A 5-HT-mediated urethral defense against urinary tract infections

Significance

Bacterial invasion of the urethra is the primary cause of urinary tract infections (UTIs), which rank among the most prevalent human infections worldwide. Efforts have focused on understanding how the bladder defends against infection, but the role of the urethra remains unclear. We show that the urethra actively defends against invading uropathogenic Escherichia coli. Neuroendocrine cells residing in the urethral epithelium are equipped to respond to bacteria and use serotonin to signal nearby cells, driving urethral contraction and bacterial expulsion. Deficiencies in this pathway could compromise urethral barrier defenses and contractility, contributing to urethral dysfunction and increasing susceptibility to recurrent UTIs, highlighting its relevance to human health.

Abstract

The urethra is considered a passive conduit for urine. Here, we reveal a surprising multicellular signaling pathway guiding the urethra’s dynamic response to an invading pathogen. Using a genetic approach in female mice, we deposited uropathogenic Escherichia coli into the distal urethra to establish a model of ascending urinary tract infection that progresses to the bladder within 4 h. We show that urethral neuroendocrine cells (UNECs), and the serotonin they synthesize, protect the bladder from bacterial colonization. We tested the hypothesis that serotonin initiates urethral contraction to expel ascending bacteria. We identified transient receptor potential cation channel subfamily A member 1, a noncanonical lipopolysaccharide receptor, in human and mouse UNECs and localized the serotonin receptors (HTR) 2B and 3, as well as the calcium-activated chloride channel anoctamin 1 (ANO1) to the pacemaker cells of the human and mouse urethra, the interstitial cells of Cajal (ICCs). HTR2B or ANO1 activation is sufficient for urethral contraction and is required for serotonin-induced mouse urethral contraction. Our results support the hypothesis that the urethra actively surveils its environment and responds to an ascending pathogen by evoking UNECs and ICC to induce urethral contraction and pathogen expulsion.

Body openings are gateways for foreign invaders (1, 2). Cells surrounding and within body openings deploy various strategies to safeguard against external threats. Mucosal epithelia, as dynamic barriers, actively exclude microbes (3). Cells within the mucosa respond to invading microbes by secreting chelators to starve them, releasing chemokines and cytokines to recruit immune cells, altering pH levels to curtail microbial growth, and generating free radicals to induce oxidative damage (4–7).

Peristalsis, often triggered by invading pathogens, acts as an additional defense mechanism for tissues near body openings. Isolates from Escherichia coli, Pseudomonas aeruginosa, and Klebsiella pneumoniae induce contraction of ureteral smooth muscle (8). Chlamydia trachomatis initiates uterine peristalsis (9). Resident microbes and invading pathogens trigger gut peristalsis (10, 11). Peristalsis is initiated and coordinated by multiple cell types. In the gut, enterochromaffin (EC) cells are armed with receptors to sense bacteria, viruses, and their products and respond by secreting multiple factors, including serotonin (5-HT), to mount a coordinated tissue response (11, 12). 5-HT receptors in the gut are widely expressed across multiple cell types including enteric nerves, smooth muscle cells (SMCs), and interstitial cells of Cajal (ICCs) (13, 14).

The urethra, which develops from similar origins as the gut, has many of the same cell types, including neuroendocrine cells that produce 5-HT (15, 16) and ICCs that regulate smooth muscle peristalsis (17, 18). Yet, it remains unclear whether the urethra actively mounts a defense against ascending microbes that drive urinary tract infection (UTI) and whether 5-HT plays a role in this response. Additional research is necessary because UTIs, most commonly caused by uropathogenic E. coli (UPEC) (19), represent a widespread and recurrent health issue among women, prompting healthcare visits, extensive antibiotic utilization, and notable economic burdens (20, 21). The surge in antibiotic resistance, fueled by widespread global antibiotic usage for UTIs, underscores the urgency for pioneering strategies in UTI management that transcend conventional antibiotic treatments (22).

Here, we identify urethral contraction as a defense mechanism initiated by ascending UPEC to clear the urinary tract of infection. Through a mouse model, we demonstrate that urethral neuroendocrine cells (UNECs) and 5-HT protect against ascending UPEC infection. We reveal that lipopolysaccharide (LPS) (outer membrane component of gram-negative bacteria) and 5-HT drive urethral contraction in vitro through a non-neurogenic smooth muscle mechanism mediated by transient receptor potential cation channel subfamily A member 1 (TRPA1), HTR2B, and anoctamin 1 (ANO1) which are localized to UNECs and ICCs in mouse and human female urethra. This research introduces potential targets for therapeutic intervention for the treatment or prevention of UTI.

Results

Neuroendocrine Cells and 5-HT in the Female Mouse Urethra Protect against Ascending UPEC Infection.

