Essential role for the urothelial plaque in Gram-negative urinary tract infections

Abstract

Over 90% of urinary tract infections (UTI) are caused by Gram-negative bacteria, especially uropathogenic Escherichia coli (UPEC). Bladder superficial cells produce Uroplakin (Upk) containing urothelial plaques that bind to UPEC type I fimbriae. The role of the plaque in UPEC urothelial invasion and triggering the host inflammatory response remains unclear. In this study, we used Upk1b KO mice to test the hypothesis that a functional urothelial plaque is required for UPEC UTI. Following UPEC inoculation, Upk1b KO mice exhibited reduced bacterial burden, impaired urothelial invasion, and absent intracellular bacterial communities, compared to wild-type hosts. Transcriptome analysis identified attenuation of innate immune pathways in Upk1b KO bladders and flow cytometry confirmed limited phagocyte recruitment. Our results demonstrate that the plaque is essential for UPEC urothelial invasion, establishment of UTI, and initiation of the innate immune response. Notable study limitations include the potential that observations made in mouse models are associative rather than causal and that our conclusions may not fully apply to human UTI. While our findings strengthen the rationale for targeted therapies that disrupt UPEC-plaque interactions, such strategies may prove challenging given the fundamental role of the plaque in preserving urothelial integrity and barrier function.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-026-59987-6.

Introduction

Bacterial UTI are common infections across the human lifespan with a global incidence exceeding 400 million cases annually^1–3. Mounting rates of antibiotic resistance pose challenges for clinicians treating UTI and provide impetus for discovery of novel antimicrobial agents to prevent future infections^4–6. Uropathogenic Escherichia coli (UPEC) account for 90% of bacterial UTI, and much of the knowledge base regarding UTI pathogenesis is based on interactions between UPEC and bladder urothelial cells. Umbrella cells (UC) are terminally differentiated, superficial urothelial cells, bordered apically by the urinary space and basally by intermediate cells^7,8. Based on their location, UC are poised to detect UPEC pathogen associated molecular patterns (PAMPs) via pattern recognition receptors (PRR) and trigger signal transduction pathways that protect the underlying bladder parenchyma from invasion by UPEC and initiate the recruitment of neutrophils to the urinary space. UC possess a host of specialized mechanisms to prevent bacterial attachment and bladder invasion including an ability to undergo apoptosis in response to UPEC, to detach and exfoliate, as well as to expel UPEC from endosomal compartments^9–12.

Conversely, UPEC have evolved sophisticated mechanisms to attach to UC, invade the urothelium, and form intracellular reservoirs such as intracellular bacterial communities (IBC) within UC^13,14. The UPEC type 1 pilus contains FimH which binds to the host Uroplakin 1a (Upk1a) protein in a mannose-dependent manner^13,15–17. UPEC lacking FimH display reduced UC adherence and invasion in vitro and attenuated virulence in vivo^18–20. This has prompted the development of mannosides, FimH-based vaccination, and anti-FimH monoclonal antibodies to interrupt mannose-dependent interactions between UPEC and urothelium and serve as anti-infectious agents to prevent and treat UTI^21–27.

Despite the foundational role for FimH in UTI pathogenesis and the identification of Upk1a as its receptor, less is known regarding the function of Upk proteins during UTI pathogenesis. UC express 4 major Upk proteins (Upk1a, Upk1b, Upk2, and Upk3a) that initially heterodimerize and are subsequently assembled into higher-order multimers that are exported to the plasma membrane as a crystalline lattice called the urothelial plaque^28,29. This highly specialized structure serves a critical role in forming the urine permeability barrier and conferring bladder compliance^30,31. We previously demonstrated that Upk1b plays a critical role in urothelial plaque formation³². In this study, we made use of Upk1b knockout (KO) mice to test the hypothesis that a functional urothelial plaque is essential for UPEC invasion of the urothelium, establishment of UTI, and initiation of the host innate immune response during experimental cystitis.

