721536-18	[53]	N/A
721402-18	[53]	N/A
CFT073 ΔfliC::FRT	[71]	TS4
CFT073 attHK022:: σ70-sfGFP	[71]	TS85
TS4 attHK022:: σ70-sfGFP	[71]	TS66
J96 attHK022:: σ70-sfGFP	Valentin Borgeat	FN27
UTI89 attHK022::σ70-sfGFP	Valentin Borgeat	FN28
721536-18 attHK022::σ70-sfGFP	This study	VB36
721402-18 attHK022::σ70-sfGFP	This study	VB37
H-6215 primary bladder epithelial cells	CellBiologics	Cat#H-6215
C57BL/6 mice	Charles River Laboratories	Cat#027
ROSAmT/mG mice	Jackson Laboratories	Cat#007576
Recombinant DNA
pSIM5	[73]	N/A
pSLC293	[70]	N/A
pSLC217	[72]	N/A
Antibodies
Antibodies	This study	see Table 4
Oligonucleotides and other sequence-based reagents
PCR primers	This study	see Table 3
Chemicals, Enzymes and other reagents
CellTracker Red CMTPX	Thermo Fisher	Cat#C34552
DAPI solution	Becton Dickinson	Cat#564907
Trypan Blue Stain	Thermo Fisher	Cat#T10282
Phalloidin-555	Thermo Fisher	Cat#A34055
Phalloidin-750	Thermo Fisher	Cat#A30105
Dextran Tetramethylrhodamine (Dextran-TMR)	Thermo Fisher	Cat#D3308
HEPES	Thermo Fisher	Cat#15630056
Gelatin-based coating solution	CellBiologics	Cat#CB6950
TryPLE	Thermo Fisher	Cat#12605010
M9 salts	Sigma-Aldrich Chemie	Cat#M6030-1 KG
Magnesium sulfate	Sigma-Aldrich Chemie	Cat#63136-250G-F
Calcium chloride	Sigma-Aldrich Chemie	Cat#C4901-500G
L-rhamnose	Sigma	Cat#R3875-25G
LB broth	Fisher Scientific	Cat#11778902
Primocin	LabForce AG	Cat#ant-pm-1
Cell recovery solution	Chemie Brunschwig	Cat#FAL354253
Bovine serum albumin (BSA)	Sigma-Aldrich Chemie GmbH	Cat#12659–500GM-M
DMSO	Chemie Brunschwig AG	Cat#67-68-5
OM-89	OM Pharma	OM-89C neutralized
Phosphate buffered saline (PBS)	Thermo Fisher	Cat#10010015
Fluorescein isothiocyanate (FITC)-tagged dextran 4000 Da	Sigma-Aldrich	Cat#46944–100MG-F
Chloroquine	LabForce AG	Cat#tlrl-chq-4
Bafilomycin A1	LuBioScience GmbH	Cat#HY-100558–500UG
16% Formaldehyde	Life Technologies	Cat#28908
Histogel	Fisher Scientific	Cat#12006679
NuPAGE LDS sample buffer	Thermo Fisher	Cat#NP0007
Dithiothreitol (DTT)	Thermo Fisher	Cat#165680050
Bis-Tris gels	Life Technologies	Cat#NW04122BOX
SuperSignal West Atto substrate solution	Life Technologies	Cat#46641
SuperSignal western blot enhancer	Life Technologies	Cat#A38554
SuperSignal West Pico PLUS chemiluminescent substrate	VWR International	Cat#PIER34577
LysoTracker DND-99 Red	Life Technologies	Cat#L7528
NucBlue Live ReadyProbes Reagent (Hoechst 33342)	Life Technologies	Cat#R37605
CellMask Plasma Membrane Stain Deep Red	Life Technologies	Cat#C10046
LysoSensor Yellow/Blue DND-160	Life Technologies	Cat#L7545
SYBRGreen PCR Master Mix	Life Technologies	Cat#4364344
Complete Epithelial Cell Medium/w Kit – 500 ML	Cell Biologics	Cat#H6621
Fetal Bovine Serum (FBS)	Thermo Fisher	Cat#A5256701
Cultrex Reduced Growth Factor Basement Membrane Extract, Type 2, Pathclear	R&D Systems	Cat#3533-010-02
3dGRO L-WRN Conditioned Media Supplement	Sigma-Aldrich	Cat#SCM105
DMEM/F-12	Thermo Fisher	Cat#11320033
DMEM/F-12, no phenol red	Thermo Fisher	Cat#21041025
GlutaMAX supplement	Thermo Fisher	Cat#35050061
Advanced DMEM/F-12	Thermo Fisher	Cat#12634010
B-27 Supplement (50X), serum free	Thermo Fisher	Cat#17504044
Human FGF-7 (KGF), Recombinant Protein, PeproTech	Thermo Fisher	Cat#100-19-100UG
Human FGF-10 Recombinant Protein, PeproTech	Thermo Fisher	Cat#100-26-100UG
A 83–01	Stemcell Technologies	Cat#100–0245
Y-27632 dihydrochloride	Bio-Techne	Cat#1254/50
N-Acetyl-L-cysteine	Sigma-Aldrich	Cat#A9165-25G
Nicotinamide	Sigma-Aldrich	Cat#N0636-500G
PeproGMP Human VEGF-165 Recombinant Protein, PeproTech	Thermo Fisher	Cat#GMP100-20-50UG
Ampicillin	Sigma-Aldrich Chemie	Cat#A9518-5G
Gentamicin	Sigma-Aldrich Chemie	Cat#G3632-10G
Fosfomycin	Sigma-Aldrich Chemie	Cat#P5396-5G
Trimethoprim	Sigma-Aldrich Chemie	Cat#T7883-5G
Sulfamethoxazole	Sigma-Aldrich Chemie	Cat#S7507-100G
Kanamycin	Sigma-Aldrich Chemie	Cat#K4000-25G
Software
STAR	[75]	v.2.7.10b
DESeq2		version 1.36.0
R studio	Posit Software
g:Profiler R client	[76]
Fgsea toolset	[77]
Genebridge MMAS (Module-Module Association Score)	[45]
GraphPad Prism		10.3.1
Fiji		2.18.0
Other
RNAeasy Plus Mini Kit	QIAGEN	Cat#74134
Pierce FITC Antibody Labeling Kit	Life Technologies	Cat#53027
Magic Red Cathepsin L kit	Bio-rad	Cat#ICT942
SuperScriptIV First-Strand Synthesis System with random hexamers	Life Technologies	Cat#18090050
Tecan Infinite M Plex	Tecan	Cat#25025547
Leica Thunder DMi8 microscope	Leica	Cat#25013617
Confocal Stellaris 5	Leica	Cat#25023819
iBlot Gel Transfer Device	Life Technologies	Cat#IB301002
QuantStudio 7 Flex Real-Time PCR System	Thermo Fisher	Cat#25008528
TapeStation 4200	Agilent	Cat#20082823
NovaSeq 6000	Illumina	Cat#25013498

Table 3. Used primers.