Gut and airway neuroendocrine cells mediate responses to commensal and pathogenic bacteria (23, 24). We hypothesized UNECs (15, 25, 26), respond to ascending pathogens to protect the urinary tract from infection. We used a genetic approach to ablate neuroendocrine cells from the mouse urethra by conditionally deleting achaete-scute achaete-scute family bHLH transcription factor 1 (Ascl1), a member of the notch signaling pathway required for neurodifferentiation (27), from Shh expressing progenitors. In a second cohort of mice, we ablated tryptophan hydroxylase 1 (Tph1) which catalyzes the rate-limiting step of 5-HT biosynthesis outside of the central nervous system (28). We used immunofluorescent staining to confirm loss of gene function (Fig. 1 C and D). ASCL1 and 5-HT copositive cells are present in the urethral epithelium of human and wild-type mouse females (Fig. 1 A and B) but absent from urethral epithelium of Ascl1 conditional null females (Fig. 1D). ASCL1 single positive cells, lacking detectable 5-HT, are present in the urethral epithelium of Tph1 null females (Fig. 1C).

Fig. 1.

Neuroendocrine cells and extraneuronal 5-HT protect against ascending UPEC infection. Five-micron female urethra tissue sections from (A and E) adult human, (B and F) wild-type mice, (C and G) Tph1 null mice, and (D and H) Ascl1 conditional null mice were labeled with antibodies against ASCL1 (green) and 5-HT (white) or mast cell tryptase (green) and 5-HT (white). DAPI nuclear staining is shown in blue. A white dashed line demarcates the epithelial (“ep”)-stromal (“st”) interface. (Scale bars are 100 µm.) Results are representative of three biological replicates. A catheter was placed into the distal urethra of adult female (I) Tph1 null, (J) Shh^Cre; Ascl1^fl/fl (Ascl1 conditional null), and age-matched control mice and E. coli UTI89 was delivered to initiate an ascending infection. Bladders were removed 3 h later and bacteria burden was determined by counting colonies on Kanamycin agar plates. Tph1 null mouse contained significantly more E. coli CFUs (n = 5; P = 0.0079) than control mouse bladders (n = 5). Ascl1 conditional null mouse bladders contained significantly more E. coli CFUs (n = 6; P = 0.0022) than control mouse bladders (n = 6). Limit of detection of 100 CFU/mL. Additionally, female urethra tissue (4 h postinfection) from (K and L) wild-type mice showed E. coli (transformed with green fluorescent protein plasmid to confer green protein fluoresce) form rod shape to a coccoid morphology in the bladder while (M and N) Tph1 null E. coli (green) form filamentous in the bladder demonstrating that 5-HT modified the complex behavior of E. coli during UTIs. Images L and N were captured at 40× magnification. (Scale bars are 100 µm.)

Mast cells can serve as a secondary source of 5-HT in the urethra. Mast cell tryptase and 5-HT double positive cells are present in urethral stroma of human, wild-type, and Ascl1 null mouse females (Fig. 1 E, F, and H), whereas mast cell tryptase single positive cells, lacking detectable 5-HT, are present outside of the urethral epithelium of Tph1 null females (Fig. 1G).

We conclude that urethras of wild-type mice and human females possess 5-HT-positive mast cells and neuroendocrine cells, located in the stromal and epithelial compartments, respectively. Ascl1 conditional null mice lack neuroendocrine cells but have 5-HT-positive mast cells, and Tph1 null mice have neuroendocrine cells and mast cells that do not synthesize 5-HT.

We next established a mouse model for ascending UTI by instilling 10 µL of E. coli UTI89 into the distal portion of the urethra of C57BL/6J (WT) female mice (SI Appendix, Fig. S1). The goal of this model was to establish an infection in the urethra which ascends over time to eventually colonize the bladder. We first validated our approach by instilling blue Davidson Tissue Dye through a PE-10 catheter 0.6 cm into the urethra of euthanized mice in accordance with a UW-Madison Animal Care and Use Committee approved protocol (SI Appendix, Fig. S1A). The dye was completely contained in the urethra and was not observed in the bladder. We used the same method to instill UPEC UTI89 (transformed to express GFP and confer kanamycin resistance) into the distal urethra of anesthetized mice. The mice were euthanized immediately after instillation or 4 h later, the urethra and bladder were carefully isolated and homogenized, and samples were plated on agar plates containing Kanamycin (50 µg/mL) and incubated for 24 h. Bacterial colony-forming units (CFUs) are observed exclusively in the urethra at 0 h posttransurethral instillation of UPEC, indicating precise delivery that did not extend to the bladder. Bacteria are detected in the bladder 4 h after instillation, confirming ascending infection (SI Appendix, Fig. S1 B and C).

To confirm that motility plays a critical role in the success of ascending infections, we conducted additional experiments using E. coli ΔflhDC, a nonmotile mutant strain. When this strain was transurethrally instilled in WT female mice, it resulted in a significantly lower number of CFUs in the bladder compared to those infected with wild-type E. coli UTI89 (Fig. 2A). This finding reinforces that in our model, bacterial motility is required for ascending infection into the bladder.

Fig. 2.