Methods

Approval for animal experiments

Animal experiments were conducted under protocols (numbers AR16-00058 and AR13-00057) that were approved by the Institutional Animal Care and Use Committee at the Abigail Wexner Research Institute at Nationwide Children’s Hospital (Public Health Service Animal Welfare Assurance number A3544-01). All experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals at the Abigail Wexner Research Institute at Nationwide Children’s Hospital.

Upk1b KO mice

Upk1b KO mice were originally generated by the GenitoUrinary Development Molecular Anatomy Project Consortium and are maintained on an FVB/NJ background³². Wild-type FVB/NJ mice (Jackson Laboratories, Bar Harbor, ME) were maintained in a colony at our institution.

Experimental UTI

UTI was induced in isoflurane anesthetized Upk1b WT and KO female mice age six to eight weeks by transurethral inoculation with UTI89, a clinical UPEC isolate from a patient with cystitis¹³. Alternatively, to model Gram-positive UTI, mice were inoculated with 0852, an Enterococcus faecalis strain isolated from the human urinary tract³³. UTI89 or 0852 was inoculated from a glycerol stock and grown statically in LB medium for 16 h at 37 °C. The inoculum was 10⁷ colony forming units (CFU) in 50 µl phosphate buffered saline (PBS). At indicated times after infection, mice were re-anesthetized and euthanized by exsanguination, urinary tract organs were isolated and homogenized, and bacterial burden was assessed³⁴. Alternatively, bladders were fixed in 10% formalin, embedded in paraffin, sectioned at four micron thickness, stained with hematoxylin and eosin, and mucosal injury and inflammation were scored by a veterinary pathologist (G.B.) in blinded fashion^35,36. IBC were visualized based on β-galactosidase activity and enumerated as described^37,38. To distinguish luminal from intracellular bacteria, bladders were harvested 16 h post infection and evaluated by ex vivo gentamicin protection assay as described³⁸.

Immunofluorescence microscopy

Immunofluorescent labeling was performed as previously described^32,39. Kidneys were formalin-fixed, paraffin-embedded and sectioned at 4 μm and then deparaffinized and rehydrated prior to heat induced antigen retrival using a citrate-based solution. Immunolabeling was performed using: goat α-Upk1a (1:500, Santa Cruz Biotechnology, Inc., Dallas, TX), mouse α-Upk1b (1:800, Sigma-Aldrich, Inc., St. Louis, MO), mouse α-Upk3a (1:200, Fitzgerald, Acton, MA), chicken α-Krt5 (1:500, BioLegend, San Diego, CA), goat α-E. coli (1:800, Abcam, Waltham, MA), rat α-Ly6G (1:200, BioLegend), rabbit α- NF-kB p65 (1:500, Abcam), and rabbit α-E-cadherin (1:800, Cell Signaling Technology, Danvers, MA). Fluorescent detection of primary antibodies was achieved using Cy3- and Alexa Fluor 488–conjugated donkey-derived secondary antibodies (1:300; Jackson ImmunoResearch Laboratories, West Grove, PA). Autofluorescence was quenched using 0.1% Sudan black B (Thermo Fisher, Wilmington, DE) and nuclei were detected using Hoechst (1:2000, Thermo Fisher). Coverslips were mounted using ProLong Gold Antifade mountant (Invitrogen, Waltham, MA) and sections were imaged at 40x using a Nikon Eclipse Ti2 microscope (Nikon Instruments, Inc., Melville, NY). Negative control sections were incubated with irrelevant species-specific Ab or secondary Ab alone.

FITC-dextran permeability assay

To evaluate urothelial barrier integrity, we used fluorescein isothiocyanate (FITC)–conjugated dextran (10,000 Dalton; Thermo Fisher Scientific). Mice were anesthetized, and the bladder was catheterized via transurethral insertion of a lubricated catheter. A 50 µL solution of FITC-dextran in sterile phosphate-buffered saline was instilled into the bladder lumen. After 20 min, bladders were harvested, fixed in 4% paraformaldehyde, embedded in optimal cutting temperature (OCT) compound, and cryosectioned. Sections were counterstained with Hoechst to identify nuclei and imaged using an Olympus BX51 microscope and CX9000 camera (Olympus Corporation, Shinjuku, Japan).