Primer name	Nucleotide sequence (5’ to 3’)	Source
attHK_F	GCTGATAAAGCGCAGGTTG	this study
attHK_R	GCGCAACAGGTTATCAGC	this study
fliC_KO_F	TAATCAACGACTTGCAATATAGGATAACGAATCATGGCACAATGTG TAGGCTGGAGCTGCTTCG	Thomas Simonet
fliC_KO_R	TGCCGTCAGTCTCAGTTAATCAGGTTACGGCGATTAACCCTGCATA TGAATATCCTCCTTAG	Thomas Simonet
fliC_F	CAGACGATAACAGGGTTGACGGC	Thomas Simonet
fliC_R	TTGCAATTCCCCTTGTAGGCCTG	Thomas Simonet
P4	GCGCCATCAGGGCAAAGCCCATCCAGAGTCTTCGGGTCAGGGTTA AATTCACGGTCGGTGCGTGTAGGCTGGAGCTGCTTC	Eshaghi et al. 2016
P5	ATAATAAGGCTTTATGCTAGATGCATTCTGCTTTGCGACTCAACCTT TTTCACCTAAAGTCATATGAATATCCTCCTTAG	Eshaghi et al. 2016
P6	GCGCCATCAGGGCAAAGCCCATCCAGAGTCTTCGGGTCAGGGTTA AATTCACGGTCGGTGCTTGACGGCTAGCTCAGTCCTA	Eshaghi et al. 2016
P7	ATAATAAGGCTTTATGCTAGATGCATTCTGCTTTGCGACTCAACCTT TTTCACCTAAAGTTTAATGATGATGATGATGATG	Eshaghi et al. 2016
UPK3A_fwd	CTCACAGATCCTGAATGCCTACC	Origene
UPK3A_rev	CCGTGGACATATTGACCAGGAC	Origene
KRT20_fwd	CTGAGGTTCAACTAACGGAGCTG	Origene
KRT20_rev	AACAGCGACTGGAGGTTGGCTA	Origene
KRT13_fwd	GATGCTGAGGAATGGTTCCACG	Origene
KRT13_rev	AGCTCCGTGATCTCTGTCTTGC	Origene
TP63_fwd	CAGGAAGACAGAGTGTGCTGGT	Origene
TP63_rev	AATTGGACGGCGGTTCATCCCT	Origene
ZO-1_fwd	GTCCAGAATCTCGGAAAAGTGCC	Origene
ZO-1_rev	CTTTCAGCGCACCATACCAACC	Origene
Gapdh_fwd	GTCTCCTCTGACTTCAACAGCG	Origene
Gapdh_rev	ACCACCCTGTTGCTGTAGCCAA	Origene

Ethical statement

Mouse primary bladder epithelial cells were obtained from C57BL/6 mice aged 9 weeks or female ROSA^mT/mG mice at age 3–4 months. Mice were housed in a specific pathogen-free facility. All animal protocols were reviewed and approved by EPFL’s Chief Veterinarian, by the Service de la Consommation et des Affaires Vétérinaires of the Canton of Vaud, and by the Swiss Office Vétérinaire Fédéral. ROSA^mT/mG mice were used under the laboratory’s approved license number VD3434c. C57BL/6 mice were not sacrificed specifically for this study; bladders were obtained as surplus tissue from mice euthanized under EPFL’s Center of PhenoGenomics organ sharing license number VD3290x2. All mice were euthanized by CO₂ inhalation, with no prior anesthesia or analgesia administered. Handling of the animals was minimized in accordance with Swiss cantonal (Canton of Vaud) animal welfare regulations to alleviate stress and suffering.

Cell culture

Human primary bladder epithelial cells (H-6215) were cultured in H-6621 medium supplemented to a final concentration of 10% heat-inactivated FBS. Cells were received from the supplier at passage 3 and expanded according to the manufacturer’s recommendations before use at passages 6–8 (maximum passage as recommended by the supplier).

Routinely, mouse and human cells were grown at 37°C and 5% CO₂ without antimicrobial or antimycotic agents. Organoids were cultured in media without FBS but with 2.5 mM GlutaMAX supplementation, whereas for monolayers a total of 10% heat-inactivated FBS was added and plastics were coated with gelatin-based coating solution. The day prior to experiments, either medium was replaced with DMEM/F12 (without phenol red) supplemented with 20 mM of HEPES.

Mouse bladder organoids.

Mouse bladder organoids (mBOs) were generated as previously described [31] with minor modifications.

Briefly, excised mouse bladders were minced with scissors and incubated in 1-1.5 ml pre-warmed TryPLE at 37°C with shaking at 170 rpm for 30 min. Following vortexing for 30 s, TryPLE was neutralized by addition of DMEM/F-12 supplemented with 10% FBS. Single cells were embedded in reduced growth factor basement membrane extract (BME) and seeded as 50 µl domes in 24-well plates in an inverted configuration for 30 min at 37°C to promote 3D growth. Thereafter, organoids were fed with mouse bladder organoid (MBO) medium (Advanced DMEM/F-12 medium supplemented with 2.5 mM GlutaMAX, 100 ng/ml FGF10, 100 ng/ml FGF7, 500 nM A83-01, 2% B27, 10 µM Y-27632 ROCK inhibitor and 100 μg/ml primocin. Typically, roughly 50,000 bladder cells were seeded per dome as not all of them have the necessary stem-cell like characters to promote growth of organoids. MBO medium was renewed every 2–3 days and upon first splitting or passaging, primocin was omitted.

When organoids became overly dense, they were split into multiple BME domes following extraction using 250 µl ice-cold Cell Recovery Solution per dome and incubating for 30 min on ice. After centrifugation at 300 g for 5 min, the supernatant was carefully removed and the organoids in remaining BME were washed once in ice-cold PBS before spinning again. Grown organoids were passaged every 5–7 days by extraction from BME with Cell Recovery Solution, followed by TryPLE digestion for 10 min at 37°C to generate single cells. At this stage, roughly 5,000–10,000 cells were seeded per dome. For cryopreservation, culture medium was removed and organoids embedded in BME were directly submerged in freezing medium (60% FBS, 30% HBO medium, 10% DMSO) prior to freezing.

tdTomato-expressing organoids derived from ROSA^mT/mG mice were used for microinjection experiments, whereas wild-type (WT) organoids dissociated into single cells were primarily used to grow as monolayers.

Generally, all plastics, except well plates, were pre-coated with 1% bovine serum albumin (BSA) in PBS to avoid organoid loss due to increased adhesion of organoids to uncoated plastic.

Human bladder organoids.

Human bladder organoids (hBOs) were generated from H-6215 cells expanded as described above. Single cells at passages 6–8 were seeded at a density of ≤ 5,000 cells per 50 µl BME dome in 24-well plates. As for mouse bladder organoids, laboratory plastic was pre-coated with 1% BSA in PBS and cells were seeded in an inverted configuration for 30 min at 37°C to promote 3D growth.

Organoids were cultured for 9 days in a 1:1 mixture of HBO medium (MBO medium as described above, supplemented with 1 mM N-acetyl L-cysteine (NAC), 10 mM Nicotinamide (NIC) and 50 ng/ml EGF) and LWRN-conditioned medium [58], referred to as 50/50 medium. Medium was refreshed every 2 days. On day 9, organoids were either differentiated or cryopreserved.