Transurethral instillation of wild-type E. coli UTI89 and ΔflhDC in WT mice and bladder infection severity following intravesical E. coli UTI89 instillation in WT and Tph1 null mice. (A) A catheter (0.6 cm, PE-10C) was placed transurethrally into the urethra of adult female wild-type mice. E. coli UTI89 (wild type) and ΔflhDC strains were instilled into the urethra to initiate infection. Bladders were collected 3 h postinstillation, and bacterial burden was assessed by counting colonies on agar plates. Bacterial CFUs were significantly greater in female mice instilled with wild-type E. coli (n = 4) compared to those instilled with the ΔflhDC strain (n = 4) (P = 0.0286). (B) No difference in bladder infection between wild-type and Tph1 null female mice 3 h after intravesical E. coli instillation. A catheter (1.2 cm, PE-10C) was placed transurethrally into the bladder of adult female Tph1 null mice and age-matched control (WT) mice. E. coli UTI89 was instilled to initiate an infection. Bladders were collected 3 h later, and bacterial burden was determined by counting colonies on Kanamycin agar plates. No statistically significant difference was found between Tph1 null mice (n = 3) and control mice (n = 3) (P > 0.9999).

We introduced an ascending UPEC infection to wild-type, Ascl1 conditional null and Tph1 null female mice and collected bladders 3 h later to quantify bacterial burden. Mice deficient in peripheral 5-HT synthesis (Tph1 null) had more severe bladder infections (median CFU = 1.5 × 10⁶, n = 5) than their respective wild-type control mice (median CFU = 6 × 10³, n = 5; P = 0.0079) (Fig. 1I). Mice deficient in neuroendocrine cells (Ascl1 conditional null) had more severe bladder infections (median CFU = 7.5 × 10⁵, n = 6) than their respective wild-type control mice (median CFU = 2.25 × 10³, n = 6; P = 0.0022) (Fig. 1J). These data support the hypothesis that UNECs and extraneuronal 5-HT produced by these and other cells protect the bladder from ascending UPEC infection. Mast cells express TPH1 in the bladder, while mast cells and neuroendocrine cells express TPH1 in the urethra. To test whether Tph1 is required in the urethra, bladder, or both to protect against infection, we next bypassed the urethra and delivered E. coli UTI89 directly into the bladders of wild-type and Tph1 null female mice. We observed no significant difference in CFUs between the groups, reinforcing that Tph1 in the urethra but not the bladder mediates protective actions against UPEC infection (Fig. 2B).

Research in mouse models indicates that UPEC replicate within bladder cells and escape to initiate a new infection cycle (29, 30). UPEC undergo morphological changes during this process transitioning from rod, to coccoid, to filamentous forms (31). The filamentous morphology predominates during UPEC escape from urothelial cells and reinfect urothelial cells after reverting to a rod-shaped morphology (30, 31). Using immunofluorescent staining against the GFP transgene in E. coli UTI89, we found, just 4 h after inoculation into the mouse urethra, that UPEC assumed filamentous structures in the bladders of Tph1 null mice but assumed a rod or coccoid morphology in WT mice (Fig. 1 K–N).

LPS and 5-HT Drive Urethral Contraction in Female Wild-Type Mice.

Airway and gut neuroendocrine cells respond to commensal and pathogenic bacteria by releasing 5-HT, which can drive tissue contraction (23, 24, 32). We hypothesized that bacteria may stimulate 5-HT secretion to drive urethral contraction, a mechanism that could expel bacteria from the urinary tract before reaching the bladder. Wild-type mouse urethra segments, approximately 2 mm in length, were mounted on a wire myograph to test their response to receptor-specific drugs. LPS, a cell wall component of gram-negative bacteria induces urethral contraction, with a maximal contraction of 11.25 ± 2.72 mN at 1.6 min. 5-HT also drives urethral contraction, with a maximal contraction of 9.87 ± 2.19 mN at 1.8 min (Fig. 3A). Dose–response stimulation employing increasing concentrations of 5-HT is depicted in Fig. 3B, illustrating the relationship between 5-HT concentration and urethral contraction force.

Fig. 3.

LPS and 5-HT initiate female muse urethral contraction. (A) Two-millimeter rings of wild-type C57BL/6J adult female mouse urethras were mounted on a wire myograph, emersed in KREBS buffer, and drugs introduced while measuring contraction force in milli-Newtons (mN). We observed that (A) 5-HT, and LPS, drive urethral contractions. (B) Dose–response stimulation employing increasing concentrations (0.3 to 30 µM) of 5-HT hydrochloride (EC50 = 5.399 µM). n = 3 tissues per group.

We considered several mechanisms by which 5-HT could induce urethral contraction. It could activate the 5-HT-sensitive spinal neurons that were shown previously to originate from Onuf’s nucleus and innervate urethral striated muscle or act directly on urethral striated muscle sphincter (33). 5-HT could also activate urethral smooth muscle contraction. Using selective inhibitors of each of these processes, we observed that 5-HT-induced contraction of the isolated urethra is insensitive to tetrodotoxin (TTX), a voltage-gated Na⁺ channel inhibitor known for blocking neurotransmission to induce nerve signal blockade (SI Appendix, Fig. S2B), and succinylcholine (SCh), a depolarizing neuromuscular blocking agent that affects cholinergic receptors in urethral striated muscle leading to muscle paralysis (SI Appendix, Fig. S3). However, 5-HT induced urethral contraction is sensitive to nifedipine (NIF), a Ca²⁺ channel blocker known to inhibit smooth muscle activity (SI Appendix, Fig. S3). We used bladder tissue as a control for the TTX experiment, to confirm that the TTX dose was sufficient to block a known axon-induced response. Bladder contraction was induced by electric field stimulation and inhibited by TTX (SI Appendix, Fig. S2A). We conclude that 5-HT induces contraction of isolated urethras by a mechanism that requires smooth muscle.