Electron microscopy

Transmission electron microscopy (TEM) was performed as previously demonstrated^32,37 to visualize the urothelial barrier and uroplakin plaques. Briefly, tissues were perfused with phosphate-buffered saline followed by 2% glutaraldehyde/PBS, then immersion-fixed in 2.5% glutaraldehyde. Samples were post-fixed in 1% osmium tetroxide, dehydrated through a graded ethanol series, cleared in propylene oxide, and embedded following sequential infiltration with a 1:1 mixture of Epon-Araldite: propylene oxide and then pure Epon-Araldite. Ultrathin Sect. (60 nm) were collected on palladium-coated copper grids, counterstained with uranyl acetate and lead citrate, and examined using a Hitachi H-7650 transmission electron microscope (Hitachi Ltd., Chiyoda, Tokyo, Japan).

Scanning electron microscopy (SEM) was performed to visualize the urothelial surface architecture. Briefly, animals were euthanized, ventrally opened, and perfused through the left ventricle of a beating heart with phosphate-buffered saline. Bladders were harvested, inverted to expose the urothelial lining, bisected, and fixed overnight in 4% paraformaldehyde, followed by 30 min in 2.5% glutaraldehyde/PBS. Samples were incubated sequentially in 1% osmium tetroxide, 1% tannic acid, and 0.1 M cacodylate buffer, with intervening washes in double-distilled water. After dehydration through a graded ethanol series, tissues were incubated briefly in hexamethyldisilazane and dried overnight. Specimens were sputter-coated with gold (Emitech K550X, Quorum Technologies Ltd., Ashford, Kent, England) and examined using a Hitachi S-4800 scanning electron microscope (Hitachi High Technologies America, Inc., Dallas, TX).

Flow cytometry

Bladders were dissected twenty-four hours after UPEC inoculation, minced and dissociated in DMEM/F12 medium containing 50 mg/mL Collagenase I and 10 mM HEPES. Next, single cell suspensions were blocked with anti-mouse Fc receptor antibody Cells were stained with blue-fluorescent reactive dye (L23105, Life Technologies) for 20 min at room temperature to exclude dead cells from analysis. After washing, cell surface receptors were stained with fluorescent antibodies: PerCP/Cy5.5 anti-mouse Ly-6G Ab (BioLegend, clone 1A8), Brilliant Violet 650 anti-mouse I-A/I-E Ab (BioLegend, clone M5/114.15.2),, Brilliant Violet 785 anti-mouse CD45 Ab (BioLegend, clone 30-F11), Alexa Fluor 700 anti-mouse/human CD11b Ab (BioLegend, clone M1/70), PerCP/Cy5.5 anti-mouse Ly6G Ab (Biolegend, clone 1A8), PE anti-mouse Cx3cr1 Ab (BioLegend, clone SA011F11), PE anti-mouse CD45 Ab (Biolegend, clone 30-F11), PE/Cy7 anti-mouse Ly6C Ab (BioLegend, clone HK1.4) and eFluor™ 450 anti-mouse Ly-6 C Ab (Invitrogen, clone HK1.4). Stained cells were detected using a LSR II cytometer (BD, Franklin Lakes, NJ). Data were analyzed with Flowjo software (Tree Star, Ashland, OR). CountBright Absolute Counting Beads (Thermo Fisher) were utilized to determine absolute cell numbers. Monocytes were identified as CD45 + CD11b+Ly6G-Ly6C+CX3CR1-; macrophages were CD45 + CD11b+Ly6G-Ly6C-CX3CR1+; and neutrophils were CD45 + CD11b+Ly6G+.

RNA extraction

RNA was isolated from female Upk1b WT and KO bladders (4 biological replicates/group) at baseline and 24 h following infected with UTI89 using the MirVana^™ Paris^™ kit (Thermo Fisher). Bulk RNA sequencing (RNA-seq) was performed by the Genomic Services Laboratory at Nationwide Children’s Hospital. RNA sample integrity was assessed using an Agilent 2100 Bioanalyzer.