For cryopreservation, culture medium was removed and organoids embedded in BME were directly submerged in freezing medium (60% FBS, 30% HBO medium, 10% DMSO) prior to freezing. Following thawing, organoids were re-embedded in 50 µl BME domes and allowed to recover overnight in 50/50 medium before differentiation.

Differentiation was induced by culturing organoids for 4–5 days in 95% HBO medium supplemented with 5% LWRN-conditioned medium (95/5 medium), with medium changes every 2 days. Differentiated hBOs were used for characterization or microinjection experiments on days 13–14.

Mouse and human bladder epithelial cell monolayers.

Single cells of mouse bladder epithelial cells were obtained by dissociating WT mBOs using TryPLE digestion for 10 min at 37°C. Single-cell suspensions of mouse bladder epithelial cells or human bladder epithelial cells (H-6215, passages 6–8) were seeded into gelatin-coated µ-Plate 96-well dishes (ibidi), unless indicated other, at densities of 1.5-2*10⁴ cells per well and 1-1.5*10⁴ cells per well, respectively. Cells were cultured in 200 µl MBO medium or H-6621 medium supplemented with 10% FBS, respectively. Upon reaching approximately 40–50% confluency (typically the following day), the respective culture media were replaced with DMEM/F-12 supplemented with 10% FBS and cells were maintained under these conditions for 3 days to promote epithelial differentiation through growth factor withdrawal. Differentiated monolayers were used for experiments upon reaching 100% confluency. Umbrella-like cells expressed cytokeratin (CK)20 and uroplakin (UP)3a, intermediate cells expressed CK13, basal cells expressed CK5 [31,61]. Umbrella-like cells were frequently binucleated.

OM-89 treatment

Organoids and monolayers were treated with OM-89 (OM-Pharma) at 500 µg/ml of protein in either MBO, 95/5 HBO/LWRN media or DMEM/F-12 media supplemented with 2.5 mM GlutaMAX and 20 mM HEPES. For control groups, equal volumes of phosphate buffered saline (PBS) was used instead of OM-89. OM-89 and PBS were added to the media according to the different treatment regimens: continuous application – 72 hour treatment pre-infection and treatment throughout the experiment, pre-application – 72 hour treatment pre-infection only with a 24 hour resting period where treatment was omitted, co-application – treatment only during antibiotic treatment. For the continuous application, bacteria were microinjected after adjusting their optical density at 600 nm (OD₆₀₀) in media with either PBS or OM-89. If not stated other, bladder epithelial cells were subjected to the continuous application.

Minimal inhibitory concentration (MIC) measurements

Bacteria were grown in DMEM/F-12 media (without phenol red) supplemented with 2.5 mM GlutaMAX and 20 mM HEPES. Overnight cultures were diluted to an OD₆₀₀ of 0.002 and grown till OD₆₀₀ of 0.01. Bacterial cultures were adjusted to an OD₆₀₀ of 0.0005 for serial dilutions of tested antibiotics. MIC was estimated based on growth after 24 hours in a plate reader (Tecan Infinite M Plex) and using the Gompertzequation for MIC determination [74] (Table 5). If not stated other, 10x the strain MIC of the indicated antibiotic was used for infection experiments.

Table 5. MICs.

Strain	Antibiotic	MIC [µg/ml]
CFT073	Ampicillin	6.43
CFT073	Fosfomycin	6.97
CFT073	Trimethoprim/Sulfamethoxazole	1.76
UTI89	Ampicillin	5.35
J96	Ampicillin	9.98
721536-18	Ampicillin	4.98
721402-18	Ampicillin	11.61

To determine differences in MIC with OM-89 treatment, OM-89 at 500 µg/ml of protein or equal volumes of PBS was added upon dilution of overnight cultures (Table 1).

Intracellular bacterial burden assay

Intracellular bacterial burden was quantified by colony-forming unit (CFU) enumeration in differentiated mouse bladder epithelial monolayers. Mouse bladder epithelial cells were seeded in gelatin-coated 48-well plates in MBO medium. Cells were differentiated and pre-treated with OM-89 as described above. Differentiated monolayers were infected with UPEC strain TS66 at approximately 1*10⁶ CFU/ml (corresponding to roughly 2*10⁵ CFU per well in 200 µl) for 4 hours on a rocking platform to facilitate bacterial uptake. Following infection, cells were washed four times with excess PBS and incubated in DMEM/F12 supplemented with 10% FBS and either OM-89 or PBS containing 100 µg/ml gentamicin for 1 hour to eliminate extracellular bacteria. Subsequently, cells were washed four times with excess PBS and either lysed immediately for CFU determination or further incubated for an additional 3 hours in medium containing 10x the MIC of ampicillin. Cells were then washed and lysed with 120 µl of 0.1% Triton X-100 in PBS for 10 min. Cell lysates were collected by scraping the wells and pipetting five times to ensure complete homogenization. Aliquots were plated on LB agar plates and incubated overnight at 37°C. Colonies were counted the following day and CFU per milliliter were calculated.

Injection experiments

Organoid injections were performed as described [31] with the modifications outlined below.

Organoid preparation.

Briefly, the day prior to the experiments, organoids were extracted from BME domes using cell recovery solution and transferred to µ-Slide 2 Well dishes (ibidi) to allow direct comparison of OM-89-treated and PBS-treated conditions. Organoids were embedded in 75 µl BME spread over an area of approximately 1.5 cm² to align organoids in a narrow focal plane rather than a dome configuration. At this stage, the media was replaced by phenol red-free DMEM/F-12 medium with 2.5 mM GlutaMAX and 20 mM HEPES with either OM-89 or PBS.

On the day of the microinjections, hBOs were stained inside the BME with CellTracker Red CMTPX at 5 µM for 1 hours, whereas mBOs derived from ROSA^mT/mG mice were left unstained due to their inherent tdTomato-expression. CellTracker was washed off with excessive volumes of warm DMEM/F-12, before putting again phenol red-free DMEM/F-12 medium with 2.5 mM GlutaMAX and 20 mM HEPES with either OM-89 or PBS.

Lid preparation for media changes.

To enable media exchange during live imaging, a custom-made perfusion lid compatible with µ-Slide 2 Well dishes was fabricated the day before microinjections with one inlet and one outlet per well. Using a drill press, two small holes were drilled at the outer corners of a DIC lids for µ-Slides (ibidi) in each well to fit cannula tips with conical, tapered ends (1.30/0.75 x 15 mm, Unimed). Cannula tips were connected to roughly 20 cm of silicone tubing (ELASTOSIL R plus 4305/60, Freudenberg). Each cannula tip was secured in place with glue with minimal distance to the bottom of the dish to allow easy media change without leaving a dead volume. The tubing was split mid-length with additional cannula connectors to allow easy routing into an H201-K-Frame incubation chamber (Okolab). At the distal end, the silicone tubing was connected to dispensing tips (LL 1/2“, ID 0.69 mm, Gonado) attached to 10 ml syringes for manual media exchange.

For every media exchange (e.g., antibiotic treatment or withdrawal), each well was washed with 9 ml of the respective media before leaving 1 ml in the well. All lid components were disassembled and cleaned with ethanol between experiments, whereas silicone tubing was replaced after each use.