UNECs Express TRPA1, and LPS Drives Urethral Contraction In Vitro by a TRPA1-Dependent Mechanism.

We posited that bacterial LPS acts upstream of 5-HT to initiate smooth muscle contraction. To further elucidate the mechanism, we stained female mouse urethra for Toll-like receptor 4 (TLR4), the canonical LPS receptor (34). TLR4-positive cells are surprisingly absent from the wild-type control mouse urothelium, instead exclusively observed in the stroma (Fig. 4A). We considered that LPS may act through a noncanonical pathway and focused on transient receptor potential cation channel, subfamily A, member 1 (TRPA1), a noncanonical receptor recently localized to EC and airway epithelial cells (10, 11, 35) and which also recognizes LPS (35). TRPA1 immunofluorescence was present in mouse and human female urethral epithelial cells but not in cells outside the urethral epithelium. 5-HT⁺ cells are copositive for TRPA1, suggesting that neuroendocrine cells have the machinery to respond to LPS through a TRPA1-dependent mechanism (Fig. 4 B–E).

Fig. 4.

LPS-driven urethral contraction is possibly mediated by TRPA1 channels expressed by UNECs. Five-micron female urethra tissue sections from wild-type mice or adult humans were labeled with antibodies against (A) TLR4 (red), CDH1 (green), and 5-HT (white) or against TRPA1 (red) and 5-HT (white) in both mouse (B and C) and human (D and E) urethral sections. DAPI nuclear staining is shown in blue. A white dashed line demarcates the epithelial (ep)-stromal (st) interface. (Scale bars are 100 µm.) Results are representative of three biological replicates. Two-millimeter rings of wild-type C57BL/6J adult female mouse urethra were mounted on a wire myograph, emersed in KREBS buffer, and drugs introduced while measuring contraction force in millinewtons (mN). (F) LPS and the TRPA1 agonist AITC induced urethral contraction while (G) a TRPA1 antagonist blocked LPS-mediated contraction (500 µg/mL LPS + 10 µM HC030031). n = 3 tissues per group. The green box in images B and D corresponds to the boxed areas in images C and E, respectively.

Using isolated urethral rings from adult female mice, we found that TRPA1 stimulation with the selective agonist AITC is sufficient to induce urethral contraction in the absence of LPS or 5-HT (Fig. 4F, maximal contraction of 1.73 ± 1.39 mN at 0.7 ± 0.17 min). We also found that TRPA1 is required for LPS-induced urethral contraction by showing that the TRPA1 antagonist (HC030031) blocks LPS-induced urethral contraction in 0.29 ± 0.05 min (Fig. 4G).

5-HT Drives Urethral Contraction In Vitro through a Mechanism Mediated by HTR2B, HTR3, and ANO1 Expressed by ICCs.

Our next goal was to decipher the downstream mechanisms by which 5-HT drives urethral contraction. Gut EC cells release 5-HT to act on ICCs, which activate 5-HT receptors and calcium-activated Cl⁻ channel ANO1 on ICCs, leading to the contraction of smooth muscle myocytes to which ICCs are electrically coupled (36). We hypothesized a similar 5-HT-dependent mechanism links UNECs to ICCs and urethral smooth muscle myocytes. We conducted multiplex immunostaining to label ICCs (localized to stroma and defined by c-KIT-immunopositivity) and examined the cellular distribution patterns of HTR2B, HTR3, and ANO1 in the human and mouse female urethra. The human and mouse urethra harbors stromal cells that are triple-positive c-KIT, HTR2B, and HTR3 (Fig. 5 A and B). We also localized ANO1 and c-KIT to stromal cells near the smooth muscle of the urethra (Fig. 5 C–F). We conclude that c-KIT positive urethral stromal cells (presumptive ICCs), like ICCs of the gut, coexpress HTR2B, HTR3, and ANO1.

Fig. 5.

5-HT mediates female mouse urethral contraction potentially via its receptors (HTR2B and HTR3) expressed by ICCs, which depolarize smooth muscle through the ANO1 channel. Paraffin-embedded urethral sections (5 μm thickness) were examined at 40× magnification (green squares); the scale bar represents 100 μm. HTR2B (white) and HTR3 (green) receptors are expressed on ICCs (red) in both mouse (A) and human (B) urethral sections. ICCs (c-KIT in red) also express the ANO1 channel (green) in both species (C–F). Direct contractile responses of female mouse urethral rings were determined in an isolated organ system (wire myograph) over time (minutes). Contraction force is expressed in millinewtons (mN). Urethral contraction is driven by 5-HT, HTR2B agonist (BW723C86), and ANO1 agonist (Eact) (G), while it is inhibited by HTR2B antagonist (SB204742), HTR3 antagonist (Ondansetron hydrochloride), as well as ANO1 antagonist (Ani9) and ICC antagonist (Imatinib) (H–J). n = 3 tissues per group. The green box in images C and E corresponds to the boxed areas in images D and F, respectively.