RNA-seq analysis

FASTQ files were processed through an RNA-seq pipeline for normalization, differential gene expression (DEG) analysis, and functional enrichment^39–43. The mm10 mouse genome and reference annotation were obtained from UCSC (https://hgdownload.soe.ucsc.edu/goldenPath/mm10/bigZips/). Reads were trimmed with Trim Galore v0.6.6⁴⁴, aligned using Tophat v2.1.29⁴⁵, and unique mapped reads retained with SAMtools v1.15.10⁴⁶. Gene-level counts were generated using Htseq-count v0.12.4⁴⁷. Low-abundance genes were filtered, and quantile normalization was applied. DEGs were identified with EdgeR_3.42.4 (fold change > 1.5, adjusted p < 0.05)⁴⁸. PCA was used to assess overall expression patterns. Functional enrichment of DEGs was conducted using clusterProfiler_4.8.2 for Gene Ontology (GO) and pathway analyses (FDR ≤ 0.05)⁴⁹. Expression of urothelial lineage markers, developmental pathways, and innate immune genes was examined. Visualization was performed in R using ggplot2, enrichplot, and pheatmap to generate heatmaps, volcano, bar, and Venn plots.

Statistics

Statistical analysis was obtained using Prism (GraphPad Software, La Jolla, CA). Continuous data were assessed for normal distribution by the D’Agostino-Pearson Omnibus test. Normally distributed data were evaluated by student’s t-test; the Mann-Whitney U test was used when comparing nonparametric data. Only p values < 0.05 were considered significant.

Results

Disruption of urothelial plaque formation and barrier function in Upk1b KO mice

Urothelial plaque formation requires adequate expression of all four major Upk subunits, as they heterodimerize and form higher order multimers to generate the plaque^28,29. We previously demonstrated that Upk1b deficiency results in absent urothelial plaque formation based on transmission electron microscopy³². In addition, Upk1b KO bladders expressed reduced levels of all major Upk subunits, including Upk1a³², which binds directly to UPEC¹⁵. To verify these findings, we performed immunofluorescence microscopy and detected reduced surface Upk1a expression by Upk1b KO UC, along with absence of Upk1b as expected (Fig. 1A). Surface electron microscopy demonstrated the smooth surface of KO apical UC in contrast to the rugged appearance of WT UC, and transmission electron microscopy confirmed absence of ultrastructure associated with urothelial plaque production (Fig. 1B-C)³². To investigate the impact of Upk1b deficiency on the urine permeability barrier, we instilled FITC dextran into WT and Upk1b KO bladders. When visualized by fluorescence microscopy, we observed that FITC-Dextran was excluded from the urothelium and entirely retained in the lumen of WT bladders, consistent with an intact urine permeability barrier, whereas urothelial penetration was exclusively observed in Upk1b KO bladders (Fig. 1D). These results establish that Upk1b deficiency disrupts expression of Upk1a, urothelial plaque formation, and the urothelial barrier.

Fig. 1.

Defects in Upk localization, urothelial plaque formation, and urothelial permeability in Upk1b KO bladders. (A) Immunofluorescence micrographs indicate uniform apical distribution of Upk1a and Upk1b in WT bladder urothelium (white arrowheads), while immunoreactivity is absent in KO bladders. (Dashed lines indicate the urothelial basement membrane; “U” indicates the urinary space; n=4 WT and KO mice); (B) Transmission electron microscopy (EM) and insets indicate presence of urothelial plaque (red arrowhead) and fusiform vesicles (red arrow) in WT urothelium and absence of these ultrastructural features in KO urothelium (n=3 WT and KO mice); (C) Scanning EM demonstrates the rugged apical surface of WT urothelium compared to the smooth surface of KO urothelium (n=3 WT and KO mice); (D) Intravesical FITC-Dextran injection led to luminal retention in WT bladders whereas KO bladders exhibited penetration of FITC-Dextran through the urothelium (green arrowheads), consistent with impaired barrier function. (Dashed lines indicate the urothelial basement membrane; n=4 WT and KO mice).