Bacterial injection.

UPEC strains with stable chromosomal integration of sfGFP were grown overnight statically at 37°C in DMEM/F-12 (phenol red-free) media with 2.5 mM GlutaMAX and 20 mM HEPES.

In the morning of the experiments, the media of the organoids was replaced with fresh phenol red-free DMEM/F-12 supplemented with 2.5 mM GlutaMAX and 20 mM HEPES with either OM-89 or PBS. The OD₆₀₀ of bacterial cultures was adjusted to 2.25 in a 1:1 dilution of media and Phenol Red solution to facilitate the visualization of injected organoids. Microinjections were performed using a Pneumatic PicoPump (WPI) using micropipettes prepared from thin wall glass capillary (TW100F-4 with length 100 mm and diameter 1 mm) using a Flaming/Brown Micropipette Puller (Sutter Instruments model P-87) set at pressure 360, heat 866, Vel 200. The micropipettes were then cut with tweezers under a stereomicroscope (Olympus SZX-16) to generate a tip size ejecting roughly 1 nl volume of the adjusted bacterial culture in mineral oil on a Zeiss coverslide (corresponding to 100 mm), corresponding to 1926 ± 150 CFU per injection (quantified from two needles, plating three replicates on LB agar plates). Injected organoids were allowed to rest for 1 hour prior to time-lapse microscopy imaging onset after assembling the µ-Slide 2 Well dishes with the custom-made perfusion lid.

Dextran injection of hBOs.

hBOs were microinjected with fluorescein isothiocyanate (FITC)-tagged dextran (4000 Da) at a final concentration of 1 mg/ml after diluting dextran dissolved in DMEM/F-12 supplemented with 2.5 mM GlutaMAX and 20 mM HEPES 1:1 with Phenol Red solution to visualize injection. Dextran diffusion out of the organoids was monitored for 2 hours post-injection. Organoids were imaged as described in the time-lapse microscopy section below. The luminal FITC fluorescence signal at the central plane of each organoid was extracted with Fiji after thresholding on the organoid boundaries using the CellTracker signal.

Time-lapse microscopy

Imaging of infected organoids was conducted using a Leica Thunder DMi8 microscope with a temperature-controlled microscope environmental chamber maintained at 37°C and 5% CO₂ in a stage-top chamber (OKOlabs) using a Leica HC FLUOTAR VISIR 25x (NA 0.95) water-immersion objective. To maintain the water immersion for the objective, water was pumped to the ring around the water objective. Microinjected organoids were identified and imaged at ex476/em519 nm (UPEC) and ex554/em594 nm (organoids) at 85 ms exposure time, acquiring the multiple channels during the same imaging sequence to improve the temporal resolution. Z stacks were acquired with 1 µm step sizes. Selected organoids were images 4 hours into the infection phase (corresponding to G3 of acquired images), throughout the antibiotic phase where indicated (corresponding to Amp0–3 of acquired images), straight after antibiotic removal (corresponding to RG0 of acquired images) and 3 hours after antibiotic withdrawal (corresponding to RG3 of acquired images). For experiments indicated, antibiotic treatment or regrowth phases were extended. UPEC fluorescence area overlapping with organoid area (meaning bacterial signal only from inside the infected organoid by masking the organoid signal on the GFP signal; the intra-organoid bacterial signal) was extracted at the indicated time-points using Fiji scripts Volumetric calc_growth and Volumetric calc_regrowth. For analysis, the obtained fluorescence area at any time-point (growth, antibiotic-mediated clearance or regrowth) of the intra-organoid bacterial signal was normalized to the same time-point of the PBS control group within one experiment. The intra-organoid masking therefore includes all bacterial signal measured within the organoid; luminal bacteria and tissue-associated bacteria together.

Blocking of acidification

Where indicated, 10 µM chloroquine (CQ) or 0.1 µM bafilomycin A1 (bafA1) was added for 3 hours prior to infection experiments. Afterwards, blocking was washed off with excessive amounts of prewarmed DMEM/F-12 before adding fresh media before UPEC infections. When enhancing for intracellular bacteria, such as co-localization experiments with Cathepsin L, blocking was added 1 hour before infection, and left throughout the infection and Cathepsin L staining (additional 4 hours).

Immunofluorescence staining

Staining of paraffin-embedded human bladder organoids.

hBOs at day 13–14 were extracted from BME with Cell Recovery Solution and fixed in 4% formaldehyde in PBS for 1 hour at 4°C. After washing, fixed organoids were resuspended in 50 µl of prewarmed Histogel at 50°C and pipetted out as a small hemispherical dome inside a 1-cm Tissue-Tek Cryomold. The cryomold was placed on a cold ice plate for solidification. Subsequently, the hemispherical Histogel was processed for paraffin embedding after overnight dehydration. Organoids embedded in paraffin were cut into 4 µm slices. The thin paraffin sections were deparaffinized and rehydrated by immersing the slides through the following solutions: xylene, three washes of 5 min each; 100% ethanol, two washes of 10 min each; 95% ethanol, two washes of 10 min each; 70% ethanol, two washes of 10 min each; 50% ethanol, two washes of 10 min each; PBS, three washes of 5 min each. Rehydrated slides were then processed for heat-induced antigen retrieval using 10 mM citrate buffer (pH 6.0). Slides were washed three times with PBS before proceeding. For histological analysis, hBO sections were stained with hematoxylin and eosin (H&E). Sections were stained in hematoxylin, differentiated in acid alcohol and blued in ammonia water, followed by counterstaining with Eosin Y. Sections were then dehydrated through a graded ethanol series and cleared in xylene before mounting and imaging on a slide-scanning microscope (Olympus VS200). For immunostaining, hBO sections were blocked with 1% BSA in PBS for 1 hour. The boundaries of paraffin sections were marked with a hydrophobic pen. Primary antibodies against cytokeratin (CK)7 for general bladder epithelium, CK5 for basal cells, CK13 for intermediate cells, and CK20 and uroplakin (UP)3a for umbrella cells were diluted 1:100 in 1% BSA in PBS and incubation was performed overnight at 4°C. After washing three times for 10 min each in PBS, slices were incubated with 1:1000 dilution of secondary antibody and DAPI in 1% BSA in PBS for 1 hour at room temperature. All used antibodies are listed in Table 4. Images were taken on a Leica Confocal Stellaris 5 using a HC PL APO 63x (NA 1.40) oil-immersion objective.

Table 4. Used antibodies.

Antigen	Used reactivity	Host	Supplier	Catalog Number
LC3b	mouse, human	rabbit	Abcam	ab192890
Rab11A	mouse	mouse	Santa Cruz	sc-166912
Rab27B	mouse	rabbit	Biorbyt	orb1264367
Rab7	mouse, human	mouse	Santa Cruz	sc-376362
Lamp1	mouse	rat	Santa Cruz	sc-19992
Lamp1	human	mouse	Abcam	ab25630
p62	mouse, human	rabbit	Abcam	ab109012
UP3A	mouse	mouse	Santa Cruz	sc-166808
UP3A	human	rabbit	Thermo Fisher	PA5–87581
CK20	mouse, human	rabbit	Biorbyt	orb256650
CK13	mouse, human	rabbit	Abcam	ab198584
CK7	human	rabbit	Abcam	ab181598
CK5	mouse, human	rabbit	Abcam	ab52635
GAPDH	mouse, human	rabbit	Abcam	ab181602
HRP	anti-rabbit IgG	goat	Agilent	P0448
Alexa Fluor-488	anti-rabbit IgG	donkey	Thermo Fisher	A-21206
Alexa Fluor-647	anti-rabbit IgG	donkey	Thermo Fisher	A-31573
Alexa Fluor-647	anti-mouse IgG	donkey	Thermo Fisher	A-31571

Differentiated monolayer stainings.