Isolated adult female mouse urethral rings were used to test the role of HTR2B, HTR3, and ANO1 in urethral contraction. HTR2B activation with BW723C86 hydrochloride and ANO1 activation with Eact were sufficient to induce urethral contraction in the absence of LPS or 5-HT (Fig. 5G; BW723C86 maximal contraction of 2.73 ± 0.5 mN was observed at 1.08 ± 0.21 min, Eact maximal contraction of 2.27 ± 0.33 mN was observed at 1.33 ± 0.17 min). We also found that HTR2B, HTR3, ICCs, and ANO1 are necessary for 5-HT mediated urethral contraction by showing that the HTR2B antagonist SB204742, the HTR3 antagonist Ondansetron hydrochloride, the ANO1 inhibitor Ani9 and the tyrosine kinase inhibitor Imatinib that blocks ICC function (37), block 5-HT-mediated contractions (Fig. 5 H–J respectively). The HTR2B and HTR3 antagonists, the ANO1 antagonist, and Imatinib inhibited 5-HT-mediated urethral contraction within 0.33 ± 0.08, 0.33 ± 0.08, 1.33 ± 0.36, and 8.0 ± 1.58 min, respectively. We conclude that 5-HT requires ICCs and activates slow wave currents driving urethral contraction.

Fig. 6 represents a hypothesized mechanism for 5-HT mediated defense against UTI in the urethra, in which UNECs sense Gram-negative bacteria LPS through the TRPA1 channel, leading to calcium influx and 5-HT release. 5-HT enhances calcium and electric activities of ICC via HTR2B and HTR3 receptors. ANO1 channels in urethral ICC depolarize SMCs, increasing their excitability for pathogen elimination.

Fig. 6.

5-HT-mediated defense against UTIs in the urethra. UNECs, equipped with the TRPA1 channel, play a pivotal role in sensing Gram-negative bacteria (E. coli) LPS. Activation of the TRPA1 channel within UNECs leads to calcium influx and subsequent release of 5-HT (5-HT). 5-HT enhances both calcium and electric activities of ICC via HTR2B and HTR3 receptors. Ano1 channels, known as calcium-activated chloride channels, are expressed in urethral ICC. Upon binding of calcium, Ano1 channels undergo a conformational change, allowing chloride ions to move down their electrochemical gradient out of the cell, contributing to cell depolarization. The depolarization signal is propagated to SMCs via gap junctions. Subsequently, the depolarization triggers the opening of L-type calcium channels, leading to calcium influx and smooth muscle contraction, ultimately contributing to the elimination of pathogens.

Discussion

UTIs pose a substantial health burden, particularly to women, and are a growing worldwide threat due to the emergence of antibiotic-resistant pathogens (2, 19, 22). Considering the ascending route of UTI and the anatomical proximity of the urethra and vagina, the vagina could serve as a source from which UPEC ascend to the urethra and bladder, influencing the dynamics of UTI development and recurrence (38, 39). Improved understanding of host–pathogen interactions has the potential to reveal new therapeutic targets. Challenging the conventional perception of the urethra as a mere conduit for urine, we show that it orchestrates an elegant defense against invading UPEC, the primary pathogen in human UTI. We describe a multicellular signaling pathway triggered by UPEC and implicating neuroendocrine cells and 5-HT in urethral contraction, which effectively expel ascending pathogens (Fig. 6). This mechanism advances our understanding of UTI pathogenesis and may also function as a target for therapeutics while potentially shedding light on other conditions such as overactive bladder and urinary incontinence. Serotonin-induced ICC overactivity could increase urethral tone, impairing urine outflow from the bladder. Additionally, if bladder ICCs embedded in the detrusor (40) are also responsive to serotonin, they may influence bladder responses during filling and emptying, potentially contributing to bladder dysfunction, such as underactive or overactive bladder. Reduced serotonin levels may impair ICC activation in the detrusor muscle, disrupting bladder contractions and potentially leading to overactive bladder. Additionally, serotonin’s role in both detrusor smooth muscle and the external urethral sphincter suggests that its deficiency could increase susceptibility to incontinence, particularly during physical stress, as seen in stress urinary incontinence. Understanding the role of serotonin derived from neuroendocrine cells and its interaction with ICCs could also help elucidate these conditions, which are often linked to UTIs.

We were uniquely positioned to find that UNEC depletion in Ascl1 conditional null mice heightened susceptibility to ascending UPEC bladder infection. Most models of mouse UTI involve deposition of pathogens directly into the bladder via a catheter (41, 42). While this approach is useful for understanding interactions between bladder cells and pathogens, it bypasses the urethra, leaving contributions of the urethral cells unknown. By establishing a model of ascending UTI that expands from the distal urethra into the bladder within 3 h, we found that UNECs (Fig. 1I), and the 5-HT they synthesize (Fig. 1J), combat bladder infection. This model will be useful for examining whether additional pathogens, beyond UPEC, trigger urethral responses, for testing whether the urethra initiates a sustained adaptive response, and whether pathogens develop methods to evade urethral defenses.