Upk1b KO mice are protected from UTI

To test the hypothesis that an intact urothelial plaque is required to establish UTI, we transurethrally inoculated WT and Upk1b KO mice with 10⁷ colony forming units (CFU) of the human cystitis UPEC isolate, UTI89¹⁰, and enumerated UPEC CFU at various times post infection (Fig. 2). Comparable UPEC were obtained 2 h post inoculation (hpi), likely due to persistent luminal UPEC in KO bladders. As early as 6 hpi, we recovered fewer UPEC from Upk1b KO bladders, and this difference was sustained up to 7 days post infection (dpi). We also identified reduced susceptibility to UPEC infection in other Upk1b KO tissues including urine (24 hpi), ureters (16 and 24 hpi), and kidneys (24 hpi), when compared to WT controls (Fig. 2). Given the marked decrease in susceptibility to UPEC cystitis, we also challenged Upk1b KO and WT mice with Enterococcus faecalis, a Gram-positive uropathogen. We did not observe any difference in bladder and kidney burden between genotypes 24 hpi (Fig. S1), indicating that Gram-positive UTI can occur in the absence of a urothelial plaque.

Fig. 2.

Upk1bKO mice are less susceptible to experimental UTI than WT controls.UPEC burden in indicated tissues from WT and Upk1b KO mice at each indicated time (hpi: hours after inoculation). Each symbol corresponds to UPEC burden from a single mouse. The horizontal line indicates the geometric mean. The dashed line indicates the limit of detection (* p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; Mann-Whitney U test; ns = not significant).

Absent UC invasion and IBC formation in Upk1b KO mice

We hypothesized that reduced bacterial recovery from Upk1b KO bladders observed 6 hpi through 7 dpi was due to failure of UPEC to invade umbrella cells. To test this, we initially performed immunofluorescence microscopy using an anti-E. coli antibody 0.5 and 2 hpi. While we easily visualized UPEC within infected urothelial cells in WT bladders, UPEC localized exclusively to the lumen of infected Upk1b KO bladders (Fig. 3A), suggesting an inability to invade the urothelium during these timepoints. To further substantiate a difference in intracellular UPEC, we performed gentamicin protection assays and enumerated intracellular UPEC 16 hpi. Results from this experiment demonstrate that fewer intracellular UPEC are recovered from Upk1b KO versus WT bladders following gentamicin treatment (Fig. 3B). Since intracellular UPEC invasion is required for UPEC to form IBC, we also visualized IBC by β-galactosidase (β-gal) staining of Upk1b KO and WT bladders 16 hpi. IBC were easily identified and enumerated in WT bladders and confirmed by immunofluorescence microscopy using an E. coli specific antibody, but we were unable to visualize IBC in Upk1b KO bladders (Fig. 3C,D).

Fig. 3.

Impaired UPEC attachment, invasion, and IBC formation in Upk1b KO bladders. (A) Immunofluorescent micrographs demonstrate UPEC attachment (0.5 hpi) and invasion (2 hpi) in WT urothelium, whereas UPEC are retained in the urinary space (U) in Upk1b KO. (B) Reduced intracellular UPEC recovery from Upk1b KO bladders 16 hpi, assessed by gentamicin protection assay. (C) Absent IBC formation in Upk1b KO compared to WT bladders 16 hpi.