Confluent mouse and human monolayers were fixed with 4% formaldehyde for 1 hour at 4°C and then washed with PBS. For umbrella-like cells, cells were blocked for 1 hour at room temperature in 1% BSA before staining with UP3A in 1% BSA in PBS. After washing, cells were permeabilized with 0.1% Triton-X in PBS for 10 min. Cells were washed and blocked again before staining with primary antibodies for the main three bladder epithelial cell types: CK5 for basal cells, CK13 for intermediate cells and CK20 for umbrella-like cells. Primary antibodies were diluted 1:100 in 1% BSA in PBS. After washing, cells were incubated with 1:1000 dilution of secondary antibody, phalloidin-555 and DAPI in 1% BSA in PBS. Each staining was carried out for 1 hour at room temperature. All used antibodies are listed in Table 4. Images were taken on a Confocal Stellaris 5 using a HC PL APO 63x (NA 1.40) oil-immersion objective.

Vesicle staining.

Differentiated monolayers were infected with UPEC strain TS4 at roughly 1*10⁶ CFU/ml (corresponding to roughly 1*10⁵ CFU per well in 100 µl) for 4 hours. Thereafter, extracellular bacteria were washed with excessive amounts of PBS before fixing. Fixing, permeabilization and staining were carried out as mentioned above. All used antibodies are listed in Table 4. Images were taken on a Confocal Stellaris 5 using a HC PL APO 63x (NA 1.40) oil-immersion objective. Foci were counted in 3D using Fiji script 3D_foci_counting.

Western blot

Proteins from differentiated and TS4 infected monolayers were isolated by resuspending cells after TryPLE digestion in 300 µl Radioimmunoprecipitation assay (RIPA) buffer. The protein mixture was then either used immediately or aliquoted and stored at -80°C. Protein concentrations were measured with the Pierce BCA protein assay kit and approximately 3 µg of proteins were boiled for 10 min at 70°C with NuPAGE LDS sample buffer and Dithiothreitol (DTT) before running them on a 4–12% Bis-Tris gel. Proteins were transferred onto a membrane using the iBlot Gel Transfer Device. The membrane was cut to separate GAPDH and LC3b bands and the membrane part containing LC3b (15–17 kDA) was pretreated with SuperSignal West Atto substrate solution. After blocking both membrane parts for 1 hour at room temperature with 1% BSA in Tris-buffered saline (TBS), they were incubating with primary antibodies overnight at 4°C – GAPDH 1:2000 in 2% BSA in TRIS-NaCl-Tween 20 (TNT) buffer; LC3b 1:1000 in SuperSignal western blot enhancer. After washing with TBS, blots were incubated with secondary antibody 1:2000 in 2% BSA in TNT for 1 hour at room temperature prior to visualizing signals with SuperSignal West Pico PLUS chemiluminescent substrate. All used antibodies are listed in Table 4.

Uptake assays

Dextran-TMR uptake.

Differentiated mouse monolayers were incubated with fixable Dextran Tetramethylrhodamine (Dextran-TMR) at a final concentration of 1 mg/ml for 4 hours before repeated washes with PBS and fixation with 4% formaldehyde for 1 hour at 4°C. Images were taken on a Confocal Stellaris 5 using a HC PL APO 63x (NA 1.40) oil-immersion objective. To reduce low-frequency background variations, a Gaussian-blurred duplicate of the projection was subtracted from the original image. Background correction was performed using separate control images acquired under identical microscope settings in the absence of the labeled compound. Intracellular fluorescence was quantified after background subtraction of images of PBS or OM-89-treated cells using Fiji script mean_fluorescence.

Labelled antibiotic uptake.

Ampicillin and gentamicin were labelled using Pierce FITC Antibody Labeling Kit according to the manufacturer’s recommendations. Labelled stocks at 2 mg/ml of each antibiotic were aliquoted and stored at -20°C. FITC-labelling rendered the antibiotics ineffective at the used concentrations.

Differentiated monolayers were incubated with FITC-labelled ampicillin and gentamicin for 3 hours at a final concentration of 64.5 µg/ml (10x MIC of CFT073 for unlabelled ampicillin) and 100 µg/ml (usual concentration used for gentamicin protection assays [10]) respectively. Monolayers were either treated with OM-89, infected or treated and infected. For infection, monolayers were infected with UPEC strain TS4 at roughly 1*10⁶ CFU/ml (corresponding to roughly 1*10⁵ CFU per well in 100 µl) for 4 hours. 1 hour into infection, FITC-labelled ampicillin at 64.5 µg/ml was added. Cells were washed several times with excess amounts of PBS washes before adding DMEM/F-12 with 20 mM HEPES omitting PBS or OM-89 and taking images on a Confocal Stellaris 5 using a HC PL APO 63x (NA 1.40) oil-immersion objective. To reduce low-frequency background variations, a Gaussian-blurred duplicate of the projection was subtracted from the original image. Background correction was performed using separate control images acquired under identical microscope settings in the absence of the labeled compound. Intracellular fluorescence was quantified after background subtraction of images of PBS or OM-89-treated cells using Fiji script mean_fluorescence.

For antibiotic uptake per cell type, cells were treated as described above but instead of live-imaging, cells were fixed to stain for cell-specific cytokeratin markers. Briefly, cells were fixed in 4% formaldehyde for 30 min at 4°C. After permeabilization, cells were incubated for 15 min in quenching buffer (2.5 M glycine in PBS) before washing and blocking. Quenching was performed to reduce autofluorescence in the labelled antibiotic channel due to fixing. Cells were stained with primary antibodies against CK13 for intermediate cells and CK20 for umbrella-like cells. Primary antibodies were diluted 1:100 in 1% BSA in PBS. After washing, cells were incubated with 1:1000 dilution of secondary antibody, phalloidin-750 and DAPI in 1% BSA in PBS. Each staining was carried out for 1 hour at room temperature. All used antibodies are listed in Table 4. Images were taken on a Confocal Stellaris 5 using a HC PL APO 63x (NA 1.40) oil-immersion objective. To reduce low-frequency background variations, a Gaussian-blurred duplicate of the projection was subtracted from the original image. Binary masks were generated from the corresponding cell-type marker channel after thresholding on image to segment marker-positive cells using Fiji script cell mask_for_fluorescence. The resulting masks were converted to binary images and used to define the cellular area of interest for cell-specific intracellular fluorescence quantification after multiplication with the binary cell masks using Fiji script mean_fluorescence_cell mask. Background correction was performed using separate control images acquired under identical microscope settings in the absence of the labeled compound using Fiji script background_mean_fluorescence_cell mask. For these controls, the same cell-type masks were applied, and mean fluorescence values were calculated. Cell-specific intracellular fluorescence was quantified from background-subtracted images of OM-89-treated cells.