We were surprised that 5-HT influences UPEC phenotype in the bladder. We observed rod-shaped UPEC in the WT mouse and filamentous UPEC in Tph1 null mouse bladders. Filamentation is a crucial survival strategy employed by various bacteria when responding to environmental stimuli. This process plays a significant role in enhancing bacterial survival during exposure to environmental stresses (43, 44). Filamentous E. coli UTI89 formed far more quickly in Tph1 null mice than in a previous study by Justice et al., in which they infected adult female C3H/HeJ mice with E. coli (31). A missense TLR4 mutation renders C3H/HeJ mice (45) less responsive to LPS, making them more susceptible to infection by E. coli UTI89 than the C57BL6/J mice used in this study. Justice et al. reported cocci E. coli UTI89 at 6 h postinfection and did not observe filamentous E. coli until 16 h postinfection (31). One possible conclusion is that the lack of peripheral 5-HT in Tph1 null mice may accelerate the transition of UPEC to its filamentous form, possibly serving as an additional mechanism to evade host innate immune responses (31). Kumar et al. demonstrated that alterations in intestinal 5-HT concentrations impact the pathogenesis of enteric bacteria, specifically decreasing virulence gene expression by enterohemorrhagic E. coli (46). Other studies have found that inhibition of 5-HT synthesis increases bacterial pathogenesis and decreases host survival (46). While our study was not specifically designed to examine the impact of 5-HT on UPEC, it is possible that 5-HT acts directly or indirectly on UPEC, a notion supported by the presence of 5-HT receptors on enterohemorrhagic E. coli (46–48).

5-HT+ UNECs were identified long ago (15, 16, 26) but their physiological significance is less clear than that of the more extensively studied intestinal EC cells and airway neuroepithelial cells (11, 49–51). EC cells express TRPA1, which respond to various stimuli, including environmental irritants (52), inflammatory mediators (11), and microbial metabolites (10), enhancing intestinal motility through calcium influx and 5-HT release (10, 52, 53). Ye et al. demonstrated that TRPA1 activation by Edwardsiella tarda enhances intestinal motility for parasite clearance and supports gut health (10). Intestinal contractions mediated by EC TRPA1 are dependent on HTR3 (52). In the airway, TRPA1 activation by LPS activates human bronchial epithelial cells, challenging the previous assumption that TLR4 was the sole pattern recognition receptor for LPS (35). Additionally, Meseguer et al. demonstrated that peripheral sensory afferents expressing TRPA1 are directly stimulated by LPS in a manner independent of TLR4 (54). Using isolated preparations of mouse urethra, we found that LPS induces urethral contractions through a TRPA1-dependent mechanism. We also found that 5-HT induces urethral contraction by a mechanism requiring HTR2B, HTR3, and apparent ICC function. While 5-HT was previously shown to mediate urethra contraction by activating pudendal nerves and the external urethral sphincter (55, 56), our findings support the notion that 5-HT also acts locally on the urethra to control ICC activity and smooth muscle tone (17, 18, 57). Our findings, in the broader context of previous studies in the gut and airway, raise the hypothesis that TRPA1 may function broadly as a pathogen sensor and act through 5-HT to drive invading pathogens from body openings.

In summary, our findings elevate the significance of UNECs in urinary health and suggest UNECs are functionally connected to UPEC through TRPA1 and to ICCs and smooth muscle through a 5-HT-dependent pathway (Fig. 6). The mechanism we propose is important for future therapies because the receptors are druggable, and because functional impairment in this pathway may underlie susceptibility to UTI or recurrent UTI.

Materials and Methods

Drugs.

5-HT hydrochloride (cat. # 153-98-0), Nifedipine (cat. # 501793791), Succinylcholine (cat. # S8251, batch #MKCR4927), Imatinib Mesylate (cat. # SML1027, batch #137544), and LPS from E. coli O111:B4 (cat. # L2630, batch #0000155607) obtained from Sigma-Aldrich Chem Co. Selective 5-HT2B receptor agonist (BW723C86 hydrochloride; sc-203428) obtained from Santa Cruz Biotechnology. Selective 5-HT2B receptor antagonist (SB20474, cat. # 1372, batch 3A/262331), activator of TMEM16A (ANO1) (Eact, cat. # 4876, batch #1), selective ANO1 blocker (Ani9, cat. # 6076, batch #1), Selective 5-HT3 antagonist (Ondansetron hydrochloride, cat. # 2891, batch #1), TRPA1 channel blocker (HC030031, cat. # 2896, batch #4), and TTX (cat. # 1078, batch #51) were purchased from Tocris Bioscience.

Human Tissues.

Samples of urethral tissue obtained from organ donors. Family members of organ donors at the Southwest Transplant Alliance provided consent for the use of tissue for research under internal approval and specimens received were deidentified except for age, sex, and ethnicity prior to receipt. Use of tissue from deidentified decedents is approved for study at University of Texas Southwestern under IRB STU 112014-033. Donor IDs: D158 (42 y old), D162 (25 y old), D212 (23 y old).

Animals.