Reduced phagocyte recruitment in Upk1b KO mice

A hallmark of the host response to acute cystitis is the recruitment of circulating phagocytes to the bladder⁵⁰. We detected significant increases in the absolute number of neutrophils and monocytes in infected WT and Upk1b KO bladders compared to their respective uninfected controls by flow cytometry, but their recruitment was blunted in infected Upk1b KO bladders (Fig. 4A,B). We also observed fewer macrophages in infected Upk1b KO bladders (Fig. 4C). Following their recruitment to the bladder, neutrophils undergo transurothelial migration toward UPEC in the bladder lumen^50–52. We detected a substantial reduction in urothelial transmigration of neutrophils in infected Upk1b KO bladders (Fig. 4D,E). Despite blunted phagocyte recruitment and transmigration, histologic grading identified comparable mucosal injury and inflammation in infected Upk1b KO and WT bladders (Fig. 4F,G). In contrast, WT kidney mucosal injury and inflammation scores exceeded those recorded in Upk1b KO kidneys following UPEC infection (Fig. 4H,I).

Fig. 4.

Impact of Upk1b deficiency on phagocyte recruitment, transurothelial migration, and urinary tract pathology. (A–C) Flow cytometry identifies reduced absolute numbers of neutrophils, monocytes and macrophages to infected Upk1b KO bladders. Each symbol represents the indicated cell population in an individual mouse. (D) Limited transurothelial migration of Ly6G+ neutrophils in Upk1b KO bladders, assessed by immunofluorescence microscopy. (E) Representative micrographs of WT and KO urothelium (Ly6G: green; Hoechst; blue). The basement membrane is indicated by white dashes. U: Urinary space. (F–I) Blinded histologic grading demonstrates comparable increases in mucosal injury and inflammation in infected WT and KO bladders, while kidney scores were significantly blunted in KO kidneys. Each symbol represents an individual mouse. * p < 0.05; ** p < 0.01, Mann-Whitney U test; ns = not significant.

Global transcriptional changes in Upk1b KO mice

To better understand the impact of Upk1b deficiency on the innate immune properties of the urothelium, we profiled global transcriptional changes in WT and Upk1b KO bladders at baseline and 24 hpi by bulk RNA-seq. Principal component analysis (PCA) revealed distinct separation among the four experimental conditions at the whole transcriptome level. PC1 (29.9%) effectively distinguished the WT and Upk1b KO bladders samples, while PC2 successfully distinguished between infected and non-infected samples (Fig.S2). Furthermore, PCA established that UTI had a decreased impact on the global transcriptome in Upk1b KO compared to WT hosts.

Upk1b deficiency alters expression of genes associated with urothelial development, homeostasis, and structural integrity in the absence of UTI

To evaluate the impact of Upk1b deficiency on the bladder transcriptome in the absence of UTI, we conducted a comparative analysis of the baseline transcriptomes across genotypes. This identified 1190 up-regulated and 777 down-regulated transcripts (Fig. S3A) with distinct clustering of transcriptomes according to genotype (Fig. S3B). Gene Ontology (GO) analysis identified significant enrichment of transcripts associated with epithelial differentiation, cell-substrate adhesion, cell-cell junction organization, intermediate filament-based processes, and mitotic cell cycle phase transition in Upk1b KO bladders (Fig.S3C). In contrast, Upk1b KO bladders expressed reduced levels of transcripts associated with protein translation and metabolic processes compared to WT controls (Fig.S3D).

To further clarify the impact of Upk1b deficiency on expression of transcripts associated with urothelial differentiation, we examined the expression of mRNA associated with specific urothelial types, including Superficial (S), Intermediate (I), and Basal (B) cells (Fig.S3E). In agreement with our previous study, Upk1b KO bladders expressed altered levels of Upk subunits that participate in urothelial plaque formation and increased levels of keratins associated with S cells (Krt20, Clu, and Abhd2) and B cells (Krt14)³². Upk1b KO bladders also showed increased expression of transcripts associated with DNA replication (Top2a, Mki67 and Birc5) and G2/M phase transition (Table S1), consistent with an enhanced proliferative index in Upk1b KO urothelium as noted previously³².