Colocalization of intracellular UPEC with acidic compartments, Lamp1 and Cathepsin L.

Differentiated mouse monolayers were infected with UPEC strain TS66 at roughly 1*10⁶ CFU/ml (corresponding to roughly 1*10⁵ CFU per well in 100 µl) for 4 hours on a rocking platform to facilitate bacterial uptake for intracellular bacteria imaging. Thereafter, extracellular bacteria were washed with excessive amounts of PBS. For staining of acidic compartments, cells were incubated with LysoTracker DND-99 Red at 50 nM for 30 min following manufacturer’s recommendations. Thereafter, monolayers were fixed with 4% formaldehyde for 1 hour at 4°C and stained with phalloidin and DAPI as mentioned above. For staining of Lamp1-positive vesicles, cells were fixed and stained with Lamp1 antibody, phalloidin and DAPI as mentioned above. For staining of enzymatically active compartments, Cathepsin L activity was stained using Magic Red Cathepsin L kit following manufacturer’s recommendations. After 3 hours of infection, the Magic Red substrate and NucBlue Live ReadyProbes Reagent (Hoechst 33342) was added on top for 60 min. After washing 3 times, cells were stained with CellMask Plasma Membrane Stain in Deep Red for 10 minutes. After washing again, cells were briefly fixed with 4% formaldehyde for 15 min at RT. Images were taken on a Confocal Stellaris 5 using a HC PL APO 63x (NA 1.40) oil-immersion objective. Intracellular bacteria were cropped by segmenting on bacterial signal inside the phalloidin signal of the bladder epithelial cells in Fiji using the orthogonal view. Colocalization of intracellular bacteria with either LysoTracker, Lamp1 or Cathepsin L signal was then analyzed using the Fiji script fluorescent_overlap_intracellular_bacteria by drawing a freehand region of interest (ROI) around bacteria and extracting the mean fluorescence intensity normalization over the area of the ROI. For random control ROIs, images were processed the same way except for moving the ROI from the bacterial cluster to an area without bacterial signal.

Intraorganelle pH

The intraorganelle pH of differentiated monolayers was estimated using LysoSensor Yellow/Blue DND-160 following manufacturer’s recommendations with minor modifications. Briefly, after infection with UPEC strain TS4 as mentioned above, LysoSensor was diluted to a final concentration of 5 µM in live cell imaging solution (140 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl₂, 1.0 mM MgCl₂, 20 mM HEPES, pH 7.4). Mouse and human bladder epithelial cells were incubated for 10 min and 20 min with the sensor, respectively, before measuring fluorescent intensity on a plate reader taking four measurements per well at Ex/Em of 329nm/440nm and 348nm/540nm.

Cathepsin assay

Cathepsin L activity in differentiated and UPEC strain TS4 infected monolayers (4h pi) was measured using Magic Red Cathepsin L kit following manufacturer’s recommendations. Cells were incubated with Magic Red substrate for 60 min and fluorescent intensity was measured on a plate reader taking four measurements per well at Ex/Em of 592nm/628nm.

RNA extraction, RT-qPCR and sequencing

Human bladder organoids.

hBOs were extracted from the BME using Cell Recovery Solution. For RNA extraction, organoids were incubated with the appropriate volume of RNA lysis buffer (RNAeasy Plus Mini Kit) and RNA was isolated following the manufacturer’s instructions. 250 ng of RNA were used to generate cDNA using the SuperScriptIV First-Strand Synthesis System with random hexamers. RT-qPCR primer sequences for UPK3A, KRT20, KRT13, TP63, ZO-1 and GAPDH are listed in Table 3. RT-qPCR reactions were prepared with SYBRGreen PCR Master Mix with 500 nM primers, and 1 µl cDNA. Reactions were run for quantification on QuantStudio 7 Flex Real-Time PCR System and amplicon specificity was confirmed by melting-curve analysis.

Mouse bladder organoids.

mBOs were either left uninfected or were infected with UPEC strain TS85 as described above. Four hours after microinjection (4h pi), the organoids were extracted from the BME using Cell Recovery Solution. For RNA extraction, organoids were incubated with the appropriate volume of RNA lysis buffer (RNAeasy Plus Mini Kit) and RNA was isolated following the manufacturer’s instructions. RNA quantification and quality control was performed on a TapeStation 4200. Libraries were prepared using the “Illumina stranded mRNA ligation” (ISML) prep, according to Illumina protocol 1000000124518 v01. To account for bacterial mRNA contaminations during the mRNA polyA capture step, 200 ng of RNA for the PBSinf (PBS+UPEC) samples was used due to the presence of a double set of rRNA peaks whereas for all other samples 125 ng of RNA was taken. Libraries were then sequenced on the NovaSeq 6000 yielding approximately 25–35 million, 60-bp, paired end reads for all samples.

RNA sequencing analysis

STAR (v.2.7.10b) [75] was used to align FASTQ reads to the mouse mm10 reference genome and count reads in genes from the annotations in Ensembl release 102. Genes were considered expressed and maintained for downstream analysis if they had more than 10 counts in at least 4 of the samples (14,114 genes remaining). Differential expression analysis was performed using DESeq2 (version 1.36.0) and genes were considered differentially expressed if they were changed by more than 2-fold (log2FC ≥ 1) and had a corrected p-value ≤ 0.05. Two methods were used for exploring affected pathways: first, an over-representation analysis, using the g:Profiler [76] R client, was used to determine which pathways were enriched for up and down-regulated genes separately. For the gene set enrichment analysis, all expressed genes were ranked by their wald statistic and analyzed using the fgsea toolset in R [77] for both GO and KEGG databases.

Genebridge MMAS analysis

Genebridge MMAS (Module-Module Association Score) within the GeneBridge platform was used to identify and quantify connections between different biological modules [45]. Briefly, GeneBridge uses cross-species transcriptome compendia and statistical analysis to determine the significance of relationships between modules. We analyzed the KEGG lysosomal category to identify models positively correlating with it in eight human bladder datasets (dataset identifier: GSE13507 – sample size 265, GSE31189 – sample size 92, GSE48276 – sample size 84, GSE32894 – sample size 308, GSE83586 – sample size 307, GSE86411 – sample size 132, GSE32548 – sample size 131, GSE31684 – sample size 93).

Statistical analysis

Statistics were performed with GraphPad Prism 10.3.1. Data and corresponding test results (e.g., normality results, T-values, degrees of freedoms, p-values) can be found in Supplementary Information.

Supporting information

S1 Fig. Controls and complementary analyses for the effects on growth, regrowth and antibiotic uptake driven by OM-89.