All experiments were conducted in strict adherence to the approved protocol from the University of Wisconsin Animal Care and Use Committee, following the guidelines outlined in the NIH Guide for the Care and Use of Laboratory Animals. All mice were from Jackson Laboratories (Bar Harbor, ME) and maintained on a C57BL/6J genetic background (Jackson stock #000664). Strains included: Tph1^tm1Bdr (also known as Tph1 null) (58), B6.Cg-Shhtm1(EGFP/cre)Cjt/J (Shhgfpcre Jackson Lab #005622) (59) backcrossed at least three generations with Ascl1^tm2Fgu (Ascl1^flox Jackson Lab #037042) (60) (also known as Ascl1 null). All subjects used in this study were sexually mature adult females between 6 to 8 wk of age.

Mice were housed in Udel® Polysulfone microisolator cages placed on racks or in Innocage® disposable mouse cages on an Innorack®. Room lighting was maintained on a 12 h light–dark cycle, with room temperature set at a consistent 20.5 ± 5 °C and humidity ranging from 30 to 70%. The mice were provided with 8604 Teklad Rodent Diet (Harlan Laboratories, Madison, WI) and had access to feed and water ad libitum. Cages were furnished with corn cob bedding.

Female Mouse Model for Transurethral Instillation—Dye Instillation.

To determine the anatomic regions of the urethra/bladder into which instillation fluid distributes, we tested mice of multiple ages and genetic backgrounds (including Tph1 null, Ascl1 null and C57BL/6J female mice) to follow the 3Rs of humane animal research and reduce animals used for this study. All female mice were transurethrally catheterized at 6 to 8 wk of age.

Mice were euthanized, a transurethral catheter (0.6 cm, PE-10) was placed, and blue Davidson Tissue Dye was instilled. The catheter was meticulously introduced 3 mm into the external urethral ostium, aligning with the distal urethral axis in a cranial to caudal direction. Subsequently, the catheter was advanced an additional 3 mm into the urethra, totaling 6 mm of insertion in the final step.

Given the anatomical horseshoe shape of rodent’s urethra, we identified key points for successful urethral catheterization, emphasizing the importance of gentle movements and catheter lubrication to minimize the risk of urethral injuries. The objective was to insert the catheter and instill blue dye into the urethra without breaching the bladder.

E. coli.

The experiments utilized E. coli UTI89 (61), a UPEC originally isolated from a human patient with cystitis (62). The bacteria were modified with a pCOMGFP plasmid (63) to enable the expression of green fluorescent protein and confer kanamycin resistance. UTI89 was obtained from J. Barasch (Columbia University) and originally reported (62). Additionally, the E. coli mutant strain UTI89 ∆flhDC, which lacks the master regulator of flagellar motility, flhDC, was used as previously described (64). Before inoculation, E. coli UTI89 was cultured statically for 18 h in antibiotic-free Luria-Bertani broth at 37 °C. The optical density (OD) was measured before inoculating mice. Subsequently, the culture was centrifuged for 15 min at 1,157 rcf, and the resulting pellet was suspended in sterile phosphate buffered saline (PBS) for instillation.

Transurethral Instillation of Sterile Saline or E. coli.

At 6 to 8 wk of age, female C57BL/6J, Tph1 null, and Ascl1 null mice underwent anesthesia with isoflurane, a transurethral catheter (0.6 cm, PE-10) was placed and sterile PBS (control) or PBS containing 10 µL of E. coli UTI89 (OD = 0.8) were instilled.

Bladder/Urethra Tissue Culture.

Mice were euthanized at fixed time intervals (0, 3, and 4 h after instillation), placed on sterile surgical drape, and drenched in 70% ethanol. Because our goal was to determine the distribution and severity of E. coli infection, great care was taken during dissection to reduce the risk of cross-contaminating organs or surgical instruments. Sterile scissors were used to open the body cavity and expose the lower urinary tract (bladder and urethra). Intestines were moved aside with a sterile gloved finger. The female urethra was lying immediately to the vagina is approximately 11 mm in length and opens independently of the vagina, it passes through and almost adjacent to the pubic symphysis, from which it is separated by a small amount of adipose tissue. The suture was placed in the bladder neck, and then, the bladder and urethra were dissected by cutting off the adipose tissue and pubic symphysis. The bladder and urethra were separated at the level of the bladder neck with a single transverse cut. The bladder and urethra were placed in a sterile 1.5-mL tube with 1 mL of sterile PBS and hand-ground with a sterile pestle and vortex. Serial dilutions were made and plated on LB agar containing kanamycin (50 µg/mL). Scissors and forceps were sterilized between mice using a bead sterilizer, and a new sterile surgical drape, sterile suture, and sterile PBS were used for each mouse. CFU values were calculated per milliliter.

Quantitative Cultures.

Tissue homogenates were 10-fold serially diluted in 1,000 µL sterile saline and 10 µL were cultured on agar plates. Plates were incubated at 37 °C for 18 to 24 h and CFUs were counted for each sample (65, 66).

Tissue Preparation, Fixation, and Sectioning.