Upk1b deficiency severely limits induction of innate immune transcripts following UPEC exposure

We next conducted pairwise RNA-seq comparisons among Upk1b KO and WT mice at baseline and 24 hpi. In WT mice, we identified 1167 upregulated genes and 1323 downregulated genes following UPEC inoculation (Fig.S4A,B). GO analysis established that transcripts induced by UTI in WT mice were associated with mitosis, cytokine-mediated signaling, leukocyte migration and the innate immune response, while downregulated genes were associated with metabolic processes (Fig.S4C,D). In contrast, only 50 upregulated genes and 464 downregulated genes were detected in infected Upk1b KO bladders (Fig.S5A,B). There was no significant association of up-regulated genes with GO enrichment terms, while UTI led to decreased expression of genes associated with muscle contraction and extracellular matrix composition in Upk1b KO bladders (Fig.S5C).

To gain insight regarding the limited induction of transcripts associated with the innate immune response in infected Upk1b KO bladders, we conducted a thorough analysis of differential gene expression and GO enrichment terms in baseline and infected transcriptomes from WT and Upk1b KO mice (Table S2–S16). In the absence of infection, Upk1b KO bladders showed increased expression of transcripts that regulate the innate immune response - including phagocytosis, neutrophil activation and migration, when compared to WT bladders. However, this apparent activation of the innate immune response was not further augmented 24 hpi in Upk1b KO bladders, whereas these pathways underwent robust activation in infected WT bladders (Fig. 5A).

Fig. 5.

Dysregulation of innate immune transcription in Upk1b KO bladders. (A) GO analysis demonstrates that WT and KO bladders differ in transcriptional regulation of innate immune pathways at baseline and following UPEC infection. All data shown are relative to WT uninfected bladders. (B) UTI leads to nuclear localization of p65 NFκB in WT superficial cells visualized by immunofluorescence microscopy (white arrowheads) that is absent in KO bladders. (C) Heatmap illustrates robust induction of chemokines and cytokines implicated in phagocyte recruitment and activation in infected WT bladders. In contrast, the induction of these NFκB transcriptional targets is attenuated in KO bladders.

GO enrichment analysis revealed that TLR and Nuclear Factor Kappa B (NF-κB) signaling pathways were significantly activated by infection in WT but not Upk1b KO bladders (Fig. 5A). This is consistent with the observation that dual ligation of Toll-like receptor 4 (by LPS) and Upk1a (by FimH) is required for optimal NF-κB induction in UC by UPEC¹⁸. Indeed, experimental UTI in WT mice led to nuclear localization of the p65 subunit of NF-κB in S cells (Fig. 5B) along with induction of canonical NF-κB transcriptional targets including cytokines (Il1a, Il1b, Tnfa) and chemokines (Cxcl1, Cxcl2, Ccl2, Ccl3, Ccl4) that regulate phagocyte recruitment and activation (Fig. 5C). However, NF-κB p65 nuclear localization and induction of NF-κB target genes were attenuated in Upk1b KO bladders (Fig. 5B,C).

Discussion

The goal of this study was to test the hypothesis that the urothelial plaque is required for Gram-negative UTI. Previous biochemical studies substantiated a role for the essential plaque component, Upk1a, in binding to UPEC-derived FimH in a mannose-dependent manner^13,15–17. The role for FimH in UPEC urothelial association, invasion and IBC formation was further supported by studies showing loss of these properties in FimH deleted UPEC^18–20. However, previous studies have not directly addressed the role of the urothelial plaque in UPEC invasion of the urothelium, establishment of UTI, and initiation of the host innate immune response. Here, we address these knowledge gaps by utilizing Upk1b KO mice with an inherent defect in urothelial plaque assembly.

The urothelial plaque is required to establish acute cystitis and ascending infections

We found that absent Upk1b expression and urothelial plaque formation was associated with reduced urothelial invasion, absent IBC formation, and overall decreased UPEC burden throughout the urinary tract. These findings are reminiscent of observations in wild-type mice inoculated with FimH deleted UPEC^18–20. Nuclear p65 NF-κB expression and transcriptional induction of NF-κB-dependent cytokines and chemokines were attenuated in infected Upk1b KO bladders, consistent with a previous study showing that dual engagement of PRR by UPEC derived LPS and FimH is required for optimal NFκB activation in urothelial cells¹⁸. Alternatively, these differences may arise from reduced UPEC binding and/or internalization by urothelial cells, thereby decreasing TLR4 stimulation.