(A-C) Mouse bladder organoids treated with OM-89 under (A) pre-application, (B) co-application, or (C) continuous application regimes. Quantification of CFT073 fluorescence area at 4h post-infection (pi). Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test. N ≥ 15 per condition for (A), n ≥ 17 per condition for (B) and n ≥ 23 per condition for (C). (D) CFT073 signal over time in the continuous application regime. Quantification of bacterial fluorescence area normalized to PBS-treated organoids at the same timepoint. Each dot represents one organoid, lines connect matched organoids. RM two-way ANOVA with Šídák’s correction; p-value shown for treatment effect. N ≥ 14 per condition. (E) Cross-sectional (left) and single central plane (right) images of organoids with regrowing UPEC from the tissue 3h after antibiotic removal (10h pi) in the continuous treatment regime. Organoids shown in blue, UPEC shown in magenta. For central plane images: organoid boundaries indicated in cyan, luminal boundaries indicated in yellow. White arrow heads indicate UPEC regrowing in the organoid wall. (F) Uptake of FITC-labelled ampicillin in OM-89 treated mouse bladder epithelial cell monolayers according to cell type. For each cell type, a binary cell mask was generated from the corresponding cell-type marker channel (CK20+ for umbrella cells, CK13+ for intermediate cells) and used to restrict measurements to the cellular area. Quantification of fluorescence after background subtraction of images without labelled antibiotics where the same cell-type masks were applied. Mean ± 95% CI. Mann-Whitney test. N ≥ 10 per condition. Z-projection (maximum intensity) of representative images. CK20 and CK13-cell staining in magenta, phalloidin in grey, DAPI in cyan. Cell mask generated from the cell-specific stainings in white. Masked areas of FITC-ampicillin in cyan. (G), (H) Uptake of (G) FITC-labeled gentamicin and (H) Dextran-TMR in mouse bladder epithelial cell monolayers. Quantification of fluorescence after background subtraction of PBS- or OM-89-treated cells without labeled antibiotic. Mean ± 95% CI. Welch’s test. N = 8 per condition. Z-projection (maximum intensity) of representative images. FITC-gentamicin and Dextran-TMR shown in cyan.

(PNG)

S2 Fig. Establishment of differentiated mouse bladder epithelial cell monolayer.

Staining of differentiated mouse monolayer for bladder epithelial cell markers: Cytokeratin (CK)20 – umbrella-like cells. Uroplakin (UP)3A – umbrella-like cells. CK13 – intermediate cells. CK13-positive, binucleated cells indicate late intermediate cells. CK5 – basal cells. Binucleated cells indicated with asterisks.

(PNG)

S3 Fig. Transcriptomic analysis of OM-89-treated bladder organoids.

(A) Volcano plots of differentially expressed genes (log2FC ≥ 1; adjusted p ≤ 0.05) comparing OM-89- (cyan) and PBS-treated organoids without infection (dark blue). (B) Gene set enrichment analysis (GSEA) of treatment and infection groups. GO-BP terms were clustered into infection-relevant pathways. NES, normalized enrichment score. Non-significant changes not highlighted. (C) Gene set enrichment analysis (GSEA) of the top 10 Kyoto Encyclopedia of Genes and Genomes (KEGG) terms comparing infection vs no infection in PBS control organoids. NES, normalized enrichment score. Ns changes not highlighted. (D) GSEA of the top 10 KEGG terms comparing OM-89- and PBS-treated organoids without infection. Ns changes not highlighted (A-D) Data from four independent wells with n ≥ 50 organoids each.

(PNG)

S4 Fig. Controls and complementary analyses for lysosomal activity and blocking experiments in mouse bladder epithelium.

(A) Quantification of vesicle numbers and size for LAMP1 in uninfected monolayers with and without OM-89 treatment. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test for vesicle size and counts. N ≥ 11 per condition. Z-projection (maximum intensity) of representative images. Foci, cyan; DAPI, grey. (B) Intraorganelle pH of uninfected monolayers with and without OM-89 treatment. Mean ± 95% CI. Mann-Whitney test. N = 12 per condition. (C)-(E), (H) Quantification of colocalization of random regions of interest (ROIs) with (C) LysoTracker, (D) LAMP1, (E) Cathepsin L and (H) Cathepsin L with bafilomycin A1 (BafA1) in monolayers of mouse bladder epithelial cells. Intensities normalized to random ROI area. (C), (E), (H) Mann-Whitney test. (D) Welch’s test. N ≥ 24 per condition for (C), n ≥ 18 per condition for (D), n ≥ 19 per condition for (E) and n ≥ 16 per condition for (H). Z-projection (maximum intensity) of inserts in XYZ cross-section images shown in Fig 4D-4G. UPEC, magenta; LysoTracker/LAMP1/Cathepsin L, cyan. ROIs drawn around intracellular UPEC used for intensity quantification of specific/active co-localization in Fig 4D-4G. Random ROIs used for unspecific/random co-localization. Scale bar 2 µm. (F) Cathepsin L activity during infection with or without OM-89, and with BafA1 or chloroquine (CQ) blocking. Values normalized to unblocked PBS controls (dashed line). Mean ± 95% CI. Brown-Forsythe ANOVA with Welch’s correction. N = 10 per condition. (G) Cathepsin L activity in uninfected monolayers with or without OM-89 treatment. Values normalized to PBS controls (dashed line). Mean ± 95% CI. Welch’s t test. N = 10 per condition. (I), (J) CFT073 growth after (I) BafA1 and (J) CQ blocking. Quantification of bacterial fluorescence area inside organoids 4h pi. Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Mann-Whitney test. N ≥ 15 per condition for (I) and n ≥ 19 per condition for (J).

(PNG)

S5 Fig. Growth of different UPEC strains in mouse bladder organoids.

(A)-(D) Growth of UPEC strain (A) UTI89, (B) J96, (C) 721536–18 and (D) 721402–18. Quantification of bacterial fluorescence area inside organoids at 4h pi. Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s test for UTI89 (A), J96 (B), 721402–18 (D). Mann-Whitney test for 721536–18 (C). N ≥ 17 per condition for (A), n ≥ 15 per condition for (B), n ≥ 14 per condition for (C), (D).

(PNG)

S6 Fig. Establishment of differentiated human bladder epithelial cell monolayer.

Staining of differentiated mouse monolayer for bladder epithelial cell markers: Cytokeratin (CK)20 – umbrella-like cells. Uroplakin (UP)3A – umbrella-like cells. CK13 – intermediate cells. CK5 – basal cells. Binucleated cells indicated with asterisks.

(PNG)

S7 Fig. Autophagy-related markers with infection and lysosomal-related markers without infection in human bladder epithelial monolayers treated with OM-89.

(A), (B) Quantification of vesicle numbers and size for (A) P62 and (B) LC3B in infected monolayers of human bladder epithelial cells with and without OM-89 treatment. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test for P62 vesicle counts (A) and LC3B vesicle size (B). Mann-Whitney test for P62 vesicle size (A) and LC3B vesicle counts (B). N ≥ 11 per condition. Z-projection (maximum intensity) of representative images. Foci, cyan; DAPI, grey. (C) Western blot analysis of LC3B during infection of human monolayers. LC3B-II/GAPDH ratio normalized to PBS control (dashed line). Mean ± 95% CI. Brown-Forsythe ANOVA with Welch’s correction. N = 3 per condition. (D) Quantification of vesicle numbers and size for LAMP1 in uninfected monolayers of human bladder epithelial cells with and without OM-89 treatment. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s t test for vesicle size and counts. N ≥ 10 per condition. Z-projection (maximum intensity) of representative images. Foci, cyan; DAPI, grey. (E) Intraorganelle pH of uninfected monolayers with and without OM-89 treatment. Mean ± 95% CI. Mann-Whitney test. N = 12 per condition. (F) Cathepsin L activity in uninfected monolayers with or without OM-89. Values normalized to PBS controls (dashed line). Mean ± 95% CI. Welch’s t test. N = 11 per condition.