Bladder and urethral tissue were removed. Bladder and urethras were fixed in 4% paraformaldehyde and dehydrated in ethanol, cleared in xylene, and infiltrated with paraffin as previously described (67). Urethras were serially sectioned starting at the bladder neck. Five-micrometer-thick sections were mounted on Superfrost Plus Gold Slides (ThermoFisher Scientific).

Immunohistochemistry (IHC).

IHC was conducted on female urethral tissue sections from mice and humans. Tissue sections were deparaffinized with xylene and hydrated through a series of graded ethanol solutions. Tissues were immersed in citrate buffer pH 6.0 and heated in a microwave for epitope decloaking. Tris-buffered saline containing 0.1% Tween 20 and 5% donkey serum was used as a blocking reagent, and primary and secondary antibodies were diluted in blocking reagent. Antibodies and dilutions are listed in the SI Appendix, Table S1. Tissue sections were imaged using an Eclipse E600 compound microscope (Nikon Instruments Inc., Melville, NY) fitted with a 20× dry objective (Plan Fluor NA = 0.75; Nikon, Melville, NY) and equipped with NIS elements imaging software (Nikon Instruments Inc.). Fluorescence was detected using 4’,6-diamidino-2-phenylindole (DAPI), Fluorescein isothiocyanate, Texas Red (Chroma Technology Corp, Bellows Fall, VT), and CY5 filter cubes (Nikon, Melville, NY).

In Vitro Urethra Preparation and Functional Studies.

Female C57BL/6J were euthanized in the CO₂ chamber, and the bladder and urethra tissues were removed as a single block to a petri dish containing Krebs solution (NaCl, NaHCO₃, KCl, MgCl, NaH₂PO₄, CaCl₂, and d-glucose). After careful removal of fat, a 2-mm ring of the urethra, located below the bladder neck was obtained. Urethral rings were mounted under a resting tension of 2 mN in a 5-mL wire myograph containing Krebs solution at 37 °C (pH 7.4, 95%O₂/5%CO₂). Isometric force was recorded using a PowerLab 400™ Data Acquisition System (Software Chart, version 6.0, AD Instrument, Milford, MA). Urethral rings were allowed to equilibrate for 1 h before starting the experiments. Tissues were then washed with Krebs solution and a maximal response was generated by adding 60 mM KCl to the chambers. A total of three urethras were tested for each data point.

After three washes with Krebs solution, increasing concentrations (0.3 to 30 µM) of 5-HT hydrochloride was added. Tissues were then washed with Krebs solution to return the tension to baseline. This step was repeated for Selective 5-HT2B receptor agonist BW723C86 (1 to 100 µM) (68), TRPA1 agonist AITC (100 µM) and ANO1 agonist Eact (10 to 50 µM) (69).

Then, urethral rings tensioned with 5-HT (30 µM) and exposed to the HTR2B antagonist SB 204741 (0.3 to 1 µM) (70, 71), Selective 5-HT3 antagonist Ondansetron hydrochloride (10 µM) (72), the ANO1 antagonist Ani9 (1 to 10 µM) (73, 74), and the ICC inhibitor Imatinib (100 µM) (75, 76).

Therefore, we measured contractile responses of female mouse urethra to LPS (500 μg/mL). Urethral rings were then exposed to the TRPA1 antagonist HC030031 (0.1 to 10 µM).

To deepen understanding of the mechanisms governing urethral contraction, we conducted supplementary experiments. A total of two urethras were tested for each data point. Specifically, we investigated the effects of TTX (1 µM) (77) combined with 5-HT and succinylcholine (2 mM) (78) combined with 5-HT to discern the roles of nerves and striated muscle in driving contractions, respectively. Additionally, we administered nifedipine (30 µM) (79) combined with 5-HT (to block smooth muscle activity). The TTX studies on bladder strips and urethra were done in a tissue bath setup, at the same time, as previously described (80). Briefly, bladder strips were maintained at 1 g of tension and urethras at 0.2 g of tension for 1 h. Electrical field stimulation was then applied to bladder strips (1, 2, 5, 10, 20, and 50 Hz) with 3 min recovery between each stimulation. Serotonin (30 µM) was applied to urethra strips and after the maximal contractile response was achieved, urethral strips were washed with Krebs solution. TTX (1 µM) was then added to bladder and urethra tissues, serotonin was added to urethral tissues and EFS applied to bladder tissues. All tissues were then washed, and KCL (60 mM) was added to confirm that tissues were still responsive at the end of the experiment.

Statistical Analysis.

Statistical analyses were conducted using GraphPad Prism 8.0.2, with significance set at P < 0.05. Bladder infection data were assessed for significance using Student’s t test, with five mice per group for Tph1 null and control groups and six mice per group for Ascl1 and control groups. Normality was evaluated using Anderson–Darling (A2*), D’Agostino–Pearson omnibus (K2), Shapiro–Wilk (W), and Kolmogorov–Smirnov (distance) tests. When the data deviated from normal distribution, further analysis was conducted using the Mann–Whitney test. In vitro study data were analyzed using nonlinear regression for fitting.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All data needed to evaluate the conclusions of the paper are present in the paper, SI Appendix, or have been deposited in Figshare (https://doi.org/10.6084/m9.figshare.28687334) (81).