We noted increased baseline expression of transcripts that regulate the innate immune response in Upk1b KO bladders, including phagocytosis, neutrophil activation and migration, when compared to WT controls. The underlying mechanism(s) that account for these differences are unclear and may relate to observed defects in urothelial maturation and/or barrier function in KO mice³². Following UPEC inoculation, Upk1b KO bladders manifested a similar degree of mucosal injury and inflammation based on histologic grading as WT controls, despite differences in phagocyte recruitment between Upk1b KO and WT mice. Possible explanations for similar pathology despite differences in cellular recruitment include: (1) Defective barrier function in Upk1b KO urothelium; and (2) Reduced expression of IL-8 like chemokines, Cxcl1/2, by Upk1b KO urothelium, which can lead to impaired transurothelial migration of neutrophils and persistence of neutrophils in the urothelium and submucosa, fomenting bladder injury^51–53. Altogether our findings illustrate the roles of an intact urothelial plaque in regulating the innate immune properties of the urothelium at baseline and following UTI.

Study limitations

There are several limitations to this study that we wish to point out. While the use of Upk1b KO mice enabled us to broadly implicate Upk1b and the urothelial plaque in UPEC invasion and the innate immune, the specific role of the plaque in governing these properties remains unclear. Moreover, the lack of a gain-of-function approach prevented us from rescuing Upk1b expression and reversing observed phenotypes in Upk1b KO mice. Finally, this study relied exclusively upon observations made in WT and Upk1b KO mice, and the extent to which our observations apply to human UTI though plausible is not fully appreciated.

Clinical significance and conclusions

In conclusion, this study establishes an essential role for Upk1b and the urothelial plaque in Gram-negative UTI. Our findings provide further rationale for plaque-directed therapies to abrogate UPEC association with urothelium and mitigate UTI. Up to this point, UTI preventative and therapeutic approaches have been directed toward UPEC-derived FimH, including FimH vaccination strategies^24,26,27, monoclonal antibodies directed against FimH²⁵, and mannosides to abrogate FimH-Upk1a interactions^21,22. In theory, based on findings in this study, therapies that temporarily prevent plaque assembly or block superficial urothelial cell differentiation may also effectively prevent UPEC infection. However, any therapeutic strategy targeting uroplakins carries potential risks considering the evolutionarily conserved role for uroplakins and the urothelial plaque in promoting urothelial integrity and barrier function. These considerations may limit the safety of such therapeutic strategies and pose challenges associated with their clinical application to human patients.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

ARJ: Performed immunofluorescence and electron microscopy. Experimental design, analysis, and interpretation of results. Manuscript writing.RL: Performed experimental UTI and enumeration of UPEC burden.ME: Performed RNA analysis. Experimental design, analysis, and interpretation of results.HC: Performed flow cytometry experiments. Experimental design, analysis, and interpretation of results.SG: Intracellular UPEC and IBC enumerationCC: Intracellular UPEC and IBC enumerationGB: Analysis of histopathology.XW: Analysis of RNAseq data, Manuscript writing.BB: Experimental design, analysis, and interpretation of results. Manuscript writing.

Funding 

This work was supported by the National Institutes of Health (NIDDK) R01 DK125469 (BB) and K01 DK126991 (ARJ).

Data availability

All raw and proceeded datasets of RNA-seq have been submitted to GEO (GSE324603). Prior to public release, the data can be accessed at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? acc=GSE324603 using the reviewer access token ipizaigqxvchbyx. This token allows anonymous, read-only access to GSE324603.

Code availability

We have deposited the complete RNAseq analysis pipeline and source code as the on Github (https://github.com/gucascau/uroplaque-uti) as the “uroplaque-uti-v1.1” software package. The underlying code was deposited on Zenodo (https://doi.org/10.5281/zenodo.20798286).

Declarations

Competing interests

The authors declare no competing interests.