(PNG)

S8 Fig. Establishment of human bladder organoids from primary human bladder epithelial cells.

(A) Schematic of the human bladder organoid culture protocol. Primary H-6215 bladder epithelial cells (passages 6–8) were seeded as single cells into basement membrane extract (BME) domes and cultured in medium containing 50% HBO medium and 50% LWRN-conditioned medium (50/50). After 9 days, organoids could be cryopreserved. Following thawing, organoids were recovered for 1 day in 50/50 HBO/LWRN medium before switching to differentiation conditions (95% HBO medium with 5% LWRN, 95/5). Differentiated human bladder organoids were characterized or used for microinjection experiments on days 13–14. Created in BioRender. Tomasek, K. (2026) https://BioRender.com/9rdwrzl. (B) Growth of human bladder organoids over time. H-6215 cells embedded in BME rapidly expand and form organoids with a central lumen visible as early as 2 days after seeding. Images show representative examples from different organoids.

(PNG)

S9 Fig. Human bladder organoids express urothelial markers and form a tight lumen.

(A) Immunostaining of stratified and differentiated human bladder organoids derived from primary H-6215 cells for urothelial markers. Cytokeratin (CK)7 marks general urothelium; CK20 and uroplakin (UP)3A mark umbrella-like cells; CK13 marks intermediate cells; CK5 marks basal cells. Representative images of paraffin-embedded sections are shown. For CK20 and UP3A, two organoids are displayed. Organoids were fixed for staining on day 13 of differentiation. (B) RT-qPCR analysis of differentiation markers in human bladder organoids. Differentiated organoids (day 13, D13) upregulate umbrella cell markers (KRT20, UPK3A), intermediate cell markers (KRT3), and progenitor cell markers (TP63) compared to undifferentiated H-6215 monolayers (day 0, D0). Mann-Whitney test for KRT20 and KRT3. Welch’s t test for UPK3A and TP63. N ≥ 4. (C) (C) RT-qPCR analysis of tight junction protein 1 TJP1 (ZO-1) in human bladder organoids. Differentiated organoids (day 13, D13) upregulate TJP1 compared to undifferentiated H-6215 monolayers (day 0, D0). Welch’s t test. N ≥ 3. (D) Dextran microinjection assay confirming luminal barrier integrity in fresh and thawed organoids. Dextran fluorescence was quantified in the central plane at t0 (approximately 30 min post-injection) and t1 (t0 + 1h). Organoids were injected either after continuous differentiation from single cells (fresh) or after freezing at day 9 followed by thawing and differentiation (thawed). Each dot represents one organoid. Values normalized to t0 for each organoid. Mean ± 95% CI. Fold change in dextran signal from t0 to t1: fresh organoids, 0.78 ± 0.12; thawed organoids, 0.78 ± 0.16. N = 7. Representative images of central plane of fresh and thawed organoids. Cell mask, magenta; Dextran, cyan.

(PNG)

S10 Fig. Growth and regrowth of UPEC in human bladder organoids.

(A) Growth of UPEC strain CFT073. Quantification of bacterial fluorescence area inside organoids at 4h pi. Each dot represents one organoid. Values normalized to PBS control (dashed line). Mean ± 95% CI. Welch’s test. N ≥ 14 per condition. (B) Cross-sectional (left) and single central plane (right) images of human organoids with regrowing UPEC from the tissue 3h after antibiotic removal (10h pi) in the continuous treatment regime. Organoids shown in blue, UPEC shown in magenta. For central plane images: organoid boundaries indicated in cyan, luminal boundaries indicated in yellow. White arrow heads indicate UPEC regrowing in the organoid wall.

(PNG)

S1 Video. Post-antibiotic UPEC regrowth in a PBS-treated mouse bladder organoid.

Related to Fig 1 and S1E Fig. Time-lapse microscopy of a representative PBS-treated mouse bladder organoid (blue) microinjected with CFT073 (magenta), imaged at the end of the regrowth phase. Bacterial regrowth originates preferentially from the organoid wall rather than from luminal compartments.

(MP4)

S2 Video. Post-antibiotic UPEC regrowth in an OM-89-treated mouse bladder organoid.

Related to Fig 1 and S1E Fig. As in S1 Video, for a representative mouse bladder organoid treated with OM-89. Bacterial regrowth remains reduced relative to PBS but continues to originate preferentially from the organoid wall.

(MP4)

S3 Video. Post-antibiotic UPEC regrowth in a PBS-treated human bladder organoid.

Related to Fig 7I and S10B Fig. Time-lapse microscopy of a representative PBS-treated human bladder organoid (blue) microinjected with CFT073 (magenta), imaged at the end of the regrowth phase. Bacterial regrowth originates preferentially from the organoid wall rather than from luminal compartments.

(MP4)

S4 Video. Post-antibiotic UPEC regrowth in an OM-89-treated human bladder organoid.

Related to Fig 7I and S10B Fig. As in S3 Video, for a representative human bladder organoid treated with OM-89 showing reduced but similarly wall-associated bacterial regrowth after antibiotic withdrawal.

(MP4)

S1 Raw Images. Original uncropped and unadjusted blot images underlying Fig 4E and S7C Fig.

(PDF)

Data Availability

Raw RNA-sequencing data have been deposited in the Gene Expression Omnibus (GEO) under accession number GSE306613. Raw and processed microscopy imaging data are available at the BioImage Archive under accession (https://www.ebi.ac.uk/biostudies/bioimages/studies/S-BIAD3885). Quantified results and statistical analyses are available at Zenodo (https://doi.org/10.5281/zenodo.21790270). Analysis code (Fiji macros) is available at Zenodo (https://doi.org/10.5281/zenodo.21789901) and GitHub (https://github.com/klebsiellat/Targeted-lysosomal-activation-in-bladder-epithelium-by-OM-89).

Funding Statement

This work was supported by the Swiss Government Excellence Scholarship (2022.0289, to K.T.), the Swiss National Science Foundation (SNSF) Postdoctoral Fellowship (TMPFP3_217144, to K.T.), an IBSA Foundation Grant (IBSA Foundation Fellowships Call 2021 in Fertility/Urology, to K.T.), Fondation Leenaards (to J.M.), the SNSF National Centre of Competence in Research (NCCR) AntiResist (51NF40_180541, to J.M.), and OM Pharma (to J.M.). The Swiss Government Excellence Scholarship, the SNSF Postdoctoral Fellowship, and the IBSA Foundation Fellowship are personal fellowships that provided salary support to K.T. OM Pharma provided salary support for K.S. and supplied OM-89 (Uro-Vaxom) as study material. C.P. and M.R. received their salary from OM Pharma in their capacity as employees of OM Pharma. The Swiss Government Excellence Scholarship, the SNSF, the IBSA Foundation and Fondation Leenaards had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.